Method for processing structured description of video image data and capable of implementing multi-target tracking

Abstract

The invention relates to a method for processing structured description of video image data and capable of implementing a function of multi-target tracking in a computer system, belonging to the technical field of video treatment. According to the method, a target detecting module is used for detecting and matching a dynamic target of video image data, and updating the target status according to a situation whether the dynamic target is shaded, and then updating the target and a track; and describing information by an information describing module, to generate structured description information which is understandable to both a computer and a human body. According to the method for processing structured description of video image data and capable of implementing the multi-target tracking, the video search speed is effectively increased to adapt to application of the video image in the network transmission on the basis of not reducing the image quality and not increasing the code rate; besides, the method for processing structured description of video image data and capable of implementing the function of multi-target tracking in the computer system is simple to implement, low in application cost and wider in application range.

Description

technical field [0001] The present invention relates to the technical field of video processing, in particular to the technical field of multi-moving target tracking based on video images, and specifically refers to a method for realizing structured description processing of video image data for multi-target tracking. Background technique [0002] Vision sensors have been widely used to monitor scenes in the field of security protection, but most scene monitoring systems based on vision sensors are still in the semi-manual simulation monitoring stage. Among them, the weak automatic processing capability of surveillance video is the bottleneck that restricts the further development of video surveillance applications. The lack of efficient and accurate video retrieval methods limits data sharing capabilities. It is difficult for video surveillance systems to achieve resource integration and interoperability with public security information systems. Video surveillance There is ...

AI Technical Summary

Problems solved by technology

The optical flow method is suitable for scenes where the camera is still and moving, but when dealing with situations such as target occlusion, multiple light sources, and noise effects, the premise is often not established, and it is difficult to apply it to real-time processing without hardware support.

Images