水下视频图像的超分辨率重建与增强

摘要

According to the low contrast, strong noise and image blur of underwater video images, a new super-resolution reconstruction and enhancement algorithm based on projection onto convex set ( POCS) algorithm is proposed. In the process of POCS iterative constraint, the wavelet transform domain nonlinear function is used to de-noise and to enhance the details of the target. Then the non-decreasing energy constraint is used to revise iteration process. The experimental results show that the proposed algorithm has a good visual effect of super-resolution reconstruction and enhancement.%针对水下视频图像对比度低、图像模糊和噪声突出的特点,提出一种新的基于凸集投影(POCS)的超分辨率图像重建与增强算法.在POCS迭代约束的过程中,采用小波域非线性函数抑制噪声并突出细节,然后用能量非递减性约束集进行修正.实验表明,该方法具有较好的超分辨率重建与增强效果.