Design And Implementation Of Shenyang’s Police Video Monitoring System

With the rapid development of information technology, especially into the 21st century, facing with the video processing and video monitoring system related technologies such as artificial intelligence, machine learning, combining challenge of realizing intelligent video monitoring system. Though video surveillance system is widely used in China, but the intellectualized degree is far behind developed countries developed a new type of intelligent video surveillance system. Aimed at this situation this paper tries to design and implement an intelligent video monitoring system with its own intellectual property rights.In this paper, according to the system design requirements and handle relevant advanced technology nowadays video, using a particle filter framework, from the theoretical research and engineering development process system design for the repeated arguments, respectively, the work includes the following several parts:1. This paper proposes a can simultaneously suitable for target tracking problem and the problem of video image theory of particle filter framework. Through the single weight in the process of computing the constraint conditions to ensure the importance sampling procedure will not appear on a large scale discrete degradation, so as to ensure the target tracking and the calculation efficiency in the process of video image.2. The application of the theory of Bayesian framework, the paper puts forward a realization method of target tracking problem in the practical application environment. And give the process of particle filter tracking timeliness theoretical criterion would not be affected. By importance sampling and sample particle count limit on a threshold, the rationality of the argument made in this paper, the use of particle filter method has a good real-time performance, meet the demand of the system to track the implementation of the time limit of demand.3. This paper reference the state-of-the-art video image method, study the process of its stability as particle representation, a particle will adopt process applied to the video image method at the same time, discussed the effect of particle number and image, the relationship between the computational times consuming.4. According to this algorithm thought with appropriate technology developed by the video image and the target tracking system, adopted the QT+OpenCV+SQL Server development tools combination, and calculation efficiency of the development system of the whole test evaluation. The algorithm effect and experience prove that we are developing system has a good robustness and practicality.