The Research Of Contrast Enhancement Processing On Wide-bits Image In Various Complex Scenes

Image enhancement is a kind of image processing method which depends on somespecific needs, enhancing some useful information while weakening or removing someirrelevant information in the original image, or providing better input for otherautomated image processing techniques. The design of enhancement methods closelyrelates to the aim of applications.Due to the influence of scene conditions, visual effects of images that taken in highdynamic range scenes, dark environment or under special lighting conditions are usuallynot satisfied. And because of the requirements in practical engineering applications,such as selection and control of the different enhancement degree, switching betweenthe various scenes, the applicability for various objects and so on, the research offinding a processing method, which has good self-adaptability within a certain rangeand could adjust the mapping curve intelligently based on the input data, has become ahotspot and difficulty in the field of digital image processing and computer vision.In addition, for the purpose of obtaining high-quality images in the project, moderncameras with wide-bits (12,14,16bits wide) are usually used to acquire enoughoriginal information. Nevertheless, to display such images on a PC (8bits wide), somekinds of data selection and conversion processing is necessary to make the outputimages with clear general contrast while not missing the important details.This paper has intensively researched several image enhancement theory and techniques,analyzed their problems and shortcomings in practical applications, and then putforward several methods to improve the enhancement effect. This paper focuses onwide-bits images in various complex scenes, analyzes the characteristics of visibleimages and infrared images as the two major categories, presents the ARPOSHEalgorithm for wide-bits visible images and the ASMM algorithm for wide-bits infraredimages with the after adaptive pseudo-color processing method. The experimentalresults show that, the algorithm proposed in this paper could keep the local details andthe mean of the original brightness. It well enhances the contrast of the images,improves their visual effects, and makes the processed images have rich level sense,more consistent with human’s perception. In addition, to better meet the actualengineering requirements, the algorithm also has a satisfying adaptability on different scenes, different objectives, different input data width, as well as a reasonable timecomplexity.