In this repository, we incorporate multi-scale features generated by different layers of U-Net and design a multi-scale skip connected and asymmetric-convolution-based U-Net (MACU-Net), for segmentation using fine-resolution remotely sensed images.
The detailed results can be seen in the MACU-Net for Semantic Segmentation of Fine-Resolution Remotely Sensed Images.
The training and testing code can refer to GeoSeg.
If our code is helpful to you, please cite:
R. Li, C. Duan, S. Zheng, C. Zhang and P. M. Atkinson, "MACU-Net for Semantic Segmentation of Fine-Resolution Remotely Sensed Images," in IEEE Geoscience and Remote Sensing Letters, doi: 10.1109/LGRS.2021.3052886.
numpy >= 1.16.5
PyTorch >= 1.3.1
sklearn >= 0.20.4
tqdm >= 4.46.1
imageio >= 2.8.0
Note: We select 15 images contained in GID, which cover the whole six categories:
GF2_PMS1__L1A0000647767-MSS1
GF2_PMS1__L1A0001064454-MSS1
GF2_PMS1__L1A0001348919-MSS1
GF2_PMS1__L1A0001680851-MSS1
GF2_PMS1__L1A0001680853-MSS1
GF2_PMS1__L1A0001680857-MSS1
GF2_PMS1__L1A0001757429-MSS1
GF2_PMS2__L1A0000607681-MSS2
GF2_PMS2__L1A0000635115-MSS2
GF2_PMS2__L1A0000658637-MSS2
GF2_PMS2__L1A0001206072-MSS2
GF2_PMS2__L1A0001471436-MSS2
GF2_PMS2__L1A0001642620-MSS2
GF2_PMS2__L1A0001787089-MSS2
GF2_PMS2__L1A0001838560-MSS2
You can download datasets and prepare the files to ./datasets folder.
Note:
Transfer the lable images form RGB Format to Grey-scale Map.
Augment the training set by the technology mentioned in our paper.
Fig. 1. Comparison of (a) U-Net, (b) U-Net++， and proposed (c) MACU-Net 3+. The depth of each node is presented below the circle.
Fig. 2. Visualization of results on the WHDLD dataset (the left) and the GID dataset (the right).