[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"project-9671":3},{"id":4,"name":5,"fullName":6,"owner":7,"repo":5,"description":8,"homepage":9,"htmlUrl":10,"language":10,"languages":10,"totalLinesOfCode":10,"stars":11,"forks":12,"watchers":13,"openIssues":14,"contributorsCount":14,"subscribersCount":14,"size":14,"stars1d":14,"stars7d":15,"stars30d":16,"stars90d":14,"forks30d":14,"starsTrendScore":17,"compositeScore":18,"rankGlobal":10,"rankLanguage":10,"license":19,"archived":20,"fork":20,"defaultBranch":21,"hasWiki":22,"hasPages":20,"topics":23,"createdAt":10,"pushedAt":10,"updatedAt":40,"readmeContent":41,"aiSummary":42,"trendingCount":14,"starSnapshotCount":14,"syncStatus":43,"lastSyncTime":44,"discoverSource":45},9671,"techniques","satellite-image-deep-learning\u002Ftechniques","satellite-image-deep-learning","Techniques for deep learning with satellite & aerial imagery","",null,10173,1642,279,0,9,37,5,44.65,"Apache License 2.0",false,"master",true,[24,25,26,27,28,29,30,31,32,33,34,35,36,37,38,39],"convolutional-neural-networks","dataset","datasets","deep-learning","deep-neural-networks","earth-observation","image-classification","machine-learning","object-detection","python","pytorch","remote-sensing","satellite-data","satellite-imagery","satellite-images","sentinel","2026-06-12 02:02:11","\u003Cdiv align=\"center\">\n  \u003Cp>\n    \u003Ca href=\"https:\u002F\u002Fwww.satellite-image-deep-learning.com\u002F\">\n        \u003Cimg src=\"images\u002Flogo.png\" width=\"700\">\n    \u003C\u002Fa>\n\u003C\u002Fp>\n\n# 👉 [satellite-image-deep-learning.com](https:\u002F\u002Fwww.satellite-image-deep-learning.com\u002F) 👈\n\n\u003C\u002Fdiv>\n\n## Introduction\n\nDeep learning has revolutionized the analysis and interpretation of satellite and aerial imagery, addressing unique challenges such as vast image sizes and a wide array of object classes. This repository provides an exhaustive overview of deep learning techniques specifically tailored for satellite and aerial image processing. It covers a range of architectures, models, and algorithms suited for key tasks like classification, segmentation, and object detection.\n\n**How to use this repository:** use `Command + F` (Mac) or `CTRL + F` (Windows) to search this page for e.g. 'SAM'\n\n## Techniques\n\n- [Classification](https:\u002F\u002Fgithub.com\u002Fsatellite-image-deep-learning\u002Ftechniques?tab=readme-ov-file#classification)\n- [Segmentation](https:\u002F\u002Fgithub.com\u002Fsatellite-image-deep-learning\u002Ftechniques?tab=readme-ov-file#segmentation)\n- [Object detection](https:\u002F\u002Fgithub.com\u002Fsatellite-image-deep-learning\u002Ftechniques?tab=readme-ov-file#object-detection)\n- [Regression](https:\u002F\u002Fgithub.com\u002Fsatellite-image-deep-learning\u002Ftechniques?tab=readme-ov-file#regression)\n- [Cloud detection & removal](https:\u002F\u002Fgithub.com\u002Fsatellite-image-deep-learning\u002Ftechniques?tab=readme-ov-file#cloud-detection--removal)\n- [Change detection](https:\u002F\u002Fgithub.com\u002Fsatellite-image-deep-learning\u002Ftechniques?tab=readme-ov-file#change-detection)\n- [Time series](https:\u002F\u002Fgithub.com\u002Fsatellite-image-deep-learning\u002Ftechniques?tab=readme-ov-file#time-series)\n- [Crop classification](https:\u002F\u002Fgithub.com\u002Fsatellite-image-deep-learning\u002Ftechniques?tab=readme-ov-file#crop-classification)\n- [Crop yield & vegetation forecasting](https:\u002F\u002Fgithub.com\u002Fsatellite-image-deep-learning\u002Ftechniques?tab=readme-ov-file#crop-yield--vegetation-forecasting)\n- [Generative networks](https:\u002F\u002Fgithub.com\u002Fsatellite-image-deep-learning\u002Ftechniques?tab=readme-ov-file#generative-networks)\n- [Autoencoders, dimensionality reduction, image embeddings & similarity search](https:\u002F\u002Fgithub.com\u002Fsatellite-image-deep-learning\u002Ftechniques?tab=readme-ov-file#autoencoders-dimensionality-reduction-image-embeddings--similarity-search)\n- [Few & zero shot learning](https:\u002F\u002Fgithub.com\u002Fsatellite-image-deep-learning\u002Ftechniques?tab=readme-ov-file#few--zero-shot-learning)\n- [Self-supervised, unsupervised & contrastive learning](https:\u002F\u002Fgithub.com\u002Fsatellite-image-deep-learning\u002Ftechniques?tab=readme-ov-file#self-supervised-unsupervised--contrastive-learning)\n- [SAR](https:\u002F\u002Fgithub.com\u002Fsatellite-image-deep-learning\u002Ftechniques?tab=readme-ov-file#sar)\n- [Explainable Ai (XAI)](https:\u002F\u002Fgithub.com\u002Fsatellite-image-deep-learning\u002Ftechniques?tab=readme-ov-file#explainable-ai-xai)\n- [Large vision & language models (LLMs & LVMs)](https:\u002F\u002Fgithub.com\u002Fsatellite-image-deep-learning\u002Ftechniques?tab=readme-ov-file#large-vision--language-models-llms--lvms)\n- [Foundational models](https:\u002F\u002Fgithub.com\u002Fsatellite-image-deep-learning\u002Ftechniques?tab=readme-ov-file#foundational-models)\n\n#\n## Classification\n\n\u003Cp align=\"center\">\n  \u003Cimg src=\"images\u002Fmerced.png\" width=\"600\">\n  \u003Cbr>\n  \u003Cb>The UC merced dataset is a well known classification dataset.\u003C\u002Fb>\n\u003C\u002Fp>\n\nClassification is a fundamental task in remote sensing data analysis, where the goal is to assign a semantic label to each image, such as 'urban', 'forest', 'agricultural land', etc. The process of assigning labels to an image is known as image-level classification. However, in some cases, a single image might contain multiple different land cover types, such as a forest with a river running through it, or a city with both residential and commercial areas. In these cases, image-level classification becomes more complex and involves assigning multiple labels to a single image. This can be accomplished using a combination of feature extraction and machine learning algorithms to accurately identify the different land cover types. It is important to note that image-level classification should not be confused with pixel-level classification, also known as semantic segmentation. While image-level classification assigns a single label to an entire image, semantic segmentation assigns a label to each individual pixel in an image, resulting in a highly detailed and accurate representation of the land cover types in an image. Read [A brief introduction to satellite image classification with neural networks](https:\u002F\u002Fmedium.com\u002F@robmarkcole\u002Fa-brief-introduction-to-satellite-image-classification-with-neural-networks-3ce28be15683)\n\n\n- [EuroSat-Satellite-CNN-and-ResNet](https:\u002F\u002Fgithub.com\u002FRumeysakeskin\u002FEuroSat-Satellite-CNN-and-ResNet) -> Classifying custom image datasets by creating Convolutional Neural Networks and Residual Networks from scratch with PyTorch\n\n- [Detecting Informal Settlements from Satellite Imagery using fine-tuning of ResNet-50 classifier](https:\u002F\u002Fblog.goodaudience.com\u002Fdetecting-informal-settlements-using-satellite-imagery-and-convolutional-neural-networks-d571a819bf44) with [repo](https:\u002F\u002Fgithub.com\u002Fdymaxionlabs\u002Fap-latam)\n\n-  [Land-Cover-Classification-using-Sentinel-2-Dataset](https:\u002F\u002Fgithub.com\u002Fraoofnaushad\u002FLand-Cover-Classification-using-Sentinel-2-Dataset) -> [well written Medium article](https:\u002F\u002Fraoofnaushad7.medium.com\u002Fapplying-deep-learning-on-satellite-imagery-classification-5f2588b932c1) accompanying this repo but using the EuroSAT dataset\n\n- [Slums mapping from pretrained CNN network](https:\u002F\u002Fgithub.com\u002Fdeepankverma\u002Fslums_detection) on VHR (Pleiades: 0.5m) and MR (Sentinel: 10m) imagery\n\n- [Comparing urban environments using satellite imagery and convolutional neural networks](https:\u002F\u002Fgithub.com\u002Fadrianalbert\u002Furban-environments) -> includes interesting study of the image embedding features extracted for each image on the Urban Atlas dataset\n\n- [RSI-CB](https:\u002F\u002Fgithub.com\u002Flehaifeng\u002FRSI-CB) -> A Large Scale Remote Sensing Image Classification Benchmark via Crowdsource Data. See also [Remote-sensing-image-classification](https:\u002F\u002Fgithub.com\u002Faashishrai3799\u002FRemote-sensing-image-classification)\n\n- [WaterNet](https:\u002F\u002Fgithub.com\u002Ftreigerm\u002FWaterNet) -> a CNN that identifies water in satellite images\n\n- [Road-Network-Classification](https:\u002F\u002Fgithub.com\u002Fualsg\u002FRoad-Network-Classification) -> Road network classification model using ResNet-34, road classes organic, gridiron, radial and no pattern\n\n- [SSTN](https:\u002F\u002Fgithub.com\u002Fzilongzhong\u002FSSTN) -> Spectral-Spatial Transformer Network for Hyperspectral Image Classification: A FAS Framework\n\n- [SatellitePollutionCNN](https:\u002F\u002Fgithub.com\u002Farnavbansal1\u002FSatellitePollutionCNN) -> A novel algorithm to predict air pollution levels with state-of-the-art accuracy using deep learning and GoogleMaps satellite images\n\n- [PropertyClassification](https:\u002F\u002Fgithub.com\u002FSardhendu\u002FPropertyClassification) -> Classifying the type of property given Real Estate, satellite and Street view Images\n\n- [remote-sense-quickstart](https:\u002F\u002Fgithub.com\u002FCarryHJR\u002Fremote-sense-quickstart) -> classification on a number of datasets, including with attention visualization\n\n- [IGARSS2020_BWMS](https:\u002F\u002Fgithub.com\u002Fjiankang1991\u002FIGARSS2020_BWMS) -> Band-Wise Multi-Scale