Research On Haze Removal Algorithm

With the rapid development of computer technology, sensor technology, image processing, computer vision and other related technologies, visual systems had been more important in every field of our life, such as video surveillance, remote sensing detection, autonomous navigation, aerial, military reconnaissance and so on. However, since the weather always changing,especially in the bad weather like haze, the image shooted by visual systems have some problems such as serious degradation, reduced contrast, low resolution, and poor color, which makes the outdoor visual system working aberrant or having a significant reduction in performance. At the same time, the image defogging involves many multidisciplinary theoretical knowledge such as optics, atmospheric science, image processing, computer vision and so on. The research on image defogging algorithm is good for the integration and development of related technologies. Therefore, the study of image defogging algorithm has the extremely important practical significance because of its widely application requirements. The contact with many other theoretical knowledge makes it great academic valuable.The author made the image obtained in the haze weather clearly by using the image processing technology, optical, and atmospheric science. The main research works in this paper are as follows:(1) At first, the research status at home and abroad of image defogging are summarized. And then the author simple introduced the main theoretical knowledge related to image defogging, such as histogram equalization, space sharpen filter, Retinex theory, atmospheric scattering model. It laid a solid theoretical basis for the later research on image defogging.(2) Image defogging algorithm based on image enhancement is studied. The haze removal algorithms based on histogram equalization, contrast-enhanced, image Retinex theory and homomorphic filtering are studied emphatically. All above algorithm were simulated. According to the results of the simulation, the author analyzed the advantages and disadvantages of various algorithms, and the scope of application of these algorithms.(3) The image defogging algorithm based on the atmospheric scattering model is studied deeply. At first, with the help of atmospheric scattering model, the image defogging algorithm based on dark channel priori is researched. The author verify the performance by simulation, and analyse the disadvantages of this algorithm. And then, the sky segmentation based image defogging, improved transmission refinement, real-time defogging algorithm based on mean filter are studied for the problems such as the ineffectiveness to sky area, rough transmittance estimation, slow processing speed. At last, the optimized contrast image defogging algorithm is studied. This method implements a new transmission rate estimation method by Building energy function. In order to verify the performance of the image defogging algorithms, all the algorithms have been simulated. The author analyze the results of the simulation, and summarize their advantages and disadvantages.