面向视频监控的视频质量异常检测系统的设计与开发

发布时间：2018-04-27 22:54

本文选题：视频监控 + 视频质量异常检测　；参考：《中山大学》2014年硕士论文

【摘要】：近年来，随着安防技术的不断发展以及市场需求的不断扩大，视频监控系统已成为当代社会安全防范中不可或缺的一部分，视频监控以其直观、准确、及时和信息内容丰富而广泛应用于许多场合。目前我国各行业各种规模大小的视频监控系统已经非常普遍，例如在银行，社区，工厂，学校以及大型园区等场所均已安装了视频监控系统。而随着监控系统的覆盖范围的扩大，系统中所容纳的摄像头的数量在逐渐增多，监控的时间也在不断延长，给视频监控系统维护工作带来了新的挑战。视频质量检测便是其中一个核心问题。在视频监控系统中，，摄像头在外界环境以及人为恶意破坏等影响下，传输的画面质量时常会出现问题，例如：视频信号中断，视频画面偏色或者亮度对比度异常等等。这些视频质量问题严重影响了监控系统的效能，如果问题不能够及时发现和解决，很可能被有心之人或不法分子利用，造成进一步的经济损失。为了快速发现视频质量异常的摄像头并进行处理，防止因摄像头视频异常产生的损失。本文设计并实现了针对大型园区等场所视频监控系统的智能视频质量诊断模块。这个智能视频质量诊断模块包含视频偏色检测，信号中断检测，强横纹干扰检测，画面冻结检测，亮度异常检测，清晰度异常检测等。通过在安防平台上使用智能视频质量诊断模块，可以实现以下几个功能：1、当有摄像头传输画面异常时，及时产生告警，缩短产生异常到发现异常的时间间隔，提高安防效率。2、自动检测视频信号是否中断或画面冻结，及时告知安保人员进行处理，防止因人为破坏而产生进一步损失。3、自动检测视频亮度，对比度，清晰度异常，及时告知相关人员前往处理，尽早修复摄像头，恢复正常的视频监控质量。针对大型园区等场所视频监控系统的智能视频质量检测模块已经应用于中国移动南方基地，经实际运行表明，本文设计的智能视频质量诊断模块能够达到实时的计算速度，且有相对较高的告警准确度。
[Abstract]:In recent years, with the continuous development of security technology and the continuous expansion of market demand, video surveillance system has become an indispensable part of modern social security prevention, video surveillance with its intuitive, accurate, Timely and information content is rich and widely used in many occasions. At present, video surveillance systems of various sizes in various industries in China are very common. For example, video surveillance systems have been installed in banks, communities, factories, schools and large parks. With the expansion of the coverage of the surveillance system, the number of cameras in the system is gradually increasing, and the monitoring time is also being extended, which brings new challenges to the maintenance of the video surveillance system. Video quality detection is one of the core problems. In the video surveillance system, under the influence of the external environment and human malice, the quality of the video transmitted will often appear problems, such as: video signal interruption, video image color deviation or brightness contrast anomalies and so on. These video quality problems have seriously affected the effectiveness of the monitoring system. If the problem can not be detected and solved in time, it may be used by interested people or illegal elements, resulting in further economic losses. In order to quickly detect and process the camera with abnormal video quality, the loss caused by the abnormal video is prevented. This paper designs and implements an intelligent video quality diagnosis module for large-scale video surveillance system. This intelligent video quality diagnosis module includes video color deviation detection, signal interrupt detection, strong cross line interference detection, picture freezing detection, brightness anomaly detection, clarity anomaly detection and so on. By using the intelligent video quality diagnosis module on the security platform, the following several functions can be realized: 1. When there is an abnormal picture transmitted by a camera, the alarm is generated in time, and the time between the abnormal occurrence and the discovery of the anomaly is shortened. Improve the security efficiency .2. automatically detect whether the video signal is interrupted or the picture is frozen, inform the security personnel in time to deal with it, prevent further loss of .3s caused by artificial destruction, automatically detect video brightness, contrast, and clarity abnormality, Timely inform relevant personnel to deal with, repair the camera as soon as possible, restore normal video monitoring quality. The intelligent video quality detection module has been applied to the south base of China Mobile. The actual operation shows that the intelligent video quality diagnosis module designed in this paper can achieve real-time computing speed. And has relatively high alarm accuracy.
【学位授予单位】：中山大学
【学位级别】：硕士
【学位授予年份】：2014
【分类号】：TN948.6

【参考文献】