Study On Pseudo-color Enhancement Of Images In Subdivision Bands Of MWIR

This paper is pre-theoretical study for color enhancement device on subdivision bands of mid-wave infrared (MWIR) detector. The purpose is to provide easily hardware-oriented color enhancement method to get natural and clear pseudo-color enhancement images.This paper based on comparative imaging features of MWIR , and combined with the basic principle of pseudo-color enhancement , structure of the human eye retina domain, YIQ and lαβcolor space, support value transform(SVT) and so on.A multilevel and multiple pseudo-color enhancement meathod is presented, which enhances its color and gray respectively, then together to enhance. Its algorithm includes color enhancement, gray enhancement and composite enhancement. In order to obtain natural and comfortable images, by means of which construct neural networks with the human visual passive membrance equation, color enhancement makes use of bionics to increase the color contrast of differences among images. Gray enhancement adopts SVT fusion, which eliminates image saturation area and composites high frequency detail information, thus we can get clear intensity enhancement images. At last, composite enhancement would come into being color and gray together enhancement images, by lαβcolor space transform and weighted average means.Pseudo-color enhancement evaluation is anther key factor of image enhancement study. In this paper, subjective evaluation adopts color contrast, comfortable feeling and nature feeling. To make comparision, objective evaluation uses parameters extension, color comparison, color image degradation and space transform,etc. By means of pseudo-color enhancement evaluation, it draws conclusions that the presented algorithm has certain advantage and its specifications are improved.For exploring the hardware, a dual-channel image enhancement system based on SoPC is designed. It includes video capture and output module, image preprocessing module, mid-wave image feature extraction module, embedded controller module, image enhancement module.