摘要
在充分研究基于时间信息的帧间差分法和基于空间信息的背景差分法的基础上,提出了一种综合利用视频图像序列的时间信息和空间信息的运动目标检测算法。该算法不仅有效地消除了噪声,还保证了运动图像的完整性。实验表明该算法检测准确、实时性强。
After studying some prevalent moving objects detection algorithms, two consecutive frames subtraction algorithm based on temporal information and background subtraction algorithm based on spatial information, a new algorithm based on information of spatial-temporal of video-frequency image sequence is presented in the paper. In tests, the method eliminates the noise and maintenances the integrality of the moving image. The results show that this algorithm combines the advantages of veracity and runtime, and fits for real time detection.
出处
《电子技术应用》
北大核心
2007年第8期69-71,共3页
Application of Electronic Technique
关键词
帧
运动目标检测
对称差分法
背景差分法
像素灰度归类
frame
moving objects detection
symmetry subtraction algorithm
background subtraction algorithm
pixel intensity classification
