The Research And Application Of Intelligent Video Analysis Technology In The Banking Business

In recent years, with the development of the computer science, image processing, artificial intelligence and pattern analysis, research of intelligent video surveillance becomes a hot point. Using computer vision and video analysis method to analyze the image sequences recorded from the camera automatically attract people’s attention and many progresses have been achieved since large numbers of researchers devoted themselves in the area. This dissertation is mainly about intelligent video analysis technology and the application in the banking business. The main content of this dissertation can be summarized as follows:1. It is a large waste of storage device for storing big number of useless video in the bank surveillance system. This dissertation proposed to preprocess by cutting the unvalued video segment. Using background differing based on the Gaussian Mixture Model to detect moving object, and then cut the video segment without moving object. This work is useful for saving storage space and retrieving.2. Some customers would forget their luggage or expensive thing when perform transaction at the bank. This dissertation an abandoned/stolen object detection algorithm based on dual background model. The first step is detecting static objects foreground by two backgrounds. One background is updated with mask and the other one without mask. The second step is tracking the static objects after object segment from foreground. The last step is classifying the static object as abandoned or stolen object and triggering an alarm signal. In addition, the algorithm is simple and computationally less intensive as it avoids the use of expensive filters while achieving better detection results.3. Objects Counting is important in video surveillance for bank. A method of people counting based on CamShift has been proposed in this dissertation. Auto tracking based on CamShift has been achieved. The method is applicable to situation of small communities.4. A video surveillance and analysis system according the above research content is achieved. The functions of the system is including video playing, moving object tracking, intelligent storage control, abandoned/stolen object detection, regional object counting. Experiments show the system is reliable and robust.