Improvement And Implementation Of Signal Image Dehazing Algorithm

As everyone knows, image is the main source for human to acquire information, therefore, the accuracy of the image information is extremely important. However, in real life, due to haze reasons the image we have obtained is degraded. The degraded image not only reduces the application value of image, but also brings many negative influences to our real life and work. So, it is necessary to do research on haze digital image dehazing technology.This paper proposes an improved dehazing algorithm based on single image, the algorithm starts from the reasons and mechanism of image degraded, according to the propagation of atmospheric light obtains a physical model of degraded image, and then inverses the physical process of image degradation, based on the principle of linear stretch and then recover a clear image close to the original, with the help of MATLAB experiment platform, implements the algorithm, and the dehazing results is compared with Fattal image dehazing algorithm. The experimental results show that the improved algorithm can achieve the image dehazing, at the same time, eliminates stains and remains the real color of the image.The main contents of this paper include:(1) The research status of image dehazing technology in domestic and international and relevant theory of image dehazing technology is depicted.(2) According to the principle of atmospheric scattering deduces the physical model of degraded image. In the constant albedo images, uses the physical model to solve the algorithm of image dehazing, and extends to the algorithm of multi-albedo image dehazing.(3) Fattal algorithm of image dehazing is improved. When a certain area of the image is partical white, the image restored by Fattal algorithm of image dehazing is easy to produce stains, and sometimes the image appears serious color distortion. To solve the problem,we proposes a improvement based on linear stretching principle.(4) With the help of MATLAB experiment platform, we have realized Fattal algorithm of image dehazing and improved algorithm, then compares and analysises the two experimental results. The experimental results show that the improved algorithm can achieve the image dehazing, while eliminates stains, and remains the real color of the image.(5) At last, the future work of image dehazing is prospected, it provides a reference for the further research and development on image dehazing technology.