Rearch Of Image Dehazing And Super-resolution Reconstruction

Human vision is the main access of access to information, and there are about more than70%information obtained through the eyes of people. With the development of multimedia technology and the use of video image, people still hope to get the picture with a higher quality. Usually in traffic scene we are required to face recognize the people’s face and the license of cars, and in the maritime supervision we are required to keep watch on the target, discover risk, fix the position and search for the target timely. But in the rain, fog and other inclement weather, most of the visual system, especially the video monitoring system due to the influence of the external environment, leading to the image signal being out of shape, reducing the ratio of signal to noise severely and reducing the imaging quality, and so on. Therefore, it is very important to deal with images effectively and improve the image’s quality. At present, the way of improving the clarity of the image can be divided into:the enhanced method based on image’s contrasts, the way based on Retinex theory, the enhanced method, the image’s restoration method of using atmospheric degradation model, the way based on atmospheric scattering theory, the method based on the technology of information’s fusion and the theory of multi frame images’ super-resolution enhancement.The main contents of this paper are:Firstly, On the basis of above, I Focus on studying the histogram equalization method, the method based on Retinex theory and super-resolution image reconstruction technical methods. And a variety of frequency and spatial advantages and disadvantages of MAP and POCS method carried out a detailed analysis and comparison.A single image, I have introduced the theoretical to remove the fog from images, and I focus on the histogram equalization method and images’enhanced method basing on the Retinex theory. Then on the basis of the theory, I use the way of writing a program to realize defogging and improve images’clarity.Super-resolution reconstruction:firstly, I have introduced the theory of super-resolution reconstruction, and emphatically introduced the POCS methods. Otherwise I have introduced the reconstruction of the reference frame, motion estimation and the reference frame modification and these aspects’effects in dealing with the pictures. This paper presents an improved way based on the original diamond search method. And then through the programming to realize the diamond search method and the improved diamond search method of POCS respectively and comparing the results of the two different approaches, I found that the improved diamond search method is better than the original diamond search method, because the effect is a little better and the average search calculation points is less. So in this paper, the modified method is a effective method.