CNN Architecture for Remote Sensing Image Scene Classification with a novel CNN architecture for the feature embedding of high-dimensional RS images\n\n- [image.classification.on.EuroSAT](https:\u002F\u002Fgithub.com\u002Fcanturan10\u002Fimage.classification.on.EuroSAT) -> solution in pure pytorch\n\n- [hurricane_damage](https:\u002F\u002Fgithub.com\u002Fallankapoor\u002Fhurricane_damage) -> Post-hurricane structure damage assessment based on aerial imagery\n\n- [ISPRS_S2FL](https:\u002F\u002Fgithub.com\u002Fdanfenghong\u002FISPRS_S2FL) -> Multimodal Remote Sensing Benchmark Datasets for Land Cover Classification with A Shared and Specific Feature Learning Model\n\n- [ensemble_LCLU](https:\u002F\u002Fgithub.com\u002Fburakekim\u002Fensemble_LCLU) -> Deep neural network ensembles for remote sensing land cover and land use classification\n\n- [Urban-Analysis-Using-Satellite-Imagery](https:\u002F\u002Fgithub.com\u002Fmominali12\u002FUrban-Analysis-Using-Satellite-Imagery) -> classify urban area as planned or unplanned using a combination of segmentation and classification\n\n- [mining-discovery-with-deep-learning](https:\u002F\u002Fgithub.com\u002Fremis\u002Fmining-discovery-with-deep-learning) -> Mining and Tailings Dam Detection in Satellite Imagery Using Deep Learning\n\n- [sentinel2-deep-learning](https:\u002F\u002Fgithub.com\u002Fd-smit\u002Fsentinel2-deep-learning) -> Novel Training Methodologies for Land Classification of Sentinel-2 Imagery\n\n- [Pay-More-Attention](https:\u002F\u002Fgithub.com\u002Fwilliamzhao95\u002FPay-More-Attention) -> Remote Sensing Image Scene Classification Based on an Enhanced Attention Module\n\n- [Remote Sensing Image Classification via Improved Cross-Entropy Loss and Transfer Learning Strategy Based on Deep Convolutional Neural Networks](https:\u002F\u002Fgithub.com\u002FAliBahri94\u002FRemote-Sensing-Image-Classification-via-Improved-Cross-Entropy-Loss-and-Transfer-Learning-Strategy)\n\n- [SKAL](https:\u002F\u002Fgithub.com\u002Fhw2hwei\u002FSKAL) -> Looking Closer at the Scene: Multiscale Representation Learning for Remote Sensing Image Scene Classification\n\n- [SAFF](https:\u002F\u002Fgithub.com\u002Fzh-hike\u002FSAFF) -> Self-Attention-Based Deep Feature Fusion for Remote Sensing Scene Classification\n\n- [GLNET](https:\u002F\u002Fgithub.com\u002Fwuchangsheng951\u002FGLNET) -> Convolutional Neural Networks Based Remote Sensing Scene Classification under Clear and Cloudy Environments\n\n- [Remote-sensing-image-classification](https:\u002F\u002Fgithub.com\u002FhiteshK03\u002FRemote-sensing-image-classification) -> transfer learning using pytorch to classify remote sensing data into three classes: aircrafts, ships, none\n\n- [remote_sensing_pretrained_models](https:\u002F\u002Fgithub.com\u002Flsh1994\u002Fremote_sensing_pretrained_models) -> as an alternative to fine tuning on models pretrained on ImageNet, here some CNN are pretrained on the RSD46-WHU & AID datasets\n\n- [OBIC-GCN](https:\u002F\u002Fgithub.com\u002FCVEO\u002FOBIC-GCN) -> Object-based Classification Framework of Remote Sensing Images with Graph Convolutional Networks\n\n- [aitlas-arena](https:\u002F\u002Fgithub.com\u002Fbiasvariancelabs\u002Faitlas-arena) -> An open-source benchmark framework for evaluating state-of-the-art deep learning approaches for image classification in Earth Observation (EO)\n\n- [droughtwatch](https:\u002F\u002Fgithub.com\u002Fwandb\u002Fdroughtwatch) -> Satellite-based Prediction of Forage Conditions for Livestock in Northern Kenya\n\n- [JSTARS_2020_DPN-HRA](https:\u002F\u002Fgithub.com\u002FB-Xi\u002FJSTARS_2020_DPN-HRA) -> Deep Prototypical Networks With Hybrid Residual Attention for Hyperspectral Image Classification\n\n- [SIGNA](https:\u002F\u002Fgithub.com\u002Fkyle-one\u002FSIGNA) -> Semantic Interleaving Global Channel Attention for Multilabel Remote Sensing Image Classification\n\n- [PBDL](https:\u002F\u002Fgithub.com\u002FUsman1021\u002FPBDL) -> Patch-Based Discriminative Learning for Remote Sensing Scene Classification\n\n- [EmergencyNet](https:\u002F\u002Fgithub.com\u002Fckyrkou\u002FEmergencyNet) -> identify fire and other emergencies from a drone\n\n- [satellite-deforestation](https:\u002F\u002Fgithub.com\u002Fdrewhibbard\u002Fsatellite-deforestation) -> Using Satellite Imagery to Identify the Leading Indicators of Deforestation, applied to the Kaggle Challenge Understanding the Amazon from Space\n\n- [RSMLC](https:\u002F\u002Fgithub.com\u002Fmarjanstoimchev\u002FRSMLC) -> Deep Network Architectures as Feature Extractors for Multi-Label Classification of Remote Sensing Images\n\n- [FireRisk](https:\u002F\u002Fgithub.com\u002FCharmonyShen\u002FFireRisk) -> A Remote Sensing Dataset for Fire Risk Assessment with Benchmarks Using Supervised and Self-supervised Learning\n\n- [flood_susceptibility_mapping](https:\u002F\u002Fgithub.com\u002Fomarseleem92\u002Fflood_susceptibility_mapping) -> Towards urban flood susceptibility mapping using data-driven models in Berlin, Germany\n\n- [Building-detection-and-roof-type-recognition](https:\u002F\u002Fgithub.com\u002Floosgagnet\u002FBuilding-detection-and-roof-type-recognition) -> A CNN-Based Approach for Automatic Building Detection and Recognition of Roof Types Using a Single Aerial Image\n\n- [SNN4Space](https:\u002F\u002Fgithub.com\u002FAndrzejKucik\u002FSNN4Space) -> project which investigates the feasibility of deploying spiking neural networks (SNN) in land cover and land use classification tasks\n\n- [vessel-classification](https:\u002F\u002Fgithub.com\u002FGlobalFishingWatch\u002Fvessel-classification) -> classify vessels and identify fishing behavior based on AIS data\n\n- [RSMamba](https:\u002F\u002Fgithub.com\u002FKyanChen\u002FRSMamba) -> Remote Sensing Image Classification with State Space Model\n\n- [BirdSAT](https:\u002F\u002Fgithub.com\u002Fmvrl\u002FBirdSAT) -> Cross-View Contrastive Masked Autoencoders for Bird Species Classification and Mapping\n\n- [EGNNA_WND](https:\u002F\u002Fgithub.com\u002Fstevinc\u002FEGNNA_WND) -> Estimating the presence of the West Nile Disease employing Graph Neural network\n\n- [cyfi](https:\u002F\u002Fgithub.com\u002Fdrivendataorg\u002Fcyfi) -> Estimate cyanobacteria density based on Sentinel-2 satellite imagery\n\n- [3DGAN-ViT](https:\u002F\u002Fgithub.com\u002Faj1365\u002F3DGAN-ViT) -> A deep learning framework based on generative adversarial networks and vision transformer for complex wetland classification\n\n- [automatic_solar_pv_detection](https:\u002F\u002Fgithub.com\u002FKennSmithDS\u002Fautomatic_solar_pv_detection) -> Automatic Solar PV Panel Image Classification with Deep Neural Network Transfer Learning\n\n- [U-netR](https:\u002F\u002Fgithub.com\u002FJonathanVSV\u002FU-netR) -> Land Use Land Cover Classification with U-Net: Advantages of Combining Sentinel-1 and Sentinel-2 Imagery [paper](https:\u002F\u002Fdoi.org\u002F10.3390\u002Frs13183600)\n\n- [nshaud\u002FDeepNetsForEO](https:\u002F\u002Fgithub.com\u002Fnshaud\u002FDeepNetsForEO) -> Deep networks for Earth Observation with PyTorch implementations of state-of-the-art architectures for remote sensing image classification\n\n- [sentinel-landslide-cls](https:\u002F\u002Fgithub.com\u002FIoannisNasios\u002Fsentinel-landslide-cls) -> Classification for Landslide Detection, using Sentinel-1 and Sentinel-2 data.\n\n#\n## Segmentation\n\n\u003Cp align=\"center\">\n  \u003Cimg src=\"images\u002Fsegmentation.png\" width=\"500\">\n  \u003Cbr>\n  \u003Cb>(left) a satellite image and (right) the semantic classes in the image.\u003C\u002Fb>\n\u003C\u002Fp>\n\nImage segmentation is a crucial step in image analysis and computer vision, with the goal of dividing an image into semantically meaningful segments or regions. The process of image segmentation assigns a class label to each pixel in an image, effectively transforming an image from a 2D grid of pixels into a 2D grid of pixels with assigned class labels. One common application of image segmentation is road or building segmentation, where the goal is to identify and separate roads and buildings from other features within an image. To accomplish this task, single class models are often trained to differentiate between roads and background, or buildings and background. These models are designed to recognize specific features, such as color, texture, and shape, that are characteristic of roads or buildings, and use this information to assign class labels to the pixels in an image. Another common application of image segmentation is land use or crop type classification, where the goal is to identify and map different land cover types within an image. In this case, multi-class models are typically used to recognize and differentiate between multiple classes within an image, such as forests, urban areas, and agricultural land. These models are capable of recognizing complex relationships between different land cover types, allowing for a more comprehensive understanding of the image content. Read [A brief introduction to satellite image segmentation with neural networks](https:\u002F\u002Fmedium.com\u002F@robmarkcole\u002Fa-brief-introduction-to-satellite-image-segmentation-with-neural-networks-33ea732d5bce). **Note** that many articles which refer to 'hyperspectral land classification' are often actually describing semantic segmentation.