基于高斯平滑直方图匹配的图像间匀光算法  

Uniform Light Between Images Based on Gaussian Smooth Histogram Matching

作　　者：孙立辉[1] 张竟雄 SUN Lihui;ZHANG Jingxiong(School of Information Technology,Hebei University of Economics and Business,Shijiazhuang Hebei 050061,China)

机构地区：[1]河北经贸大学信息技术学院,河北石家庄050061

出　　处：《信息与电脑》2021年第21期45-47,共3页Information & Computer

摘　　要：由于不同的光照条件及各传感器之间的特性差异等原因,进行全景拼接的各图像间往往存在曝光差异。为了使拼接后的各图能够自然过渡,笔者首先分别统计待拼接两图重叠区域的直方图并对其进行高斯平滑预处理,然后进行灰度级的匹配。实验证明,该方法不仅能够有效消除图像间的曝光差异,还能改善图像中可能出现的灰度值分块不连续现象。Due to different lighting conditions and the characteristics differences between sensors, there are often exposure differences among panoramic mosaic images. In order to make the spliced images transition naturally, this paper first counts the histograms of the overlApping areas of the two images to be spliced, preprocesses them with Gaussian smoothing, and then matches the gray level. Experiments show that this method can not only effectively eliminate the exposure difference between images, but also improve the possible gray value block discontinuity in the image.

关 键 词：图像拼接 图像间匀光 直方图匹配 

分 类 号：TP751[自动化与计算机技术—检测技术与自动化装置]