CAJ | 학술논문

Retinex算法是图像增强的常用方法.基于Retinex理论提出一种新的视频光照增强算法,结合图像全局及局部灰度特性引入补偿系数,增强暗区域光照的同时克服了传统算法处理后的人工痕迹,使处理结果更加自然;同时,算法用引导滤波估计照度图像,在避免光晕现象的同时可以加快算法处理速度.实验结果表明,该方法简单有效,视觉效果提升显著.
Retinex산법시도상증강적상용방법.기우Retinex이론제출일충신적시빈광조증강산법,결합도상전국급국부회도특성인입보상계수,증강암구역광조적동시극복료전통산법처리후적인공흔적,사처리결과경가자연;동시,산법용인도려파고계조도도상,재피면광훈현상적동시가이가쾌산법처리속도.실험결과표명,해방법간단유효,시각효과제승현저.
Retinex algorithm is a common image enhancement method. Based on Retinex theory,a new video illumination enhancement algorithm is proposed,in which the compensation factor is introduced in combination with global and local gray-scale characteristics of images. It can enhance the illumination in dark areas and overcome the artificial marks on images pro-cessed by traditional algorithms,which can make the processing results more natural. The guided filtering is applied to estima-tion illumination image by the algorithm,which can avoid halo phenomenon and accelerate the processing speed of the algo-rithm. The experimental results show that this method is simple and effective,with which the visual effect is promoted obviously.