\n\n### Segmentation - Land use & land cover\n\n- [Automatic Detection of Landfill Using Deep Learning](https:\u002F\u002Fgithub.com\u002FAnupamaRajkumar\u002FLandfillDetection_SemanticSegmentation)\n\n- [CDL-Segmentation](https:\u002F\u002Fgithub.com\u002Fasimniazi63\u002FCDL-Segmentation) -> Deep Learning Based Land Cover and Crop Type Classification: A Comparative Study. Compares UNet, SegNet & DeepLabv3+\n\n- [LoveDA](https:\u002F\u002Fgithub.com\u002FJunjue-Wang\u002FLoveDA) -> A Remote Sensing Land-Cover Dataset for Domain Adaptive Semantic Segmentation\n\n- [DeepGlobe Land Cover Classification Challenge solution](https:\u002F\u002Fgithub.com\u002FGeneralLi95\u002Fdeepglobe_land_cover_classification_with_deeplabv3plus)\n\n- [CNN_Enhanced_GCN](https:\u002F\u002Fgithub.com\u002Fqichaoliu\u002FCNN_Enhanced_GCN) -> CNN-Enhanced Graph Convolutional Network With Pixel- and Superpixel-Level Feature Fusion for Hyperspectral Image Classification\n\n- [LULCMapping-WV3images-CORINE-DLMethods](https:\u002F\u002Fgithub.com\u002Fesertel\u002FLULCMapping-WV3images-CORINE-DLMethods) -> Land Use and Land Cover Mapping Using Deep Learning Based Segmentation Approaches and VHR Worldview-3 Images\n\n- [MCANet](https:\u002F\u002Fgithub.com\u002Fyisun98\u002FSOLC) -> A joint semantic segmentation framework of optical and SAR images for land use classification. Uses [WHU-OPT-SAR-dataset](https:\u002F\u002Fgithub.com\u002FAmberHen\u002FWHU-OPT-SAR-dataset)\n\n- [land-cover](https:\u002F\u002Fgithub.com\u002Flucashu1\u002Fland-cover) -> Model Generalization in Deep Learning Applications for Land Cover Mapping\n\n- [generalizablersc](https:\u002F\u002Fgithub.com\u002Fdgominski\u002Fgeneralizablersc) -> Cross-dataset Learning for Generalizable Land Use Scene Classification\n\n- [SSLTransformerRS](https:\u002F\u002Fgithub.com\u002FHSG-AIML\u002FSSLTransformerRS) -> Self-supervised Vision Transformers for Land-cover Segmentation and\n  Classification\n\n- [LULCMapping-WV3images-CORINE-DLMethods](https:\u002F\u002Fgithub.com\u002Fburakekim\u002FLULCMapping-WV3images-CORINE-DLMethods) -> Land Use and Land Cover Mapping Using Deep Learning Based Segmentation Approaches and VHR Worldview-3 Images\n\n- [DCSA-Net](https:\u002F\u002Fgithub.com\u002FJulia90\u002FDCSA-Net) -> Dynamic Convolution Self-Attention Network for Land-Cover Classification in VHR Remote-Sensing Images\n\n- [CHeGCN-CNN_enhanced_Heterogeneous_Graph](https:\u002F\u002Fgithub.com\u002FLiuzhizhiooo\u002FCHeGCN-CNN_enhanced_Heterogeneous_Graph) -> CNN-Enhanced Heterogeneous Graph Convolutional Network: Inferring Land Use from Land Cover with a Case Study of Park Segmentation\n\n- [TCSVT_2022_DGSSC](https:\u002F\u002Fgithub.com\u002FB-Xi\u002FTCSVT_2022_DGSSC) -> DGSSC: A Deep Generative Spectral-Spatial Classifier for Imbalanced Hyperspectral Imagery\n\n- [DeepForest-Wetland-Paper](https:\u002F\u002Fgithub.com\u002Faj1365\u002FDeepForest-Wetland-Paper) -> Deep Forest classifier for wetland mapping using the combination of Sentinel-1 and Sentinel-2 data, GIScience & Remote Sensing\n\n- [Wetland_UNet](https:\u002F\u002Fgithub.com\u002Fconservation-innovation-center\u002FWetland_UNet) -> UNet models that can delineate wetlands using remote sensing data input including bands from Sentinel-2 LiDAR and geomorphons. By the Conservation Innovation Center of Chesapeake Conservancy and Defenders of Wildlife\n\n- [DPA](https:\u002F\u002Fgithub.com\u002Fx-ytong\u002FDPA) -> DPA is an unsupervised domain adaptation (UDA) method applied to different satellite images for large-scale land cover mapping.\n\n- [dynamicworld](https:\u002F\u002Fgithub.com\u002Fgoogle\u002Fdynamicworld) -> Dynamic World, global 10 m land use land cover mapping from Google. [dynamic_world_pytorch](https:\u002F\u002Fgithub.com\u002Fcalebrob6\u002Fdynamic_world_pytorch) is a pytorch implementation.\n\n- [spada](https:\u002F\u002Fgithub.com\u002Flinks-ads\u002Fspada) -> Land Cover Segmentation with Sparse Annotations from Sentinel-2 Imagery\n\n- [M3SPADA](https:\u002F\u002Fgithub.com\u002Fecapliez\u002FM3SPADA) ->  Multi-Sensor Temporal Unsupervised Domain Adaptation for Land Cover Mapping with spatial pseudo labelling and adversarial learning\n\n- [GLNet](https:\u002F\u002Fgithub.com\u002FVITA-Group\u002FGLNet) -> Collaborative Global-Local Networks for Memory-Efficient Segmentation of Ultra-High Resolution Images\n\n- [LoveNAS](https:\u002F\u002Fgithub.com\u002FJunjue-Wang\u002FLoveNAS) -> LoveNAS: Towards Multi-Scene Land-Cover Mapping via Hierarchical Searching Adaptive Network\n\n- [FLAIR-2 challenge](https:\u002F\u002Fgithub.com\u002FIGNF\u002FFLAIR-2) -> Semantic segmentation and domain adaptation challenge proposed by the French National Institute of Geographical and Forest Information (IGN)\n\n- [flair-2 8th place solution](https:\u002F\u002Fgithub.com\u002Fassociation-rosia\u002Fflair-2)\n\n- [igarss-spada](https:\u002F\u002Fgithub.com\u002Flinks-ads\u002Fspada) -> Dataset and code for the paper Land Cover Segmentation with Sparse Annotations from Sentinel-2 Imagery [IGARSS 2023](https:\u002F\u002Farxiv.org\u002Fabs\u002F2306.16252)\n\n- [cnn-land-cover-eco](https:\u002F\u002Fgithub.com\u002FDGalexander\u002Fcnn-land-cover-eco) -> Multi-stage semantic segmentation of land cover in the Peak District using high-resolution RGB aerial imagery\n\n### Segmentation - Vegetation, deforestation, crops & field boundaries\n\nNote that deforestation detection may be treated as a segmentation task or a change detection task\n\n- [DetecTree](https:\u002F\u002Fgithub.com\u002Fmartibosch\u002Fdetectree) -> Tree detection from aerial imagery in Python, a LightGBM classifier of tree\u002Fnon-tree pixels from aerial imagery\n\n- [kenya-crop-mask](https:\u002F\u002Fgithub.com\u002Fnasaharvest\u002Fkenya-crop-mask) -> Annual and in-season crop mapping in Kenya - LSTM classifier to classify pixels as containing crop or not, and a multi-spectral forecaster that provides a 12 month time series given a partial input. Dataset downloaded from GEE and pytorch lightning used for training\n\n- [Tree species classification from from airborne LiDAR and hyperspectral data using 3D convolutional neural networks](https:\u002F\u002Fgithub.com\u002Fjaeeolma\u002Ftree-detection-evo)\n\n- [Find sports fields using Mask R-CNN and overlay on open-street-map](https:\u002F\u002Fgithub.com\u002Fjremillard\u002Fimages-to-osm)\n\n- [An LSTM to generate a crop mask for Togo](https:\u002F\u002Fgithub.com\u002Fnasaharvest\u002Ftogo-crop-mask)\n\n- [DeepSatModels](https:\u002F\u002Fgithub.com\u002Fmichaeltrs\u002FDeepSatModels) -> Context-self contrastive pretraining for crop type semantic segmentation\n\n- [DeepTreeAttention](https:\u002F\u002Fgithub.com\u002Fweecology\u002FDeepTreeAttention) -> Implementation of Hang et al. 2020 \"Hyperspectral Image Classification with Attention Aided CNNs\" for tree species prediction\n\n- [Crop-Classification](https:\u002F\u002Fgithub.com\u002Fbhavesh907\u002FCrop-Classification) -> crop classification using multi temporal satellite images\n\n- [crop-mask](https:\u002F\u002Fgithub.com\u002Fnasaharvest\u002Fcrop-mask) -> End-to-end workflow for generating high resolution cropland maps, uses GEE & LSTM model\n\n- [DeepCropMapping](https:\u002F\u002Fgithub.com\u002FLab-IDEAS\u002FDeepCropMapping) -> A multi-temporal deep learning approach with improved spatial generalizability for dynamic corn and soybean mapping, uses LSTM\n\n- [ResUnet-a](https:\u002F\u002Fgithub.com\u002FAkhilesh64\u002FResUnet-a) -> a deep learning framework for semantic segmentation of remotely sensed data\n\n- [DSD_paper_2020](https:\u002F\u002Fgithub.com\u002FJacobJeppesen\u002FDSD_paper_2020) -> Crop Type Classification based on Machine Learning with Multitemporal Sentinel-1 Data\n\n- [MR-DNN](https:\u002F\u002Fgithub.com\u002Fyasir2afaq\u002FMulti-resolution-deep-neural-network) -> extract rice field from Landsat 8 satellite imagery\n\n- [deep_learning_forest_monitoring](https:\u002F\u002Fgithub.com\u002Fwaldeland\u002Fdeep_learning_forest_monitoring) -> Forest mapping and monitoring of the African continent using Sentinel-2 data and deep learning\n\n- [global-cropland-mapping](https:\u002F\u002Fgithub.com\u002FCharly-tian\u002Fglobal-cropland-mapping) -> global multi-temporal cropland mapping\n\n- [Landuse_DL](https:\u002F\u002Fgithub.com\u002Fyghlc\u002FLanduse_DL) -> delineate landforms due to the thawing of ice-rich permafrost\n\n- [canopy](https:\u002F\u002Fgithub.com\u002Fjonathanventura\u002Fcanopy) -> A Convolutional Neural Network Classifier Identifies Tree Species in Mixed-Conifer Forest from Hyperspectral Imagery\n\n- [forest_change_detection](https:\u002F\u002Fgithub.com\u002FQuantuMobileSoftware\u002Fforest_change_detection) -> forest change segmentation with time-dependent models, including Siamese, UNet-LSTM, UNet-diff, UNet3D models\n\n- [cultionet](https:\u002F\u002Fgithub.com\u002Fjgrss\u002Fcultionet) -> segmentation of cultivated land, built on PyTorch Geometric and PyTorch Lightning\n\n- [sentinel-tree-cover](https:\u002F\u002Fgithub.com\u002Fwri\u002Fsentinel-tree-cover) -> A global method to identify trees outside of closed-canopy forests with medium-resolution satellite imagery\n\n- [crop-type-detection-ICLR-2020](https:\u002F\u002Fgithub.com\u002FRadiantMLHub\u002Fcrop-type-detection-ICLR-2020) -> Winning Solutions from Crop Type Detection Competition at CV4A workshop, ICLR 2020\n\n- [S4A-Models](https:\u002F\u002Fgithub.com\u002FOrion-AI-Lab\u002FS4A-Models) -> Various experiments on the Sen4AgriNet