Enhancing underwater scene visualization through a biological vision-inspired approach

Abstract In response to the challenge of degraded underwater imagery, we introduce a novel bio-inspired vision-based method for improving and rendering underwater scenes more visible. Unlike previous methods that concentrate solely on enhancing a single aspect of the image, our system incorporates a variety of techniques, such as bio-inspired color adjustment, visual scene enhancement, contrast improvement, and scene merging components. By harnessing the characteristics of biological vision, our method substantially enhances the visibility of underwater scenes with improved contrast and color accuracy. This leads to a more intuitive and comprehensive perception of the underwater environment and offers high interpretability along with minimal computational demands. Extensive testing reveals that our technique surpasses 14 advanced underwater image enhancement methods across four well-known datasets, highlighting its potential utility in fields including marine biology, oceanography, and underwater archaeology.