Semantic-segmentation-methods-for-landslide-detection
The descriptions of principal files in this project are introduced as follows:
- ./landslide.kml : the construction of landslide imagery datasets is of great significance to the monitoring and analysis of landslides. Up to now, open landslide imagery datasets are still so scarce that research in this field is developing slowly. Therefore, we create an open labelled remote sensing landslide dataset in which landslides are determined based on field investigation and remote sensing imagery interpretation, as supported by the China Geological Survey (CGS) of China.
- ./data/ : processed data are under in this folder, and the results of experiments will be generated here. The download instructions of dataset are shown in "download.txt".
- ./data_util/
- dataset.py : codes for generating the dataset for our model
- ./models/
- resnet/ : codes for resnet models
- tools/ : codes for some model tools
- vgg/ : codes for vgg models
- deeplab3plus.py : codes for DeepLab v3+ model
- deeplabv32.py : codes for DeepLab v3 model
- fcn.py : codes for FCN models
- gcn.py : codes for GCN model
- padded_unet.py : codes for UNet model
- pspnet.py : codes for PSPNet model
- ./pretrained_models/ : pretrained_models (e.g. 3x3resnet50-imagenet.pth). The download instructions of pretrained models are shown in "download.txt".
- ./caffe_pb2.py : codes generated by the protocol buffer compiler for PSPNet building. source: caffe.proto
- ./demo.py : codes for demo
- ./evaluate.py : codes for evaluating models
- ./main_withargu.py : the main function for training and testing models
- ./plot.py : code for the visualization of experimental results
- ./utils.py : containing useful codes that are required in the project (e.g. dataloader, model save, model train, model test, metrics calculation)
- PyTorch 1.5.0
- tqdm
- numpy
- pandas
- torchvision
- pillow
- pydensecrf
- albumentations
- opencv