dataset\n\n- [attention-mechanism-unet](https:\u002F\u002Fgithub.com\u002Fdavej23\u002Fattention-mechanism-unet) -> An attention-based U-Net for detecting deforestation within satellite sensor imagery\n\n- [SummerCrop_Deeplearning](https:\u002F\u002Fgithub.com\u002FAgriRS\u002FSummerCrop_Deeplearning) -> A Transferable Learning Classification Model and Carbon Sequestration Estimation of Crops in Farmland Ecosystem\n\n- [DeepForest](https:\u002F\u002Fdeepforest.readthedocs.io\u002Fen\u002Flatest\u002Findex.html) is a python package for training and predicting individual tree crowns from airborne RGB imagery\n\n- [Official repository for the \"Identifying trees on satellite images\" challenge from Omdena](https:\u002F\u002Fgithub.com\u002Fcienciaydatos\u002Fai-challenge-trees)\n\n- [PTDM](https:\u002F\u002Fgithub.com\u002Fhr8yhtzb\u002FPTDM) -> Pomelo Tree Detection Method Based on Attention Mechanism and Cross-Layer Feature Fusion\n\n- [urban-tree-detection](https:\u002F\u002Fgithub.com\u002Fjonathanventura\u002Furban-tree-detection) -> Individual Tree Detection in Large-Scale Urban Environments using High-Resolution Multispectral Imagery. With [dataset](https:\u002F\u002Fgithub.com\u002Fjonathanventura\u002Furban-tree-detection-data)\n\n- [BioMassters_baseline](https:\u002F\u002Fgithub.com\u002Ffnands\u002FBioMassters_baseline) -> a basic pytorch lightning baseline using a UNet for getting started with the [BioMassters challenge](https:\u002F\u002Fwww.drivendata.org\u002Fcompetitions\u002F99\u002Fbiomass-estimation\u002F) (biomass estimation)\n\n- [Biomassters winners](https:\u002F\u002Fgithub.com\u002Fdrivendataorg\u002Fthe-biomassters) -> top 3 solutions\n\n- [kbrodt biomassters solution](https:\u002F\u002Fgithub.com\u002Fkbrodt\u002Fbiomassters) -> 1st place solution\n\n- [biomass-estimation](https:\u002F\u002Fgithub.com\u002Fazavea\u002Fbiomass-estimation) -> from Azavea, applied to Sentinel 1 & 2\n\n- [3DUNetGSFormer](https:\u002F\u002Fgithub.com\u002Faj1365\u002F3DUNetGSFormer) -> A deep learning pipeline for complex wetland mapping using generative adversarial networks and Swin transformer\n\n- [SEANet_torch](https:\u002F\u002Fgithub.com\u002Flong123524\u002FSEANet_torch) -> Using a semantic edge-aware multi-task neural network to delineate agricultural parcels from remote sensing images\n\n- [arborizer](https:\u002F\u002Fgithub.com\u002FRaffiBienz\u002Farborizer) -> Tree crowns segmentation and classification\n\n- [ReUse](https:\u002F\u002Fgithub.com\u002Fpriamus-lab\u002FReUse) -> REgressive Unet for Carbon Storage and Above-Ground Biomass Estimation\n\n- [unet-sentinel](https:\u002F\u002Fgithub.com\u002Feliasqueirogavieira\u002Funet-sentinel) -> UNet to handle Sentinel-1 SAR images to identify deforestation\n\n- [MaskedSST](https:\u002F\u002Fgithub.com\u002FHSG-AIML\u002FMaskedSST) -> Masked Vision Transformers for Hyperspectral Image Classification\n\n- [UNet-defmapping](https:\u002F\u002Fgithub.com\u002Fbragagnololu\u002FUNet-defmapping) -> master's thesis using UNet to map deforestation using Sentinel-2 Level 2A images, applied to Amazon and Atlantic Rainforest dataset\n\n- [cvpr-multiearth-deforestation-segmentation](https:\u002F\u002Fgithub.com\u002Fh2oai\u002Fcvpr-multiearth-deforestation-segmentation) -> multimodal Unet entry to the CVPR Multiearth 2023 deforestation challenge\n\n- [TransUNetplus2](https:\u002F\u002Fgithub.com\u002Faj1365\u002FTransUNetplus2) -> TransU-Net++: Rethinking attention gated TransU-Net for deforestation mapping. Uses the Amazon and Atlantic forest dataset\n\n- [A high-resolution canopy height model of the Earth](https:\u002F\u002Fgithub.com\u002Flangnico\u002Fglobal-canopy-height-model#a-high-resolution-canopy-height-model-of-the-earth) -> A high-resolution canopy height model of the Earth\n\n- [Radiant Earth Spot the Crop Challenge](https:\u002F\u002Fgithub.com\u002Fradiantearth\u002Fspot-the-crop-challenge) -> Winning models from the Radiant Earth Spot the Crop Challenge, uses a time-series of Sentinel-2 multispectral data to classify crops in the Western Cape of South Africa. [Another solution](https:\u002F\u002Fgithub.com\u002FDariusTheGeek\u002FRadiant-Earth-Spot-the-Crop-XL-Challenge)\n\n- [transfer-field-delineation](https:\u002F\u002Fgithub.com\u002Fkerner-lab\u002Ftransfer-field-delineation) -> Multi-Region Transfer Learning for Segmentation of Crop Field Boundaries in Satellite Images with Limited Labels\n\n- [crop-field-segmentation-ukan](https:\u002F\u002Fgithub.com\u002FDarthReca\u002Fcrop-field-segmentation-ukan) -> KANs and Sentinel for Effective and Explainable Crop Field Segmentation\n\n- [mowing-detection](https:\u002F\u002Fgithub.com\u002Flucas-batier\u002Fmowing-detection) -> Automatic detection of mowing and grazing from Sentinel images\n\n- [PTAViT3D and PTAViT3DCA](https:\u002F\u002Fgithub.com\u002Ffeevos\u002Ftfcl) -> Tackling fluffy clouds: field boundaries detection using time series of S2 and\u002For S1 imagery\n\n- [ai4boundaries](https:\u002F\u002Fgithub.com\u002Fwaldnerf\u002Fai4boundaries) -> a Python package that facilitates download of the AI4boundaries data set\n\n- [Nasa_harvest_field_boundary_competition](https:\u002F\u002Fgithub.com\u002Fradiantearth\u002FNasa_harvest_field_boundary_competition) -> Nasa Harvest Rwanda Field Boundary Detection Challenge Tutorial\n\n- [UTB_codes](https:\u002F\u002Fgithub.com\u002Fzhu-xlab\u002FUTB_codes) -> The Urban Tree Canopy Cover in Brazil [article](https:\u002F\u002Fnkszjx.github.io\u002Fprojects\u002FUTB.html)\n\n- [nasa_harvest_boundary_detection_challenge](https:\u002F\u002Fgithub.com\u002Fgeoaigroup\u002Fnasa_harvest_boundary_detection_challenge) -> the 4th place solution for NASA Harvest Field Boundary Detection Challenge on Zindi.\n\n- [rainforest-segmentation](https:\u002F\u002Fgithub.com\u002Fjcblsn\u002Frainforest-segmentation) -> Identifying and tracking deforestation in the Amazon Rainforest using state-of-the-art deep learning models and multispectral satellite imagery.\n\n- [Delineate Anything: Resolution-Agnostic Field Boundary Delineation on Satellite Imagery](https:\u002F\u002Fgithub.com\u002FLavreniuk\u002FDelineate-Anything)\n\n- [Semantic_segmentation_for_LCLUC](https:\u002F\u002Fgithub.com\u002Fwaterdmd\u002FSemantic_segmentation_for_LCLUC) -> Semantic Segmentation for Simultaneous Crop and Land Cover Land Use Classification Using Multi-Temporal Landsat Imagery\n\n- [boundary-sam](https:\u002F\u002Fgithub.com\u002Fawadbahaa\u002Fboundary-sam) -> parcel boundary delineation using SAM, image embeddings and detail enhancement filters\n\n- [TOFMapper](https:\u002F\u002Fgithub.com\u002FMoerizzy\u002FTOFMapper) -> a semantic segmentation tool for mapping and classifying Trees outside Forest in high resolution aerial images\n\n- [Mask-PSTIN](https:\u002F\u002Fgithub.com\u002FBruceKai\u002FMask-PSTIN) -> Improving crop type mapping by integrating LSTM with temporal random masking and pixel-set spatial information\n\n- [paddy_identification](https:\u002F\u002Fgithub.com\u002Fkayathri4\u002Fpaddy_identification) -> Paddy Field Instance Segmentation using Multi-Temporal SAR Time Series\n\n- [CropSight](https:\u002F\u002Fgithub.com\u002Frssiuiuc\u002FCropSight) -> towards a large-scale operational framework for object-based crop type ground truth retrieval using street view and PlanetScope satellite imagery\n\n- [ftw-prue](https:\u002F\u002Fgithub.com\u002Ffieldsoftheworld\u002Fftw-prue) -> PRUE: A Practical Recipe for Field Boundary Segmentation at Scale.\n\n- [agribound](https:\u002F\u002Fgithub.com\u002Fmontimaj\u002Fagribound) -> An AI-powered field boundary delineation toolkit combining satellite foundation models, embeddings, and global training data for accurate agricultural parcel\u002Ffield boundary mapping.\n\n- [s2-forest-browning-monitoring](https:\u002F\u002Fgithub.com\u002FSamanthaBiegel\u002Fs2-forest-browning-monitoring) -> Monitoring forest browning using Sentinel-2 imagery.\n\n- [LacunaLabels](https:\u002F\u002Fgithub.com\u002Fagroimpacts\u002Flacunalabels) -> A region-wide, multi-year set of crop field boundary labels for Africa.\n\n- [Pseudo-fields](https:\u002F\u002Fgithub.com\u002Fphilipperufin\u002Fpseudo-fields\u002F) -> Generating pseudo labels for satellite-based crop field delineatio.\n\n### Segmentation - Water, coastlines, rivers & floods\n\n- [sat-water](https:\u002F\u002Fgithub.com\u002Fbusayojee\u002Fsat-water) -> Semantic segmentation of water bodies in satellite imagery, producing pixel-wise water masks from remote sensing images using a U-Net–style deep learning pipeline (data preparation, training, inference, and evaluation).