Underwater fish image segmentation is the key technology to realizing intelligent fisheries and ecological monitoring. However, the problems of light attenuation, blurred boundaries, and low contrast caused by complex underwater environments seriously restrict the segmentation accuracy. In this paper, RetinalConet, an underwater fish segmentation network based on bionic retina dual-channel and multi-module cooperation, is prop...