Research About Image Enhancement Algorithm In Video Image Restoration

Video image restoration is one of the technologies that need urgent solution. It can be used to restore aged video images stored in simulate films or tapes. The image enhancement algorithm discussed in this paper is the fore work of video image restoration, and it is also important for the image quality. The aim of image enhancement is to process the degenerated images to recover to the ideal image before degeneration. The main enhancement method used nowadays is image filtering, which should remove the noise and preserve the details as well. Because of noise and detail's blending on the frequency range, it is incompatible to removing the noise and preserving the details during the process. The Partial Differential Equation method developed recently can remove the noise and preserve the details on edges. It attracts a lot of attention.The main work of this paper is composed of the following two parts:Firstly, the integration of the Curvature-Preserving PDE and the shockfilter. The former has its special advantage in remove the noise, which get abetter result between remove the noise and preserve the details. But in videoimage restoration, the amount of the images is very large and only a sequence of images can form a vivid video. Enhancing the edges is extremely important. Therefore, we emphasize on the improvement of the PDE by adding a shock filter operator. The shock filter's main function is creating a shock when detecting an edge to enhance the edge information.According to the contaminative effect of the image, we handle them differently. For example, for the images with low contrast we can strength the contrast first, or equalize the Histogram of the image to enrich the information on edges and get a better effect, then restore it using the improved algorithm.Secondly, because of the connection between the frames before and after the current frame, we use a sequence of low-resolution images to reconstruct an image with high resolution. Making use of the continuity of the images and the connection in contents, we propose to use a super-resolution method to reconstruct a more detailed image. Then we handle the noise in the image and enhance the edges at the same time.Based on the classical algorithm, we discuss a PDE method to denoise and enhance the images, which has a better visual effect than classical methods. From the results of the experiments, we can find that the images are much clear, and richer in edge information.