面向人眼探测识别的灰度图像伪彩色化方法  

Pseudo-color processing of gray images for human visual detection and recognition

作　　者：蔡铁峰[1,2,3] 朱枫[1,3] 郝颖明[1,3] 范慧杰[1,3] 

机构地区：[1]中国科学院沈阳自动化研究所,辽宁沈阳110016 [2]中国科学院大学,北京100049 [3]中国科学院光电信息处理重点实验室,辽宁沈阳110016

出　　处：《红外与激光工程》2015年第B12期213-219,共7页Infrared and Laser Engineering

基　　金：国家自然科学基金青年项目(61401455)

摘　　要：人眼能分辨的颜色数量远多于人眼能分辨的灰度级,把灰度图像用伪彩色显示出来,人眼可以更好地感知图像中的景物信息。区别于现有方法,明确以有利于人眼探测识别为目标,给出了灰度图像伪彩色化后人眼能感知到的景物信息量的两项评价指标,在评价指标指导下,提出了一种灰度图像伪彩色化方法。该方法在确保灰度图像中各灰度级映射成人眼可分辨的不同颜色前提下,使尽量多的景物信息被人眼感知到。实验结果表明,相较于现有方法,该方法能使更多景物被人眼看清。该方法满足实时处理图像的要求。The colors that can be distinguished by human eyes are much more than the gray levels that can be distinguished. More objects information of the gray image can be obtained through the pseudo-color processing. In a fundamental departure from the current pseudo-color processing methods,quantitative metrics for human visual detection and recognition were devised to measure perceivable objects information after pseudo-color processing, and a pseudo-color processing method was proposed under the metrics. At first, it was ensured that different gray levels were mapped to distinguishable colors. Then, a gray-to-color mapping function was created to make more objects information be perceived. In the comparison experiments, the proposed method can make more objects noticed by people.The proposed method can process images in real time.

关 键 词：探测识别 图像对比度增强 伪彩色 

分 类 号：TP391.41[自动化与计算机技术—计算机应用技术]