Mild Haze Weather Degraded Image Processing And Optimization Studies

Lead to reduced visibility in fog and haze, affecting people’s daily life and travel, In order to better restore the scene information of the image of haze, It is very necessary to improve the quality of the original image for the haze image processing recovery. In this paper, bad fog and haze image processing recovery technology is of great significance.Currently, There are two main method for the haze image processing recovery:One is the use of digital image processing method for processing, the other is from the physical principles of the analysis of fog and haze image degradation. Mathematical and physical model through the establishment of degradation of image processing recovery. Use of mathematics and physical model of deeper mining the nature of image degradation. Therefore, in the analysis of existing physical model based on, Presents an improved Processing Method, Achieved very good recovery results.The paper is mainly carried out the following:1. Access to relevant information and literature, analysis of domestic and foreign digital image processing and the advantages and disadvantages of the atmospheric degeneration mathematical and physical models, Specific research presented in this article.2. Physical principles described in the fog and haze, Physical properties,describe the atmosphere and atmospheric scattering foundation theory.3. The introduction of the system for the basic theory of digital image processing, using the Histogram linearization processing method, homomorphic filtering processing method, color image processing method, of the Retinex theory methods haze image processing. By objective image quality assessment parameters contrast and subjective visual contrast, the Retinex theory treatment is relatively good.4. To establish a mathematical model method using minimum mean square error filtering recovery method, Autocorrelation function of the minimum mean square error filtering recovery method and under the influence of atmospheric turbulence create a physical model to deal with recovery haze image. Found that the autocorrelation function of the minimum mean square error filter restoration method in comprehensive evaluation of image quality is better than the Retinex theory process recovery method.Create a physical model of the atmospheric turbulence under the influence of processing recovery works best.5. Proposed the improved method of the physical model, Optimization to reduce atmospheric turbulence part of the impact, Subjective and objective evaluation for the haze image processing recovery and image quality, found in the subjective and objective evaluation of the restored image entropy, clarity, contrast restoration effect of the physical model is superior to the effects of atmospheric turbulence.