\n\n- [Houston_flooding](https:\u002F\u002Fgithub.com\u002FLichtphyz\u002FHouston_flooding) -> labeling each pixel as either flooded or not using data from Hurricane Harvey. Dataset consisted of pre and post flood images, and a ground truth floodwater mask was created using unsupervised clustering (with DBScan) of image pixels with human cluster verification\u002Fadjustment\n\n- [ml4floods](https:\u002F\u002Fgithub.com\u002Fspaceml-org\u002Fml4floods) -> An ecosystem of data, models and code pipelines to tackle flooding with ML\n\n- [1st place solution for STAC Overflow: Map Floodwater from Radar Imagery hosted by Microsoft AI for Earth](https:\u002F\u002Fgithub.com\u002Fsweetlhare\u002FSTAC-Overflow) -> combines Unet with Catboostclassifier, taking their maxima, not the average\n\n- [hydra-floods](https:\u002F\u002Fgithub.com\u002FServir-Mekong\u002Fhydra-floods) -> an open source Python application for downloading, processing, and delivering surface water maps derived from remote sensing data\n\n- [CoastSat](https:\u002F\u002Fgithub.com\u002Fkvos\u002FCoastSat) -> tool for mapping coastlines which has an extension [CoastSeg](https:\u002F\u002Fgithub.com\u002Fdbuscombe-usgs\u002FCoastSeg) using segmentation models\n\n- [deepwatermap](https:\u002F\u002Fgithub.com\u002Fisikdogan\u002Fdeepwatermap) -> a deep model that segments water on multispectral images\n\n- [rivamap](https:\u002F\u002Fgithub.com\u002Fisikdogan\u002Frivamap) -> an automated river analysis and mapping engine\n\n- [deep-water](https:\u002F\u002Fgithub.com\u002Fmaxbeber\u002Fdeep-water) -> track changes in water level\n\n- [WatNet](https:\u002F\u002Fgithub.com\u002Fxinluo2018\u002FWatNet) -> A deep ConvNet for surface water mapping based on Sentinel-2 image, uses the [Earth Surface Water Dataset](https:\u002F\u002Fzenodo.org\u002Frecord\u002F5205674#.YoMjyZPMK3I)\n\n- [A-U-Net-for-Flood-Extent-Mapping](https:\u002F\u002Fgithub.com\u002Fjorgemspereira\u002FA-U-Net-for-Flood-Extent-Mapping)\n\n- [floatingobjects](https:\u002F\u002Fgithub.com\u002FESA-PhiLab\u002Ffloatingobjects) -> TOWARDS DETECTING FLOATING OBJECTS ON A GLOBAL SCALE WITHLEARNED SPATIAL FEATURES USING SENTINEL 2. Uses U-Net & pytorch\n\n- [SpaceNet8](https:\u002F\u002Fgithub.com\u002FSpaceNetChallenge\u002FSpaceNet8) -> baseline Unet solution to detect flooded roads and buildings\n\n- [dlsim](https:\u002F\u002Fgithub.com\u002Fnyokoya\u002Fdlsim) -> Breaking the Limits of Remote Sensing by Simulation and Deep Learning for Flood and Debris Flow Mapping\n\n- [Water-HRNet](https:\u002F\u002Fgithub.com\u002Ffaye0078\u002FWater-Extraction) -> HRNet trained on Sentinel 2\n\n- [semantic segmentation model to identify newly developed or flooded land](https:\u002F\u002Fgithub.com\u002FAzure\u002Fpixel_level_land_classification) using NAIP imagery provided by the Chesapeake Conservancy, training on MS Azure\n\n- [BandNet](https:\u002F\u002Fgithub.com\u002FIamShubhamGupto\u002FBandNet) -> Analysis and application of multispectral data for water segmentation using machine learning. Uses Sentinel-2 data\n\n- [mmflood](https:\u002F\u002Fgithub.com\u002Fedornd\u002Fmmflood) -> MMFlood: A Multimodal Dataset for Flood Delineation From Satellite Imagery (Sentinel 1 SAR)\n\n- [Urban_flooding](https:\u002F\u002Fgithub.com\u002Fomarseleem92\u002FUrban_flooding) -> Towards transferable data-driven models to predict urban pluvial flood water depth in Berlin, Germany\n\n- [MECNet](https:\u002F\u002Fgithub.com\u002Fzhilyzhang\u002FMECNet) -> Rich CNN features for water-body segmentation from very high resolution aerial and satellite imagery\n\n- [SWRNET](https:\u002F\u002Fgithub.com\u002Ftrongan93\u002Fswrnet) -> A Deep Learning Approach for Small Surface Water Area Recognition Onboard Satellite\n\n- [elwha-segmentation](https:\u002F\u002Fgithub.com\u002FStefanTodoran\u002Felwha-segmentation) -> fine-tuning Meta's Segment Anything (SAM) for bird's eye view river pixel segmentation\n\n- [RiverSnap](https:\u002F\u002Fgithub.com\u002FArminMoghimi\u002FRiverSnap) -> code for paper: A Comparative Performance Analysis of Popular Deep Learning Models and Segment Anything Model (SAM) for River Water Segmentation in Close-Range Remote Sensing Imagery\n\n- [SAR-water-segmentation](https:\u002F\u002Fgithub.com\u002Fmyeungun\u002FSAR-water-segmentation) -> Deep Learning based Water Segmentation Using KOMPSAT-5 SAR Images\n\n- [TerraMind-Flood](https:\u002F\u002Fgithub.com\u002FR1-AK\u002Fterramind-flood) -> DEM-Enhanced Flood Detection with Physics-Aware Learning, applied to Sen1Flood11\n\n- [Prithvi-CAFE](https:\u002F\u002Fgithub.com\u002FSk-2103\u002FPrithvi-CAFE) -> Transformer-based global reasoning (Prithvi-EO-2.0) with CNN-based local spatial sensitivity, enabling high-resolution, reliable flood inundation mapping across multi-channel\u002Fsensor inputs, applied to Sen1Flood11\n\n- [SMAGNet](https:\u002F\u002Fgithub.com\u002FASUcicilab\u002FSMAGNet) -> A Spatially Masked Adaptive Gated Network for Multimodal Post-Flood Water Extent Mapping using SAR and Incomplete Multispectral Data. Uses c2smsfloods dataset\n\n- [IBM BlueSky Challenge - ZeroFlood](https:\u002F\u002Fgithub.com\u002Foroikono\u002FIBM-BlueSky-Challenge-ZeroFlood)\n\n- [OmniWaterMask-training](https:\u002F\u002Fgithub.com\u002FDPIRD-DMA\u002FOmniWaterMask-training) -> Training code for the deep learning model used in [OmniWaterMask](https:\u002F\u002Fgithub.com\u002FDPIRD-DMA\u002FOmniWaterMask) - a Python library for detecting water bodies in satellite and aerial imagery.\n\n### Segmentation - Fire, smoke & burn areas\n\n- [SatelliteVu-AWS-Disaster-Response-Hackathon](https:\u002F\u002Fgithub.com\u002FSatelliteVu\u002FSatelliteVu-AWS-Disaster-Response-Hackathon) -> fire spread prediction using classical ML & deep learning\n\n- [A Practical Method for High-Resolution Burned Area Monitoring Using Sentinel-2 and VIIRS](https:\u002F\u002Fgithub.com\u002Fmnpinto\u002FFireHR)\n\n- [IndustrialSmokePlumeDetection](https:\u002F\u002Fgithub.com\u002FHSG-AIML\u002FIndustrialSmokePlumeDetection) -> using Sentinel-2 & a modified ResNet-50\n\n- [burned-area-detection](https:\u002F\u002Fgithub.com\u002Fdymaxionlabs\u002Fburned-area-detection) -> uses Sentinel-2\n\n- [rescue](https:\u002F\u002Fgithub.com\u002Fdbdmg\u002Frescue) -> Attention to fires: multi-channel deep-learning models for wildfire severity prediction\n\n- [smoke_segmentation](https:\u002F\u002Fgithub.com\u002Fjeffwen\u002Fsmoke_segmentation) -> Segmenting smoke plumes and predicting density from GOES imagery\n\n- [wildfire-detection](https:\u002F\u002Fgithub.com\u002Famanbasu\u002Fwildfire-detection) -> Using Vision Transformers for enhanced wildfire detection in satellite images\n\n- [Burned_Area_Detection](https:\u002F\u002Fgithub.com\u002Fprhuppertz\u002FBurned_Area_Detection) -> Detecting Burned Areas with Sentinel-2 data\n\n- [burned-area-baseline](https:\u002F\u002Fgithub.com\u002Flccol\u002Fburned-area-baseline) -> baseline unet model accompanying the Satellite Burned Area Dataset (Sentinel 1 & 2)\n\n- [burned-area-seg](https:\u002F\u002Fgithub.com\u002Flinks-ads\u002Fburned-area-seg) -> Burned area segmentation from Sentinel-2 using multi-task learning\n\n- [chabud2023](https:\u002F\u002Fgithub.com\u002Fdevelopmentseed\u002Fchabud2023) -> Change detection for Burned area Delineation (ChaBuD) ECML\u002FPKDD 2023 challenge\n\n- [Post Wildfire Burnt-up Detection using Siamese-UNet](https:\u002F\u002Fgithub.com\u002Fkavyagupta\u002Fchabud) -> on Chadbud dataset\n\n- [vit-burned-detection](https:\u002F\u002Fgithub.com\u002FDarthReca\u002Fvit-burned-detection) -> Vision transformers in burned area delineation\n\n- [ai4good25-wildfire](https:\u002F\u002Fgithub.com\u002FVSainteuf\u002Fai4good25-wildfire) -> AI4GOOD Class Fall 2025 : Wildfire spread prediction project\n\n### Segmentation - Landslides\n\n- [landslide-sar-unet](https:\u002F\u002Fgithub.com\u002Fiprapas\u002Flandslide-sar-unet) -> Deep Learning for Rapid Landslide Detection using Synthetic Aperture Radar (SAR) Datacubes\n\n- [landslide-mapping-with-cnn](https:\u002F\u002Fgithub.com\u002Fnprksh\u002Flandslide-mapping-with-cnn) -> A new strategy to map landslides with a generalized convolutional neural network\n\n- [Landslide-mapping-on-SAR-data-by-Attention-U-Net](https:\u002F\u002Fgithub.com\u002Florenzonava96\u002FLandslide-mapping-on-SAR-data-by-Attention-U-Net) -> Rapid Mapping of landslide on SAR data by Attention U-net\n\n- [SAR-landslide-detection-pretraining](https:\u002F\u002Fgithub.com\u002FVMBoehm\u002FSAR-landslide-detection-pretraining) -> SAR-based landslide classification pretraining leads to better segmentation\n\n- [Landslide mapping from Sentinel-2 imagery through change detection](https:\u002F\u002Fgithub.com\u002Flinks-ads\u002Figarss-landslide-delineation)\n\n- [landslide4sense-solution](https:\u002F\u002Fgithub.com\u002Fiamtekson\u002Flandslide4sense-solution) -> solution of Tek Kshetri\n\n- [DiGATe-UNet-LandSlide-Segmentation](https:\u002F\u002Fgithub.com\u002Fmishaown\u002FDiGATe-UNet-LandSlide-Segmentation) -> Lightweight Dual-Stream Framework for Landslide Segmentation\n\n- [Erosion-detection](https:\u002F\u002Fgithub.com\u002FAndrii-Radyhin\u002FErosion-detection) -> using Sentinel-2 to detect erosion\n\n### Segmentation - Glaciers\n\n- [HED-UNet](https:\u002F\u002Fgithub.com\u002Fkhdlr\u002FHED-UNet) -> a model for simultaneous semantic segmentation and edge detection, examples provided are glacier fronts and building footprints using the Inria Aerial Image Labeling dataset\n\n- [glacier_mapping](https:\u002F\u002Fgithub.com\u002Fkrisrs1128\u002Fglacier_mapping) -> Mapping glaciers in the Hindu Kush Himalaya, Landsat 7 images, Shapefile labels of the glaciers, Unet with dropout\n\n- [GlacierSemanticSegmentation](https:\u002F\u002Fgithub.com\u002Fn9Mtq4\u002FGlacierSemanticSegmentation)\n\n- [Antarctic-fracture-detection](https:\u002F\u002Fgithub.com\u002Fchingyaolai\u002FAntarctic-fracture-detection) -> uses UNet with the MODIS Mosaic of Antarctica to detect surface fractures\n\n- [sentinel_lakeice](https:\u002F\u002Fgithub.com\u002Fprs-eth\u002Fsentinel_lakeice) -> Lake Ice Detection from Sentinel-1 SAR with Deep Learning\n\n- [MCD-Net](https:\u002F\u002Fgithub.com\u002FLyra-alpha\u002FMCD-Net) -> a lightweight deep learning framework for optical-only moraine segmentation\n\n### Segmentation - methane\n\n- [Methane-detection-from-hyperspectral-imagery](https:\u002F\u002Fgithub.com\u002Fsatish1901\u002FMethane-detection-from-hyperspectral-imagery) -> Deep Remote Sensing Methods for Methane Detection in Overhead Hyperspectral Imagery\n\n- [methane-emission-project](https:\u002F\u002Fgithub.com\u002Fstlbnmaria\u002Fmethane-emission-project) -> Classification CNNs was combined in an ensemble approach with traditional methods on tabular data\n\n- [CH4Net](https:\u002F\u002Fgithub.com\u002Fannavaughan\u002FCH4Net) -> A fast, simple model for detection of methane plumes using sentinel-2\n\n- [STARCOP: Semantic Segmentation of Methane Plumes with Hyperspectral Machine Learning models](https:\u002F\u002Fgithub.com\u002Fspaceml-org\u002FSTARCOP)\n\n- [Project-Eucalyptus](https:\u002F\u002Fgithub.com\u002FOrbio-Earth\u002FProject-Eucalyptus) -> pipelines for satellite-based methane detection. Includes trained segmentation models, a synthetic plume generator, and benchmarking tools for Sentinel-2, Landsat 8\u002F9, and EMIT.