Research On Pseudo High Dynamic Range Technology Based On Image Enhancement

At present,people's demand for high quality image is growing,as a result,image enhancement technology has become a hot spot of research.High dynamic range technology is an effective image enhancement technology,it can not only enhance the details of the image but also improve the contrast of it obviously.However,because of the high cost of hardware implementation,the technology is difficult to be generalized and applied.Therefore,it is of great significance to study an image enhancement method,which is effective and easy to be generalized and applied.Based on the practical application requirements,we propose a pseudo high dynamic range algorithm based on image enhancement,which can effectively improve the image quality and the enhaced image can achieve the effect similar to that of high dynamic range image.The main content of this paper is as follow: firstly,in order to improve the robustness of the enhancement algorithm,a classification method of images and videos is proposed.The proposed method divides images and videos into achromatic,bright,foggy and clear by analyzing the features of images and videos,e.g.,color channel difference feature,dark channel feature,furthermore,the appropriate enhancement methods are studied according to the characteristics of different types.Secondly,adaptive parameter setting methods for enhancement algorithms of different types of images and videos are proposed.As for the contrast limited adaptive histogram equalization algorithm,the block number and the shear height are calculated automatically according to the brightness distribution and contrast of the image;as for the dark channel prior algorithm,the intensity of image enhancement is calculated automatically according to histogram feature.Thirdly,the dark channel prior algorithm is improved and a transmission optimization method is proposed.In order to accuratly estimate the transmission in the bright area,the bright area is detected according to the brightness and saturation characteristics of the image,and the transmission of the bright area is optimized by the tolerance method,moreover the transmission optimization rate is improved by guided filter.Finally,a pseudo high dynamic range algorithm is proposed based on the combination of the above classification method,adaptive parameter setting method and transmission optimization method,and inter-frame weighting method is used to calculate the airlight to improve the smoothness of video.The proposed algorithm can effectively enhance different types of images and videos,and it is robust and easy to be implemented,so it could be easy to be generalized and applied.In order to verify the effectiveness of the proposed pseudo high dynamic range algorithm,a large number of images and videos are enhanced,after that,the enhancement results are evaluated and analyzed.The subjective enhancement effect evaluation of typical images shows that 91.83% of the testers think that the images enhanced by our algorithm are better than the original images,and 68.5% of the testers think that the images enhanced by our algorithm are better than those of the unimproved algorithm.The objective enhancement effect evaluation of typical images shows that the average increase ratio of image entropy is 3.66%.The enhancement effect evaluation of typical video shows that the enhanced videos are smooth without gain mutation.The above experimental results indicate that pseudo high dynamic range algorithm is an effective enhancement algorithm.