Research And Implementation Of High Dynamic Range Image Mapping Algorithm Based On Bilateral Filtering

With the continuous development of digital image technology, the demand for the image of the real scene is further increased. The color space of high dynamic range image is large,and the information level is rich. It can better record and display the real scene of the light and dark details than traditional images and has been widely concerned by researchers. It has been applied in different fields, such as meteorology, satellite, military, medical and so on. The existing display device can not directly display high dynamic range images, so it need to make the high dynamic range image into low dynamic range image by the tone mapping algorithm. The tone mapping algorithm is a hot research area of high dynamic range image compression mapping.This paper studies the basic theory and the method of high dynamic range image, and analyzes the multi storage structure of high dynamic range image, and makes a detailed study on the global and local tone mapping algorithm, and analyzes the advantages and disadvantages of various tone mapping algorithms. On the basis of the analysis of the bilateral filtering algorithm, this paper proposes an improved scheme of bilateral filtering algorithm based on the single character of the algorithm parameters and the low degree of the results. According to experimental analysis, the bilateral filter range parameters increases with the increase of image dynamic range and variance. This paper can set the bilateral filter parameter according to the characteristics of the image, realizes the adaptive adjustment of parameters, thus enhancing the level of detail of information. The high dynamic range image is processed by the bilateral filtering algorithm and is divided into the detail layer and the basic layer. This paper mainly carries on the image compression mapping on the basic level of the image.In order to increase the contrast of the image, the paper carries on the function of the log domain mapping and the local contrast stretching of the image base layer. In order to solve the problem of high computational complexity and detail information of image details in local mapping, this paper introduces a weighted guided filter, which is based on the original image as the guide image, and after a linear processing method the information of the filtered image is calculated. The filter on the edge of the image has good treatment effect, effectively remove halo phenomenon, and its time complexity is a linear level, withfaster processing speed. At the end, the image tone mapping is completed by the fusion of the compressed basic level image and the detail.In order to verify the effectiveness of the proposed algorithm, a multi group of high dynamic range image is tested and compared with a variety of tone mapping algorithms,and the results are given in this paper. In this thesis, we use subjective evaluation and objective evaluation method to evaluate the image quality. Experimental results show that the proposed algorithm has faster processing speed and better image results.This paper has completed the improvement and simulation verification based on the bilateral filtering tone mapping algorithm. It has shown the effectiveness of the proposed algorithm. This thesis has some reference value for the research and optimization of the high dynamic range image tone mapping algorithm.