摘要
提出了一种室内静止摄像机条件下的运动目标检测和阴影抑制方法。该方法采用一种自适应的背景估计方法来实时更新背景,用基于概率分类法检测运动目标,并在联合HMMD色彩空间和光度特征来抑制阴影之后,用Sobel边缘检测来修正运动目标。实验结果表明,该方法能够有效地检测运动目标和抑制阴影。
An approach to detect moving object and suppress shadow with a static camera indoors is presented in this paper, which uses a self-adaptive background subtraction method to update background and detects moving object based on the classification of probability,associates with HMMD color space and luminosity characteristic for shadow suppression,and then Sobel＇s edge-detecting algorithm is used to amend moving object The experimental results show that the method is effective in detecting moving targets and suppressing shadow.
出处
《计算机工程与应用》
CSCD
北大核心
2007年第21期232-234,238,共4页
Computer Engineering and Applications
基金
建设部科技攻关项目(No.03-2-043)。
