基于水下光照不均匀成像模型的图像清晰化算法

A visibility improving algorithm based on underwater imaging model with non-uniform illumination

为了克服水下图像清晰度低和光照不均问题,首先在水下光照均匀成像模型的基础上建立水下光照不均匀条件下的成像模型,进而提出新的水下图像清晰化算法。在算法中,首先在小波变换低频子带上实现了介质散射光和光照变化混合图像的快速估计与去除,然后将得到的图像分割成亮斑区和散射区,并分别进行增强处理。实验结果表明,本文提出的算法可以显著降低处理时间;既能有效提高水下图像的清晰度,又能同时去除水下图像的光照不均问题;相邻视频帧处理结果具有良好的一致性。

In order to overcome the problems of low visibility and non-uniform illumination for underwater images,an underwater imaging model with non-uniform illumination was first derived based on the underwater imaging model with uniform illumination, and then a visibility enhancing algorithm was proposed from the developed model. In the algorithm, the mixing image with medium scattering light and light change was first estimated and eliminated quickly by using the low-frequency sub-band information of wavelet transform, and then the obtained image was segmented to bright speck region, and was scattering region and enhanced according to the segmented result respectively. The experimental results show that the proposed algorithm can significantly reduce the processing time and effectively improve the visi bility of underwater degraded images while eliminating the effect of non-uniform illumination. Moreover, the processing results of consecutive frames of video sequence have good consistency.