摘要
通过分析视频处理流程,设计并实现了一套基于OpenCV的视频目标检索方案。首先使用经过改良的背景差分法获取视频中的运动前景,接着根据前景的轮廓得到前景的位置,然后根据前景的颜色直方图特征对前景物体进行分类跟踪,最后将分好类别的目标进行排序,以便根据要求得到最迫切需要的目标及其运动过程。实验结果表明,该方法科学合理,思路清晰,功能完备,对视频目标检索的深入研究具有很高的研究价值。
By analyzing the video processing flow,a video object retrieval scheme based on OpenCV is designed and implemented. First-ly,get foreground objects in the video by an improved background differencing method. Secondly,recognize the contour of foreground objects to get moving objects＆#39; position. Then,track different categories of objects separately after classifying objects by comparing their color histogram features. Finally,put the classified objects in order to make sure to get the most needed object as well as its motion. Ex-perimental results show that the method is scientific and reasonable,clearly-formed and fully functional,which has a high value for deep study of video object detection.
出处
《计算机技术与发展》
2014年第11期210-213,共4页
Computer Technology and Development
基金
安徽省科技攻关计划科技强警专项资金资助项目(1301b042020)
安徽大学青年骨干教师培养对象经费资助项目
高等学校博士学科点专项科研基金联合资助课题(20133401110009)
关键词
目标检索
背景差分
分类
跟踪
object retrieval
background difference
classifying
tracking
