DOPNet: Dense Object Prediction Network for Multi-Class Object Counting and Localization in Remote Sensing Images

Code DOPNet: Dense Object Prediction Network for Multi-Class Object Counting and Localization in Remote Sensing Images.

Pre-trained models

Baidu Cloud : n551

Environment

We are good in the environment:

python 3.8

pytorch 1.10.0

numpy 1.21.4

matplotlib 3.6.0

mmcv-full 1.4.8

Usage

We provide the test code for our model. The DOPNet_RSOC.pth model is adapted on the RSOC dataset. We randomly select an image from the RSOC_small-vehicle dataset and place it in the image folder. And you can either choose the other images for a test.

We are good to run:

python test.py --model_state ./weights/DOPNet_RSOC.pth --out ./out/result.png

We will release more trained models soon. The core code will be released after the journal paper is accepted. Please see the paper for more details.

Acknowledgement

Thanks to these repositories

• C-3 Framework
• mmcv

If you have any question, please feel free to contact us. (ceoilmp@whu.edu.cn and gcding@whu.edu.cn)