Quaternion special least squares with Tikhonov regularization method in image restoration
摘要
In this paper, a color image restoration method based on Quaternion Special Least Squares with Tikhonov regularization method (QSLST) is proposed. In order to solve the problems of blur damage and noise degradation that may occur in the process of color image imaging, this method uses quaternion to represent the color image, and can consider the multi-channel spatial information of the image while processing the image restoration task. By introducing Tikhonov regularization term, the ill-conditioned problem in image restoration is effectively solved, and the stability and robustness of the restoration results are improved. The experimental results show that the proposed algorithm performs well in both quality and efficiency of color image restoration, and can effectively remove noise and blur in the image and recover clear image details.