Research Of Assessment Methods About Degraded Image Processing Effect

Image has become one of the most important sources of information in today’s society with its intuitive, efficient and precise characteristics and extensive adaptability. The accuracy of image quality assessment is an important research topic as well as a research challenge in the field of image processing. With the development of digital technology, the demand on image quality assessment technology is becoming higher and higher. Subjective assessment can identify the actual situation of image, but it is time-consuming and labor-intensive and can not be applied to engineering practice. The traditional objective image quality assessment, such as Peak Signal to Noise Ratio (PSNR), has been widely used as it is simple and easy to implement.But the assessment result of this method is not satisfactory because it ignores human visual characteristics. Objective image quality assessment includes the full-reference type that needs all of the original reference image information, semi-reference type that only needs part of the original image information and no-reference type that totally do not need the original reference image information. This paper focuses on the full-reference assessment method and no-reference assessment method.The theory of structural similarity is a new idea of image quality assessment. As the realization of the structural similarity theory, the structural similarity (SSIM) index has the advantages of simple and efficient. This thesis conducted an in-depth study of the structural similarity theory and the experimental result shows that the accuracy of SSIM is higher than that of simply calculate the image peak signal to noise ratio. So far, for the degraded images, the various clear treatments can only improve them to some extent, that is, there are still varying degrees of quality degradation in the images after processed by clear treatment. Therefore, this paper first starts from grayscale images, assessed the degraded images with structural similarity theory. The structural similarity assessment method based on gradient and the structural similarity assessment method based on dual-scale edge highlights the image edge and texture information and strengthens the proportion of the rich regions of the image edge texture in assessment. This thesis also involved the assessment of color degraded images, the experimental result shows that the structural similarity theory can also be applied in the assessment of color images.The no-reference image quality assessment is difficult to achieve because there is no original image information as reference. This paper applied the structural similarity theory into the no-reference image quality assessment, generated reference image with pending assessment image, assessed the image before and after clear treatment, integrated the assessment value and the integration result is the final assessment value of the processing effect of degraded image.