Joint Residual Learning for Underwater Image Enhancement

• Published in: Proceedings - International Conference on Image Processing pp. 4043 - 4047
• Main Authors: Hou, Minjun, Liu, Risheng, Fan, Xin, Luo, Zhongxuan
• Format: Conference Proceeding
• Language: English
• Published: IEEE 01.10.2018
• Subjects: Attenuation; Convolution; illumination balance; Image color analysis; Image enhancement; Image quality; joint residual learning; Task analysis; Training; transmission map estimation; Underwater image enhancement
• ISSN: 2381-8549
• DOI: 10.1109/ICIP.2018.8451209

Improving the quality of underwater image has a significant impact on many signal processing and computer vision applications, while haze-effect and color shift are main handicaps need to be surmounted. Due to the complexity of the underwater environmental factors, most existing image enhancement techniques cannot be directly applied to address this task. In this work, we develop a novel framework to jointly performing residual learning on transmission and image domains for underwater scene entrenchment. Indeed, our deep model consists of a data-driven residual architecture for transmission estimation and a knowledge-driven scene residual formulation for underwater illumination balance. Therefore, we can aggregate the prior knowledge and data information to investigate the underlying underwater image distribution. Moreover, by introducing adaptive exposure map, image colors will also be corrected accordingly. Experimentally, both quantitative and qualitative analysis can indicate outstanding effectiveness of the proposed algorithm, against state-of-the-art approaches.