This repository contains the implementation of our paper, "ERN: Edge Loss Reinforced Semantic Segmentation Network for Remote Sensing Images".
In this repository you will find:
- ERN prototxt and caffemodel;
- Python codes for training and testing;
- C++ code (VS2013 + OpenCV3.0) for simple shadow detection;
1. Caffe
- Please follow Yu Liu's instruction and install the modified version of Caffe from following link: https://github.com/Walkerlikesfish/CaffeGeo.git.
- Notice that if you use Caffe, please cite their papers.
- Prepare the data;
- Make sure the file path in solver.prototxt & train.prototxt & run_training.py is correct;
- Run python run_training.py in terminal.
- Prepare the data;(In our experiments, we first splited the original image into 256*256 )
- Make sure the file path in infer.prototxt & VH_infer.py & UAV_infer.py is correct;
- Run python VH_infer.py in terminal.
- We have split the images using https://github.com/Walkerlikesfish/HSNRS/tree/master/script/data_preprocessing
- Make sure the file path in VH_assemble.py & UAV_assemble.py is correct;
- Run python VH_assemble.py in terminal.
- Shadow detection for ISPRS Vaihingen Dataset (Windows)
- VS2013 + OpenCV3.0
- The contrast preserving decolorization has been used.
- Evaluation (Linux)
- Make sure the file path in testperformance.py (def xf_set_test_vh())is correct;
- Run python testperformance.py (def xf_set_test_vh()) in terminal.
This code is shared under a Attribution-NonCommercial 4.0 International (CC BY-NC 4.0) Creative Commons licensing.
You are free to:
- Share — copy and redistribute the material in any medium or format
- Adapt — remix, transform, and build upon the material The licensor cannot revoke these freedoms as long as you follow the license terms.
Under the following terms:
- Attribution — You must give appropriate credit, provide a link to the license, and indicate if changes were made. You may do so in any reasonable manner, but not in any way that suggests the licensor endorses you or your use.
- NonCommercial — You may not use the material for commercial purposes.
- No additional restrictions — You may not apply legal terms or technological measures that legally restrict others from doing anything the license permits.