背景模型选择的目标检测  

Target detection based on background model selection

作　　者：王俊卜 宋英磊[1] 张冰[1] 孙陶莹 WANG Junbo;SONG Yinglei;ZHANG Bing;SUN Taoying(College of Electronics and Information,Jiangsu University of Science and Technology,Zhenjiang 212003,China)

机构地区：[1]江苏科技大学电子信息学院,江苏镇江212003

出　　处：《现代电子技术》2018年第13期34-37,共4页Modern Electronics Technique

基　　金：国家自然科学基金项目(61671222);江苏省特聘教授基金项目(1034901501)~~

摘　　要：针对视频图像中的目标检测问题,提出基于模型选择的背景模型建模方法,并在此基础上实现了目标检测。该方法能够有效地根据视频图像的复杂度对其背景进行建模。对简单视频图像选择改进的均值法背景模型,对复杂视频图像则选择码本模型。以信息论中的熵和边缘比率作为依据,判断图像的复杂度,进而选择建立相应的背景模型。实验结果表明,提出的算法模型具有很好的目标检测效果,并可以较精确地处理不同复杂度的视频图像。A background model modeling method based on model selection is proposed to realize target detection in video image. With the proposed method,the background of the video image is effectively modeled according to the complexity of the video image. The improved mean value method background model is used to process the simple video image,and the codebook model is used to process the complex video image. The complexity of the image is judged on the basis of entropy in information theory and edge ratio,so as to select the corresponding background model. The experimental results show that the algorithm model has perfect target detection effect,and can process the video image with different complexities accurately.

关 键 词：背景模型 码本模型 均值法 信息熵 视频图像 目标检测 

分 类 号：TP391.41[自动化与计算机技术—计算机应用技术]