\n\n- [plume-hunter: Towards Methane Detection On Board Satellites ](https:\u002F\u002Fgithub.com\u002Fspaceml-org\u002Fplume-hunter)\n\n\n### Segmentation - Other environmental\n\n- [Detection of Open Landfills](https:\u002F\u002Fgithub.com\u002Fdymaxionlabs\u002Fbasurales) -> uses Sentinel-2 to detect large changes in the Normalized Burn Ratio (NBR)\n\n- [sea_ice_remote_sensing](https:\u002F\u002Fgithub.com\u002Fsum1lim\u002Fsea_ice_remote_sensing) -> Sea Ice Concentration classification\n\n- [EddyNet](https:\u002F\u002Fgithub.com\u002Fredouanelg\u002FEddyNet) -> A Deep Neural Network For Pixel-Wise Classification of Oceanic Eddies\n\n- [schisto-vegetation](https:\u002F\u002Fgithub.com\u002Fdeleo-lab\u002Fschisto-vegetation) -> Deep Learning Segmentation of Satellite Imagery Identifies Aquatic Vegetation Associated with Snail Intermediate Hosts of Schistosomiasis in Senegal, Africa\n\n- [Earthformer](https:\u002F\u002Fgithub.com\u002Famazon-science\u002Fearth-forecasting-transformer) -> Exploring space-time transformers for earth system forecasting\n\n- [weather4cast-2022](https:\u002F\u002Fgithub.com\u002Fiarai\u002Fweather4cast-2022) -> Unet-3D baseline model for Weather4cast Rain Movie Prediction competition\n\n- [WeatherFusionNet](https:\u002F\u002Fgithub.com\u002FDatalab-FIT-CTU\u002Fweather4cast-2022) -> Predicting Precipitation from Satellite Data. weather4cast-2022 1st place solution\n\n- [marinedebrisdetector](https:\u002F\u002Fgithub.com\u002FMarcCoru\u002Fmarinedebrisdetector) -> Large-scale Detection of Marine Debris in Coastal Areas with Sentinel-2\n\n- [kaggle-identify-contrails-4th](https:\u002F\u002Fgithub.com\u002Fselimsef\u002Fkaggle-identify-contrails-4th) -> 4th place Solution, Google Research - Identify Contrails to Reduce Global Warming\n\n- [MineSegSAT](https:\u002F\u002Fgithub.com\u002Fmacdonaldezra\u002FMineSegSAT) -> An automated system to evaluate mining disturbed area extents from Sentinel-2 imagery\n\n- [asos](https:\u002F\u002Fgitlab.jsc.fz-juelich.de\u002Fkiste\u002Fasos) -> Recognizing protected and anthropogenic patterns in landscapes using interpretable machine learning and satellite imagery\n\n- [SinkSAM](https:\u002F\u002Fgithub.com\u002Fosherr1996\u002FSinkSAM) -> Knowledge-Driven Self-Supervised Sinkhole Segmentation Using Topographic Priors and Segment Anything Model\n\n### Segmentation - Roads & sidewalks\nExtracting roads is challenging due to the occlusions caused by other objects and the complex traffic environment\n\n- [ChesapeakeRSC](https:\u002F\u002Fgithub.com\u002Fisaaccorley\u002FChesapeakeRSC) -> segmentation to extract roads from the background but are additionally evaluated by how they perform on the \"Tree Canopy Over Road\" class\n\n- [ML_EPFL_Project_2](https:\u002F\u002Fgithub.com\u002FLucasBrazCappelo\u002FML_EPFL_Project_2) -> U-Net in Pytorch to perform semantic segmentation of roads on satellite images\n\n- [Winning Solutions from SpaceNet Road Detection and Routing Challenge](https:\u002F\u002Fgithub.com\u002FSpaceNetChallenge\u002FRoadDetector)\n\n\n- [awesome-deep-map](https:\u002F\u002Fgithub.com\u002Fantran89\u002Fawesome-deep-map) -> A curated list of resources dedicated to deep learning \u002F computer vision algorithms for mapping. The mapping problems include road network inference, building footprint extraction, etc.\n\n- [RoadTracer: Automatic Extraction of Road Networks from Aerial Images](https:\u002F\u002Fgithub.com\u002Fmitroadmaps\u002Froadtracer) -> uses an iterative search process guided by a CNN-based decision function to derive the road network graph directly from the output of the CNN\n\n- [road_detection_mtl](https:\u002F\u002Fgithub.com\u002Fntelo007\u002Froad_detection_mtl) -> Road Detection using a multi-task Learning technique to improve the performance of the road detection task by incorporating prior knowledge constraints, uses the SpaceNet Roads Dataset\n\n- [road_connectivity](https:\u002F\u002Fgithub.com\u002Fanilbatra2185\u002Froad_connectivity) -> Improved Road Connectivity by Joint Learning of Orientation and Segmentation (CVPR2019)\n\n- [SPIN_RoadMapper](https:\u002F\u002Fgithub.com\u002Fwgcban\u002FSPIN_RoadMapper) -> Extracting Roads from Aerial Images via Spatial and Interaction Space Graph Reasoning for Autonomous Driving\n\n- [road_extraction_remote_sensing](https:\u002F\u002Fgithub.com\u002Fjiankang1991\u002Froad_extraction_remote_sensing) -> pytorch implementation, CVPR2018 DeepGlobe Road Extraction Challenge submission. See also [DeepGlobe-Road-Extraction-Challenge](https:\u002F\u002Fgithub.com\u002Fzlckanata\u002FDeepGlobe-Road-Extraction-Challenge)\n\n- [RoadDetections dataset by Microsoft](https:\u002F\u002Fgithub.com\u002Fmicrosoft\u002FRoadDetections)\n\n- [CoANet](https:\u002F\u002Fgithub.com\u002Fmj129\u002FCoANet) -> Connectivity Attention Network for Road Extraction From Satellite Imagery. The CoA module incorporates graphical information to ensure the connectivity of roads are better preserved\n\n- [Satellite Imagery Road Segmentation](https:\u002F\u002Fmedium.com\u002F@nithishmailme\u002Fsatellite-imagery-road-segmentation-ad2964dc3812) -> intro article on Medium using the kaggle [Massachusetts Roads Dataset](https:\u002F\u002Fwww.kaggle.com\u002Fdatasets\u002Fbalraj98\u002Fmassachusetts-roads-dataset)\n\n- [Label-Pixels](https:\u002F\u002Fgithub.com\u002Fvenkanna37\u002FLabel-Pixels) -> for semantic segmentation of roads and other features\n\n- [Satellite-image-road-extraction](https:\u002F\u002Fgithub.com\u002Famanhari-projects\u002FSatellite-image-road-extraction) -> Road Extraction by Deep Residual U-Net\n\n- [road_building_extraction](https:\u002F\u002Fgithub.com\u002Fjeffwen\u002Froad_building_extraction) -> Pytorch implementation of U-Net architecture for road and building extraction\n\n- [RCFSNet](https:\u002F\u002Fgithub.com\u002FCVer-Yang\u002FRCFSNet) -> Road Extraction From Satellite Imagery by Road Context and Full-Stage Feature\n\n- [SGCN](https:\u002F\u002Fgithub.com\u002Ftist0bsc\u002FSGCN) -> Split Depth-Wise Separable Graph-Convolution Network for Road Extraction in Complex Environments From High-Resolution Remote-Sensing Images\n\n- [ASPN](https:\u002F\u002Fgithub.com\u002Fpshams55\u002FASPN) -> Road Segmentation for Remote Sensing Images using Adversarial Spatial Pyramid Networks\n\n- [cresi](https:\u002F\u002Fgithub.com\u002Favanetten\u002Fcresi) -> Road network extraction from satellite imagery, with speed and travel time estimates\n\n- [D-LinkNet](https:\u002F\u002Fgithub.com\u002FNekoApocalypse\u002Froad-extraction-d-linknet) -> LinkNet with Pretrained Encoder and Dilated Convolution for High Resolution Satellite Imagery Road Extraction\n\n- [Sat2Graph](https:\u002F\u002Fgithub.com\u002Fsongtaohe\u002FSat2Graph) -> Road Graph Extraction through Graph-Tensor Encoding\n\n- [RoadTracer-M](https:\u002F\u002Fgithub.com\u002Fastro-ck\u002FRoadTracer-M) -> Road Network Extraction from Satellite Images Using CNN Based Segmentation and Tracing\n\n- [ScRoadExtractor](https:\u002F\u002Fgithub.com\u002Fweiyao1996\u002FScRoadExtractor) -> Scribble-based Weakly Supervised Deep Learning for Road Surface Extraction from Remote Sensing Images\n\n- [RoadDA](https:\u002F\u002Fgithub.com\u002FLANMNG\u002FRoadDA) -> Stagewise Unsupervised Domain Adaptation with Adversarial Self-Training for Road Segmentation of Remote Sensing Images\n\n- [DeepSegmentor](https:\u002F\u002Fgithub.com\u002Fyhlleo\u002FDeepSegmentor) -> A Pytorch implementation of DeepCrack and RoadNet projects\n\n- [Cascaded Residual Attention Enhanced Road Extraction from Remote Sensing Images](https:\u002F\u002Fgithub.com\u002Fliaochengcsu\u002FCascade_Residual_Attention_Enhanced_for_Refinement_Road_Extraction)\n\n- [NL-LinkNet](https:\u002F\u002Fgithub.com\u002FSIAnalytics\u002Fnia-road-baseline) -> Toward Lighter but More Accurate Road Extraction with Non-Local Operations\n\n- [IRSR-net](https:\u002F\u002Fgithub.com\u002Fyangzhen1252\u002FIRSR-net) -> Lightweight Remote Sensing Road Detection Network\n\n- [hironex](https:\u002F\u002Fgithub.com\u002Fjohannesuhl\u002Fhironex) -> A python tool for automatic, fully unsupervised extraction of historical road networks from historical maps\n\n- [Road_detection_model](https:\u002F\u002Fgithub.com\u002FJonasImazon\u002FRoad_detection_model) -> Mapping Roads in the Brazilian Amazon with Artificial Intelligence and Sentinel-2\n\n- [DTnet](https:\u002F\u002Fgithub.com\u002Fhuzican695\u002FDTnet) -> Road detection via a dual-task network based on cross-layer graph fusion modules\n\n- [Automatic-Road-Extraction-from-Historical-Maps-using-Deep-Learning-Techniques](https:\u002F\u002Fgithub.com\u002FUrbanOccupationsOETR\u002FAutomatic-Road-Extraction-from-Historical-Maps-using-Deep-Learning-Techniques) -> Automatic Road Extraction from Historical Maps using Deep Learning Techniques\n\n- [Istanbul_Dataset](https:\u002F\u002Fgithub.com\u002FTolgaBkm\u002FIstanbul_Dataset) -> segmentation on the Istanbul, Inria and Massachusetts datasets\n\n- [D-LinkNet](https:\u002F\u002Fgithub.com\u002FShenweiXie\u002FD-LinkNet) -> 1st place solution in DeepGlobe Road Extraction Challenge\n\n- [PaRK-Detect](https:\u002F\u002Fgithub.com\u002FShenweiXie\u002FPaRK-Detect) -> PaRK-Detect: Towards Efficient Multi-Task Satellite Imagery Road Extraction via Patch-Wise Keypoints Detection\n\n- [tile2net](https:\u002F\u002Fgithub.com\u002FVIDA-NYU\u002Ftile2net) -> Mapping the walk: A scalable computer vision approach for generating sidewalk network datasets from aerial imagery\n\n- [sam_road](https:\u002F\u002Fgithub.com\u002Fhtcr\u002Fsam_road) -> Segment Anything Model (SAM) for large-scale, vectorized road network extraction from aerial imagery.