Liying Lang

The polarization imaging technique leverages the disparity between target and background polarization information to mitigate the impact of backward scattered light, thereby enhancing image quality. However, the imaging model of this method exhibits limitations in extracting inter-image features, resulting in less-than-optimal outcomes in turbid underwater environments. In recent years, machine learning methodologies, particularly neural networks, have gained traction.