摘要
在彩色图像特别是窄动态范围的彩色图像增强中,为了保留更多图像细节,提出一种直方图受限的窄动态范围的彩色图像细节增强方法。该算法在RGB空间分别对三分量进行直方图受限运算,设置分量直方图的累积概率分布阈值C_T,将分量图像直方图分割成2个直方图H_1和H_2;将H_2(受限的直方图)非线性映射至H_1的均匀分布直方图中;合并RGB三分量输出彩色图像。实验结果表明,与同类算法相比,该算法在提高对比度的同时,具有最高的信息熵值,且输出的彩色图像视觉效果自然、清晰。
In order to preserve more image details in colour image enhancement,especially in narrow dynamic range,we proposed a histogram-limited color image detail enhancement for narrow dynamic range.The algorithm performed histogram constrained operation on three-components in RGB space and set the cumulative probability distribution threshold CT of component histogram to divide the component image histogram into two histograms H1 and H2.We mapped H2(constrained histogram)nonlinear to H1 uniform distribution histogram,and merged three components of RGB to output color images.The experimental results show that compared with other similar algorithms,the proposed algorithm has the highest entropy value while improving the contrast,and the visual effect of the output color image is natural and clear.
作者
王海峰
章怡
杜卓明
Wang Haifeng;Zhang Yi;Du Zhuoming(Information Center,Jiangsu University of Technology,Changzhou 213001,Jiangsu,China;School of Computer Engineering,Jiangsu University of Technology,Changzhou 213001,Jiangsu,China)
出处
《计算机应用与软件》
北大核心
2019年第5期187-193,248,共8页
Computer Applications and Software
基金
国家自然科学基金项目(61402206)
江苏省高校自然科学基金项目(18KJB520012)
关键词
图像细节
对比度
直方图受限
信息熵
动态范围
Image details
Contrast
Limited histogram
Information entropy
Dynamic range
