Remote Sensing Image Restoration: An Adaptive Reciprocal Cell Recovery Technique

Improving the quality of remote sensing images is of great value for subsequent applications. In order to reduce the noise and blur of remote sensing images, the deblurring restoration technology is studied in this paper. Firstly, the acquisition of remote sensing image and the causes of image degradation are briefly analyzed. It was found that noise, blur and aliasing had great influence on image quality. Then, based on adaptive reciprocal cell, a method of titling mode restoration is proposed, which was achieved by generating adaptive reciprocal cell, extracting effective spectrum and deblurring. In order to verify the validity of the method, TV model, ARCTV model and ARCNLM model were used to restore a synthetic image and two scene images. The results showed that the Signal Noise Ratio (SNR) of the ARCNLM model was higher than that of other models, and its mean square error (MSE) was lower than that of other models. Moreover, ARCNLM model had better deblurring effect than other models, and it could not only effectively remove aliasing, but also could maintain the texture and details of the image. The experimental results suggested the effectiveness of ARCNLM model in image restoration and provided some basis for its practical application in image restoration.