CAJ | 학술논문

为了实现人眼感知信息最大化，结合人眼视觉系统亮度阈值特性提出一种具有目标函数的图像增强算法。首先在全局灰度映射约束条件下给出感知图像的计算方法，并以感知图像与原图像之间的互信息作为目标函数；然后采用动态规划算法实现了全局最优的灰度合并策略，以合并图像中人眼无法分辨的灰度；最后通过调整合并后灰度的取值实现目标函数最大化。实验结果表明，该算法具有良好的图像增强效果，所提目标函数合理有效。
위료실현인안감지신식최대화，결합인안시각계통량도역치특성제출일충구유목표함수적도상증강산법。수선재전국회도영사약속조건하급출감지도상적계산방법，병이감지도상여원도상지간적호신식작위목표함수；연후채용동태규화산법실현료전국최우적회도합병책략，이합병도상중인안무법분변적회도；최후통과조정합병후회도적취치실현목표함수최대화。실험결과표명，해산법구유량호적도상증강효과，소제목표함수합리유효。
In order to achieve the maximum of human eye perceived image information, an image enhancement algorithm with objective function was proposed considering human visual system brightness threshold charac-teristic. Firstly, put forward a method to obtain perceived image under the global grayscale map constraints, and make the mutual information between perceived image and the original image as objective function; Sec-ondly, achieve a global optimum grayscale merger strategy by using dynamic programming algorithm to merge the grayscales which human eye could not distinguish; Finally, adjust the combined grayscale to maximize the objective function. Experimental results demonstrate the proposed algorithm has a good treatment effect, and the proposed objective function is reasonable and effective.