针对传统水下偏振差分成像模型在图像处理过程中出现失真的问题,提出改进的水下偏振差分成像模型。该方法通过增加目标透射光辐射强度的反演过程,同时引入修正参数,对目标透射光辐射强度进行校正,有效避免了传统水下偏振差分成像模型中由全局参量直接反演场景图像造成的图像失真和局部灰度值为负数的情况。利用FD-1665偏振成像仪,开展基于线偏振光主动照明的水下目标成像探测实验,获取偏振图像数据,并利用客观图像质量评估参数对处理结果进行对比分析。结果表明,改进后的模型相较于传统偏振差分模型,图像对比度提高了40%以上,故改进后的偏振差分成像模型能更有效地识别水下目标。

This study proposes an improved underwater polarization-difference imaging model to overcome the problem of distortion that occurs in image processing based on conventional underwater polarization-difference imaging model. By this method, the inversion process and correction parameters are introduced for the correction of target forward-scatter intensity,thereby avoiding the image distortion caused by the global parameters’ direct inversion and the negative local gray values associated with the traditional underwater polarization-difference imaging model. Using active linear-polarization illumination and underwater-target imaging detection, we obtain the polarization image data with an FD-1665 polarization imager. Using objective image-quality evaluation parameters, we compare multiple sets of experimentally processed images. Results show that the proposed model effectively improves the image-contrast value by more than 40% in comparison with those obtained by the conventional model, indicating the effectiveness of the improved polarization difference imaging model.