Detecting Method of Moving Human in Intelligent Video Monitoring of Coal Preparation Plant
-
摘要: 针对数字视频监控系统人工监视方式可能因监视人员疲劳而导致失误的问题,提出了一种选煤厂运动人体目标检测方法。该方法采用背景减除法来分割运动目标,选用混合高斯模型对背景进行建模,通过不断更新背景模型来提高运动目标检测的准确性;在准确检测出运动目标的基础上,结合人体形状信息,将人体目标检测并标记出来。实验结果表明,该方法能较准确地分割出运动人体目标,并且满足实时性要求,但是无法分割遮挡或连接在一起的运动人体目标,需进一步研究。Abstract: Manual surveillance method of digital video monitoring system may lead to fatigue failure. The paper proposed a detecting method of moving human in coal preparation plant. The method uses background subtraction method to separate moving targets, uses Gaussian mixture model for background modeling, and continuously updates the background model to improve accuracy of motion detection. On the basis of accurate detection of moving targets, the method would detect and mark the human target combined with body shape information. Experimental results showed the method can accurately segment some targets of moving human, and meet with real-time requirements. But the method cannot segment some obscure or connected targets of moving human which needs further studying.
点击查看大图
计量
- 文章访问数: 53
- HTML全文浏览量: 12
- PDF下载量: 10
- 被引次数: 0