An efficient evaluation scheme for remote sensing images

During image post processing, the restoration operation is required to increase the quality of the image. According to traditional evaluation methods, the evaluated object is the whole image. Obviously, some image details are not favorable for quality evaluation, unfortunately, these maybe cause evaluation error. Furthermore, it is inefficient because of lots of computation for entire image. In this paper, we puts forward using a specified region of interest (ROI) to evaluate the image quality instead of whole image. The experimental results show that, the selection method of the specified ROI is effective, and the evaluation result is consistent with the objective evaluation of entire image.