\n\n- [LRDNet](https:\u002F\u002Fgithub.com\u002Fdyl96\u002FLRDNet) -> A Lightweight Road Detection Algorithm Based on Multiscale Convolutional Attention Network and Coupled Decoder Head\n\n- [Fine–Grained Extraction of Road Networks via Joint Learning of Connectivity and Segmentation](https:\u002F\u002Fgithub.com\u002FYXu556\u002FRoadExtraction) -> uses SpaceNet 3 dataset\n\n- [Satellite-Image-Road-Segmentation](https:\u002F\u002Fgithub.com\u002Faavek\u002FSatellite-Image-Road-Segmentation) -> Graph Reasoned Multi-Scale Road Segmentation in Remote Sensing Imagery [paper](https:\u002F\u002Fieeexplore.ieee.org\u002Fdocument\u002F10281660)\n\n- [PathFinder](https:\u002F\u002Fgithub.com\u002FOraegbuayomide10\u002FPathFinder) -> A Foundation Model for Road Mapping in Support of United Nations Humanitarian Affairs\n\n### Segmentation - Buildings & rooftops\n\n- [Road and Building Semantic Segmentation in Satellite Imagery](https:\u002F\u002Fgithub.com\u002FPaulymorphous\u002FRoad-Segmentation) uses U-Net on the Massachusetts Roads Dataset & keras\n\n- [find unauthorized constructions using aerial photography](https:\u002F\u002Fmedium.com\u002Ftowards-artificial-intelligence\u002Ffind-unauthorized-constructions-using-aerial-photography-and-deep-learning-with-code-part-2-b56ca80c8c99) -> [Dataset creation](https:\u002F\u002Fpub.towardsai.net\u002Ffind-unauthorized-constructions-using-aerial-photography-and-deep-learning-with-code-part-1-6d3ca7ff6fa0)\n\n- [SRBuildSeg](https:\u002F\u002Fgithub.com\u002Fxian1234\u002FSRBuildSeg) -> Making low-resolution satellite images reborn: a deep learning approach for super-resolution building extraction\n\n\n- [automated-building-detection](https:\u002F\u002Fgithub.com\u002Frodekruis\u002Fautomated-building-detection) -> Input: very-high-resolution (\u003C= 0.5 m\u002Fpixel) RGB satellite images. Output: buildings in vector format (geojson), to be used in digital map products. Built on top of robosat and robosat.pink.\n\n- [JointNet-A-Common-Neural-Network-for-Road-and-Building-Extraction](https:\u002F\u002Fgithub.com\u002FThomasWangWeiHong\u002FJointNet-A-Common-Neural-Network-for-Road-and-Building-Extraction)\n\n- [Mapping Africa’s Buildings with Satellite Imagery: Google AI blog post](https:\u002F\u002Fai.googleblog.com\u002F2021\u002F07\u002Fmapping-africas-buildings-with.html). See the [open-buildings](https:\u002F\u002Fsites.research.google\u002Fopen-buildings\u002F) dataset\n\n- [nz_convnet](https:\u002F\u002Fgithub.com\u002Fweiji14\u002Fnz_convnet) -> A U-net based ConvNet for New Zealand imagery to classify building outlines\n\n- [polycnn](https:\u002F\u002Fgithub.com\u002FLydorn\u002Fpolycnn) -> End-to-End Learning of Polygons for Remote Sensing Image Classification\n\n- [spacenet_building_detection](https:\u002F\u002Fgithub.com\u002Fmotokimura\u002Fspacenet_building_detection) solution by [motokimura](https:\u002F\u002Fgithub.com\u002Fmotokimura) using Unet\n\n- [Semantic-segmentation repo by fuweifu-vtoo](https:\u002F\u002Fgithub.com\u002Ffuweifu-vtoo\u002FSemantic-segmentation) -> uses pytorch and the [Massachusetts Buildings & Roads Datasets](https:\u002F\u002Fwww.cs.toronto.edu\u002F~vmnih\u002Fdata\u002F)\n\n- [Extracting buildings and roads from AWS Open Data using Amazon SageMaker](https:\u002F\u002Faws.amazon.com\u002Fblogs\u002Fmachine-learning\u002Fextracting-buildings-and-roads-from-aws-open-data-using-amazon-sagemaker\u002F) -> With [repo](https:\u002F\u002Fgithub.com\u002Faws-samples\u002Faws-open-data-satellite-lidar-tutorial)\n\n- [TF-SegNet](https:\u002F\u002Fgithub.com\u002Fmathildor\u002FTF-SegNet) -> AirNet is a segmentation network based on SegNet, but with some modifications\n\n- [rgb-footprint-extract](https:\u002F\u002Fgithub.com\u002Faatifjiwani\u002Frgb-footprint-extract) -> a Semantic Segmentation Network for Urban-Scale Building Footprint Extraction Using RGB Satellite Imagery, DeepLavV3+ module with a Dilated ResNet C42 backbone\n\n- [SpaceNetExploration](https:\u002F\u002Fgithub.com\u002Fyangsiyu007\u002FSpaceNetExploration) -> A sample project demonstrating how to extract building footprints from satellite images using a semantic segmentation model. Data from the SpaceNet Challenge\n\n- [Rooftop-Instance-Segmentation](https:\u002F\u002Fgithub.com\u002FMasterSkepticista\u002FRooftop-Instance-Segmentation) -> VGG-16, Instance Segmentation, uses the Airs dataset\n\n- [solar-farms-mapping](https:\u002F\u002Fgithub.com\u002Fmicrosoft\u002Fsolar-farms-mapping) -> An Artificial Intelligence Dataset for Solar Energy Locations in India\n\n- [poultry-cafos](https:\u002F\u002Fgithub.com\u002Fmicrosoft\u002Fpoultry-cafos) -> This repo contains code for detecting poultry barns from high-resolution aerial imagery and an accompanying dataset of predicted barns over the United States\n\n- [ssai-cnn](https:\u002F\u002Fgithub.com\u002Fmitmul\u002Fssai-cnn) -> This is an implementation of Volodymyr Mnih's dissertation methods on his Massachusetts road & building dataset\n\n- [Remote-sensing-building-extraction-to-3D-model-using-Paddle-and-Grasshopper](https:\u002F\u002Fgithub.com\u002FYoussef-Harby\u002FRemote-sensing-building-extraction-to-3D-model-using-Paddle-and-Grasshopper)\n\n- [segmentation-enhanced-resunet](https:\u002F\u002Fgithub.com\u002Ftranleanh\u002Fsegmentation-enhanced-resunet) -> Urban building extraction in Daejeon region using Modified Residual U-Net (Modified ResUnet) and applying post-processing\n\n- [Mask RCNN for Spacenet Off Nadir Building Detection](https:\u002F\u002Fgithub.com\u002Fashnair1\u002FMask-RCNN-for-Off-Nadir-Building-Detection)\n\n- [GRSL_BFE_MA](https:\u002F\u002Fgithub.com\u002Fjiankang1991\u002FGRSL_BFE_MA) -> Deep Learning-based Building Footprint Extraction with Missing Annotations using a novel loss function\n\n- [FER-CNN](https:\u002F\u002Fgithub.com\u002Frunnergirl13\u002FFER-CNN) -> Detection, Classification and Boundary Regularization of Buildings in Satellite Imagery Using Faster Edge Region Convolutional Neural Networks\n\n- [Vector-Map-Generation-from-Aerial-Imagery-using-Deep-Learning-GeoSpatial-UNET](https:\u002F\u002Fgithub.com\u002FManishSahu53\u002FVector-Map-Generation-from-Aerial-Imagery-using-Deep-Learning-GeoSpatial-UNET) -> applied to geo-referenced images which are very large size > 10k x 10k pixels\n\n- [building-footprint-segmentation](https:\u002F\u002Fgithub.com\u002Ffuzailpalnak\u002Fbuilding-footprint-segmentation) -> pip installable library to train building footprint segmentation on satellite and aerial imagery, applied to Massachusetts Buildings Dataset and Inria Aerial Image Labeling Dataset\n\n- [FCNN-example](https:\u002F\u002Fgithub.com\u002Femredog\u002FFCNN-example) -> overfit to a given single image to detect houses\n\n- [SAT2LOD2](https:\u002F\u002Fgithub.com\u002Fgdaosu\u002Flod2buildingmodel) -> an open-source, python-based GUI-enabled software that takes the satellite images as inputs and returns LoD2 building models as outputs\n\n- [SatFootprint](https:\u002F\u002Fgithub.com\u002FPriyanK7n\u002FSatFootprint) -> building segmentation on the Spacenet 7 dataset\n\n- [Building-Detection](https:\u002F\u002Fgithub.com\u002FEL-BID\u002FBuilding-Detection) -> Raster Vision experiment to train a model to detect buildings from satellite imagery in three cities in Latin America\n\n- [Multi-building-tracker](https:\u002F\u002Fgithub.com\u002Fsebasmos\u002FMulti-building-tracker) -> Multi-target building tracker for satellite images using deep learning\n\n- [Boundary Enhancement Semantic Segmentation for Building Extraction](https:\u002F\u002Fgithub.com\u002Fhin1115\u002FBEmodule-Satellite-Building-Segmentation)\n\n- [Spacenet-Building-Detection](https:\u002F\u002Fgithub.com\u002FIdanC1s2\u002FSpacenet-Building-Detection)\n\n- [LGPNet-BCD](https:\u002F\u002Fgithub.com\u002FTongfeiLiu\u002FLGPNet-BCD) -> Building Change Detection for VHR Remote Sensing Images via Local-Global Pyramid Network and Cross-Task Transfer Learning Strategy\n\n- [MTL_homoscedastic_SRB](https:\u002F\u002Fgithub.com\u002Fburakekim\u002FMTL_homoscedastic_SRB) -> A