基于凸包和背景先验的图像显著性检测  

Image significance detection based on convex hull and background prior

作　　者：李春华[1] 秦云凡 Li Chunhua;Qing Yunfan(school of information science and engineering,hebei university of science and technology,Shijiazhuang 050000,China)

机构地区：[1]河北科技大学信息科学与工程学院

出　　处：《信息通信》2020年第5期50-52,共3页Information & Communications

摘　　要：视觉显著性检测是一种模仿人类视觉系统并从复杂场景中选择感兴趣区域的机制,为使显著性区域检测更加准确,提出了基于凸包和背景先验的图像显著性检测。首先对图像进行超像素分割,利用Harris颜色增强角点算法检测出图像的角点,绘制凸包得到目标的大致范围,并根据中心-周围算法得到凸包中心显著图;其次利用边界信息获取图像边界超像素的背景种子,并计算一个基于背景的显著性图;通过融合以上两种显著图得到最终的显著图。在现有公开的MSRA-1000图像测试数据库中进行测试,本文方法与当前几种经典的显著性检测算法进行了对比实验,实验结果表明,本文方法得到的显著性区域更准确。Visual significance detection is a mechanism that imitates the human visual system and selects the region of interest from the complex scene.Firstly,the image is segmented into super pixels.Harris color-enhanced corner algorithm is used to detect corner points of the image,and the convex hull is drawn to get the general range of the target.Secondly,we use the boundary information to obtain the background seed of the image boundary superpixel and calculate a significance map based on the background.The final significance map was obtained by merging the above two images.In the existing public msra-1000 image test database,this method is compared with several classical significance detection algorithms,and the experimental results show that the significance region obtained by this method is more accurate.

关 键 词：计算机图象处理 显著性检测 凸包 超像素 欧氏距离 背景先验 

分 类 号：TP391.41[自动化与计算机技术—计算机应用技术]