Research On Color Image Super Resolution And Parallel Processing Technology

Image super resolution (SR) reconstruction is the technique that constructs high resolution images from one or several observed low resolution images of the same scene. The goal is to increasing the high frequency components and removing degradations caused by imaging process. Since the SR technique can overcome the resolution limitation on the image equipments in a certain extent, it has become one of the most active research areas in digital image processing. This dissertation is mainly about color image SR and parallel implementation of the SR algorithm. The major work is as follows:(1). Based on colorization, the SR method for color image is proposed. This method makes use of the relation between the luminance and the chrominance. The back projection algorithm can be employed to further improve this method.(2). Combining the joint bilateral filter, adaptive joint bilateral filter, which ships over the parameter-tuning problem in a certain extent, is proposed for color image SR. Iterative process using cross bilateral filter and colorization is introduced to improve the quality of reconstructed high resolution color image.(3). As the quality and processing speed of SR algorithms can hardly both satisfy users'expectation, openMP is employed to parallel optimize the SR methods.This thesis presents some algorithms based on the research of color image SR. Simulation experiments results demonstrate that these new algorithms can preserve the color edges and avoid the color aberration preferably on reconstructed image. Parallel optimization results show that the parallel processing method can increase the speed by a large margin and does not affect the reconstruction quality.