请问论文中所述Synthetic dataset 选取6200张图像，具体是怎么实现的？

danfenghong / IEEE_TIP_UIU-Net

Xin Wu, Danfeng Hong, Jocelyn Chanussot. UIU-Net: U-Net in U-Net for Infrared Small Object Detection, IEEE Trans. Image. Process., 2022, 32: 364-376.

Apache License 2.0

136 stars 15 forks source link

Open BLUE-coconut opened 1 year ago

BLUE-coconut commented 1 year ago

请问论文中所述Synthetic dataset 选取6200张图像，具体是怎么实现的？有开源的代码吗？文中只说了手动去除预处理后具有零值或负值的图像，那具体是保留了哪些图像呢？