基于改进DCP算法的水下机器人视觉增强

Underwater Robot Visual Enhancements Based on the Improved DCP Algorithm

对水下机器人的视觉增强问题进行了研究,提出了基于改进暗通道优先(IDCP)的图像增强算法对单目视觉图像进行预处理.首先对水下图像的色偏与雾化等现象进行退化建模,然后通过计算亮暗通道的视差来获取图像的景深信息,以精确估计水体的背景颜色,同时计算得到相应的透射图.在此基础上,采取自适应比例因子选择策略对透射图进行后处理以得到具有更高对比度的图像恢复效果.此外,进一步采用颜色校正方法以去除残余色偏,并提高图像的整体亮度.针对水下多场景的实验结果表明,与常规方法相比,本文的IDCP方法能够在纠正色偏的前提下实现具有更高清晰度和亮度的视觉增强效果.

The visual enhancement problem of underwater robots is studied, and an image enhancement algorithm based on the improved dark channel prior(IDCP) algorithm is proposed to preprocess monocular vision images. Firstly, a degradation model of underwater image color slants and atomization phenomenon is developed. Then, the image of depth information is obtained through calculating the parallax between the brightness channel and the darkness channel to estimate the background color of the water accurately, and the corresponding transmission diagram is computed meanwhile. On this basis, the adaptive scaling factor selection strategy is adopted to postprocess the transmission diagram for the higher contrast image restoration effect. In addition, the color correction method is taken to remove residual color slants and enhance the overall brightness of the image. The experiment results in multiple underwater scenes show that the IDCP method can remove color slants and achieve visual enhancement with better resolution and brightness than the conventional ones.