Intestinal Endoscopic Polyp Segmentation based on Histogram Equalization ResNet (PE-ResNet)

The early screening of lesions is of great significance for the prevention of colorectal cancer, where the polyp segmentation is essential to it. Although the deep-learning based polyp segmentation has gained acceptable performance, the color difference between intestinal endoscopic images affects it much. Based on ResNet architecture, this study proposes an advanced PE-ResNet where the histogram equalization is used to decrease color influence. Experimental results on 5 datasets including ClinicDB demonstrate that the PE-ResNet achieves improvement in intestinal polyp segmentation.