The key technologies of intelligent surveillance system camera include moving target extraction, moving target tracking, moving target classification and moving target behavior analysis. Today, Xiaobian will share with you a brief introduction to these technologies.

1、 Moving target extraction

Moving target extraction is the preparatory work for intelligent analysis. Based on this work, the security monitoring system camera can extract the changed area from the background area from the image sequence. The effective extraction of moving target will greatly reduce the amount of calculation in the subsequent process, which is of great significance for target recognition and behavior analysis in the later stage. At present, the more mainstream methods are background subtraction Time difference method and optical flow method.

2、 Moving target tracking

Moving target tracking is the process of finding the position of candidate target area similar to the target template in the image sequence through the effective expression of the target. In short, it is to locate the target in the sequence image.

3、 Moving target classification

Moving target classification, as the name suggests, extracts specific types of objects from the detected moving area, such as classifying different targets such as people, motor vehicles and people in the scene. At present, the mainstream methods include classification based on motion characteristics and classification based on shape information.

4、 Moving target behavior analysis
Behavior analysis is one of the key objectives of intelligent security surveillance camera, and it is also the key and difficult problem of video surveillance system in maintaining public security. Behavior analysis involves many fields, such as computer vision, pattern recognition, artificial intelligence and so on. On the basis of low-level processing of video image sequence, it obtains the information of the monitoring scene or the information of moving targets in the scene by analyzing and processing the images and videos of the monitoring scene, and further studies the nature and relationship of each target in the image, so as to obtain the interpretation of the objective scene and high-level semantic description, Behavior analysis is often carried out with the help of neural networks and decision-making.