摘要
分析了现有人工视频监控的不足和智能视频监控的优越性,提出了检测视频中可疑情形的可行方法。为了能够在长时间不间断视频中稳定快速准确地识别出当前监控所遇到的情形,提出了基于视频帧间差分的视频监控处理技术;同时对视频处理进行了区域分割,提高对视频中可疑情况的识别率。实验在Matlab环境下对于录制的视频进行处理,利用改进的帧间差分法检测到视频中的异常情形,得到监控视频中有异常情况的画面,验证了方法的可行性和有效性。
The shortcomings of the existing artificial video monitoring and the superiority of the intelligent video monito- ring are analyzed, and the feasible method to detect the suspicious circumstances in the video is proposed. In order to identify the current situation steadily, quickly and accurately in the long time continuous video, the video monitoring technology based on inter-frame difference is proposed. At the same time, region segmentation is used to process the video, and the recognition rate of the suspicious circumstances in the video is improved. The video is processed in the MATLAB environment, and the abnormal situation is detected by the improved inter-frame difference method. The frames of abnormal condition in the video monitoring is obtained, and then the feasibility and validity of the method are verified.
出处
《四川理工学院学报（自然科学版）》
CAS
2015年第6期58-62,共5页
Journal of Sichuan University of Science & Engineering(Natural Science Edition)
基金
四川理工学院研究生创新基金项目(y2015015)
关键词
视频监控
帧间差分法
区域分割
video monitoring
inter-frame difference
region segmentation
