摘要
为在视频监控中更加有效地检测出行人运动目标阴影,进行行人的步态识别,提出一种基于图割技术(graph cuts)的多特征运动阴影检测方法。通过颜色与纹理特征的融合,运用图割技术进行全局最优求解,对图像中行人实体部分和阴影部分进行正确分割,检测出行人阴影。实验结果表明,该方法分割适中,基本将行人的轮廓表现出来,充分体现了行人的步态特征。
To detect motion target shadow more effectively in the video surveillance so as to progress the gait recognition,a graph cuts techniques based moving shadow detection method using multi feature was proposed.Through the fusion of color and texture features,graph cuts technique was used to solve the global optimal problem of the method,thus,the entities and shaded parts of image were correctly segmented and the shadow of pedestrian was detected.Experimental results show that the proposed method is suitable for the segmentation and the contour of the pedestrian is presented,which fully reflects the characteristics of pedestrian＇s gait.
出处
《计算机工程与设计》
北大核心
2016年第8期2151-2154,2164,共5页
Computer Engineering and Design
基金
国家自然科学基金民航联合基金项目(U1433130)
民航局科技基金项目(20150215)
四川省科技厅科技基金项目(2015JY0188)
关键词
图割
阴影检测
多特征
步态
识别
graph cuts
shadow detection
multi-feature
gait
distinguish