Multi-Task Deep Learning Framework for Building Footprint Segmentation\n\n- [FDANet](https:\u002F\u002Fgithub.com\u002Fdaifeng2016\u002FFDANet) -> Full-Level Domain Adaptation for Building Extraction in Very-High-Resolution Optical Remote-Sensing Images\n\n- [CBRNet](https:\u002F\u002Fgithub.com\u002FHaonanGuo\u002FCBRNet) -> A Coarse-to-fine Boundary Refinement Network for Building Extraction from Remote Sensing Imagery\n\n- [ASLNet](https:\u002F\u002Fgithub.com\u002FggsDing\u002FASLNet) -> Adversarial Shape Learning for Building Extraction in VHR Remote Sensing Images\n\n- [BRRNet](https:\u002F\u002Fgithub.com\u002Fwangyi111\u002FBuilding-Extraction) -> A Fully Convolutional Neural Network for Automatic Building Extraction From High-Resolution Remote Sensing Images\n\n- [Multi-Scale-Filtering-Building-Index](https:\u002F\u002Fgithub.com\u002FThomasWangWeiHong\u002FMulti-Scale-Filtering-Building-Index) -> A Multi - Scale Filtering Building Index for Building Extraction in Very High - Resolution Satellite Imagery\n\n- [Models for Remote Sensing](https:\u002F\u002Fgithub.com\u002Fbohaohuang\u002Fmrs) -> long list of unets etc applied to building detection\n\n- [boundary_loss_for_remote_sensing](https:\u002F\u002Fgithub.com\u002Fyiskw713\u002Fboundary_loss_for_remote_sensing) -> Boundary Loss for Remote Sensing Imagery Semantic Segmentation\n\n- [Open Cities AI Challenge](https:\u002F\u002Fwww.drivendata.org\u002Fcompetitions\u002F60\u002Fbuilding-segmentation-disaster-resilience\u002F) -> Segmenting Buildings for Disaster Resilience. Winning solutions [on Github](https:\u002F\u002Fgithub.com\u002Fdrivendataorg\u002Fopen-cities-ai-challenge\u002F)\n\n- [MAPNet](https:\u002F\u002Fgithub.com\u002Flehaifeng\u002FMAPNet) -> Multi Attending Path Neural Network for Building Footprint Extraction from Remote Sensed Imagery\n\n- [dual-hrnet](https:\u002F\u002Fgithub.com\u002FSIAnalytics\u002Fdual-hrnet) -> localizing buildings and classifying their damage level\n\n- [ESFNet](https:\u002F\u002Fgithub.com\u002Fmrluin\u002FESFNet-Pytorch) -> Efficient Network for Building Extraction from High-Resolution Aerial Images\n\n- [CVCMFFNet](https:\u002F\u002Fgithub.com\u002FJiankun-chen\u002FCVCMFFNet-master) -> Complex-Valued Convolutional and Multifeature Fusion Network for Building Semantic Segmentation of InSAR Images\n\n- [STEB-UNet](https:\u002F\u002Fgithub.com\u002FBrightGuo048\u002FSTEB-UNet) -> A Swin Transformer-Based Encoding Booster Integrated in U-Shaped Network for Building Extraction\n\n- [dfc2020_baseline](https:\u002F\u002Fgithub.com\u002Flukasliebel\u002Fdfc2020_baseline) -> Baseline solution for the IEEE GRSS Data Fusion Contest 2020. Predict land cover labels from Sentinel-1 and Sentinel-2 imagery\n\n- [Fusing multiple segmentation models based on different datasets into a single edge-deployable model](https:\u002F\u002Fgithub.com\u002Fmarkusmeingast\u002FSatellite-Classifier) -> roof, car & road segmentation\n\n- [ground-truth-gan-segmentation](https:\u002F\u002Fgithub.com\u002Fzakariamejdoul\u002Fground-truth-gan-segmentation) -> use Pix2Pix to segment the footprint of a building. The dataset used is AIRS\n\n- [UNICEF-Giga_Sudan](https:\u002F\u002Fgithub.com\u002FKamal-Eldin\u002FUNICEF-Giga_Sudan) -> Detecting school lots from satellite imagery in Southern Sudan using a UNET segmentation model\n\n- [building_footprint_extraction](https:\u002F\u002Fgithub.com\u002Fshubhamgoel27\u002Fbuilding_footprint_extraction) -> The project retrieves satellite imagery from Google and performs building footprint extraction using a U-Net.\n\n- [projectRegularization](https:\u002F\u002Fgithub.com\u002Fzorzi-s\u002FprojectRegularization) -> Regularization of building boundaries in satellite images using adversarial and regularized losses\n\n- [PolyWorldPretrainedNetwork](https:\u002F\u002Fgithub.com\u002Fzorzi-s\u002FPolyWorldPretrainedNetwork) -> Polygonal Building Extraction with Graph Neural Networks in Satellite Images\n\n- [dl_image_segmentation](https:\u002F\u002Fgithub.com\u002Fharry-gibson\u002Fdl_image_segmentation) -> Uncertainty-Aware Interpretable Deep Learning for Slum Mapping and Monitoring. Uses SHAP\n\n- [UBC-dataset](https:\u002F\u002Fgithub.com\u002FAICyberTeam\u002FUBC-dataset) -> a dataset for building detection and classification from very high-resolution satellite imagery with the focus on object-level interpretation of individual buildings\n\n- [UNetFormer](https:\u002F\u002Fgithub.com\u002FWangLibo1995\u002FGeoSeg) -> A UNet-like transformer for efficient semantic segmentation of remote sensing urban scene imagery\n\n- [BES-Net](https:\u002F\u002Fgithub.com\u002FFlyC235\u002FBESNet) -> Boundary Enhancing Semantic Context Network for High-Resolution Image Semantic Segmentation. Applied to Vaihingen and Potsdam datasets\n\n- [CVNet](https:\u002F\u002Fgithub.com\u002Fxzq-njust\u002FCVNet) -> Contour Vibration Network for Building Extraction\n\n- [CFENet](https:\u002F\u002Fgithub.com\u002Fdjzgroup\u002FCFENet) -> A Context Feature Enhancement Network for Building Extraction from High-Resolution Remote Sensing Imagery\n\n- [HiSup](https:\u002F\u002Fgithub.com\u002FSarahwXU\u002FHiSup) -> Accurate Polygonal Mapping of Buildings in Satellite Imagery\n\n- [BuildingExtraction](https:\u002F\u002Fgithub.com\u002FKyanChen\u002FBuildingExtraction) -> Building Extraction from Remote Sensing Images with Sparse Token Transformers\n\n- [CrossGeoNet](https:\u002F\u002Fgithub.com\u002Flqycrystal\u002Fcoseg_building) -> A Framework for Building Footprint Generation of Label-Scarce Geographical Regions\n\n- [AFM_building](https:\u002F\u002Fgithub.com\u002Flqycrystal\u002FAFM_building) -> Building Footprint Generation Through Convolutional Neural Networks With Attraction Field Representation\n\n- [RAMP (Replicable AI for MicroPlanning)](https:\u002F\u002Fgithub.com\u002Fdevglobalpartners\u002Framp-code) -> building detection in low and middle income countries\n\n- [Building-instance-segmentation](https:\u002F\u002Fgithub.com\u002Fyuanqinglie\u002FBuilding-instance-segmentation-combining-anchor-free-detectors-and-multi-modal-feature-fusion) -> Multi-Modal Feature Fusion Network with Adaptive Center Point Detector for Building Instance Extraction\n\n- [CGSANet](https:\u002F\u002Fgithub.com\u002FMrChen18\u002FCGSANet) -> A Contour-Guided and Local Structure-Aware Encoder–Decoder Network for Accurate Building Extraction From Very High-Resolution Remote Sensing Imagery\n\n- [building-footprints-update](https:\u002F\u002Fgithub.com\u002Fwangzehui20\u002Fbuilding-footprints-update) -> Learning Color Distributions from Bitemporal Remote Sensing Images to Update Existing Building Footprints\n\n- [RAMP](https:\u002F\u002Frampml.global\u002F) -> model and buildings dataset to support a wide variety of humanitarian use cases\n\n- [Thesis_Semantic_Image_Segmentation_on_Satellite_Imagery_using_UNets](https:\u002F\u002Fgithub.com\u002Frinkwitz\u002FThesis_Semantic_Image_Segmentation_on_Satellite_Imagery_using_UNets) -> This master thesis aims to perform semantic segmentation of buildings on satellite images from the SpaceNet challenge 1 dataset using the U-Net architecture\n\n- [HD-Net](https:\u002F\u002Fgithub.com\u002Fdanfenghong\u002FISPRS_HD-Net) -> High-resolution decoupled network for building footprint extraction via deeply supervised body and boundary decomposition\n\n- [RoofSense](https:\u002F\u002Fgithub.com\u002FDimitrisMantas\u002FRoofSense\u002Ftree\u002Fmaster) -> A novel deep learning solution for the automatic roofing material classification of the Dutch building stock using aerial imagery and laser scanning data fusion\n\n- [IBS-AQSNet](https:\u002F\u002Fgithub.com\u002Fzhilyzhang\u002FIBS-AQSNet) -> Enhanced Automated Quality Assessment Network for Interactive Building Segmentation in High-Resolution Remote Sensing Imagery\n\n- [DeepMAO](https:\u002F\u002Fgithub.com\u002FSumanth181099\u002FDeepMAO) -> Deep Multi-scale Aware Overcomplete Network for Building Segmentation in Satellite Imagery\n\n- [CMGFNet-Building_Extraction](https:\u002F\u002Fgithub.com\u002Fhamidreza2015\u002FCMGFNet-Building_Extraction) -> Deep Learning Code for Building Extraction from very high resolution (VHR) remote sensing images\n\n- [Global Collinearity-aware Polygonizer for Polygonal Building Mapping in Remote Sensing](https:\u002F\u002Fgithub.com\u002Fzhu-xlab\u002FGCP)\n\n- [Building Segmentation on LR-HR-SR Satellite Imagery](https:\u002F\u002Fgithub.com\u002FESAOpenSR\u002FSegmentation-Models-Benchmark) -> perform building delineation on different types of satellite imagery: Low-Resolution (LR), High-Resolution (HR), and Super-Resolution (SR). The goal is to compare the performance of segmentation models across these varying resolutions.\n\n- [UrbanGraphSAGE](https:\u002F\u002Fgithub.com\u002FOMUZ9924\u002FUrbanGraphSAGE) -> Graph Neural Network (GraphSAGE) for urban building footprint extraction from Sentinel-2 satellite imagery\n\n- [terratorch-building-segmentation](https:\u002F\u002Fgithub.com\u002FOMUZ9924\u002Fterratorch-building-segmentation) -> Fine-tuning Geospatial Foundation Models (Prithvi, Terra","该项目专注于卫星和航空影像的深度学习技术，提供了针对这些特定图像处理任务的全面技术概览。其核心功能涵盖了分类、分割、目标检测等关键任务，并介绍了适用于大规模图像数据集的各种神经网络架构与算法，如卷积神经网络、自编码器及生成对抗网络等。特别强调了在遥感领域的应用，包括云检测与移除、变化监测以及作物分类预测等。此外，还探讨了一些前沿技术，比如零样本学习和可解释的人工智能方法。此项目非常适合需要利用深度学习技术来分析卫星或无人机拍摄图像的研究人员和开发者使用。",2,"2026-06-11 03:24:07","top_topic"]