Dynamic target identification and real-time tracking system based on FPGA

Abstract

The invention relates to a dynamic target identification and real-time tracking system based on an FPGA. The system comprises an image acquisition module used for acquiring images in real time, an image library used for updating and storing the images acquired in real time, an image processing module, an FPGA module, an analysis module and a cache module, and the image processing module is used for preprocessing the images acquired in real time to obtain processed real-time images; the FPGA module is used for performing target identification; the analysis module is used for performing target tracking and trajectory prediction; the cache module is used for caching data and providing mutual exchange between the data. Compared with the prior art, on the basis of the FPGA technology, filteringbinarization processing, cascading updating and distributed caching are adopted, and processing, target recognition, target tracking and trajectory prediction are carried out on images collected in real time, so that the recognition speed can be effectively increased, the recognition error is reduced, the recognition accuracy is improved, and the purpose of real-time target tracking can be achieved.

Description

technical field [0001] The invention relates to the technical field of target detection and tracking, in particular to an FPGA-based dynamic target recognition and real-time tracking system. Background technique [0002] Dynamic target detection and tracking technology is a research hotspot in the fields of computer vision, image processing, pattern recognition and artificial intelligence in recent years. Among them, pedestrian detection based on computer vision is regarded as the most important application value in vehicle auxiliary driving systems. It is widely used in the field of intelligent vehicles, as well as in the field of target search in public places such as stations and airports. Existing target detection and tracking usually requires the help of image recognition technology. Image recognition is generally divided into face recognition and commodity recognition. Face recognition is mainly used in security inspection, identity verification and mobile payment, and...

AI Technical Summary

Problems solved by technology

Chinese patent CN201910461306.X discloses a method and device for tracking a target object. By controlling the camera to switch to different preset positions, the searched preset position corresponds to whether there is a target in the monitoring screen. When the target is found, the The target is tracked and monitored. This patent needs to constantly change the position of the camera to track the target trajectory. In fact, it is a passive and lagging target tracking method, and it is easy to lose the tracked target.

[0004] To sum up, the existing technology usually has the problems of large recognition error, low recognition accuracy, and slow recognition speed when identifying dynamic targets. When tracking dynamic targets, it can only identify and track the trajectory, and cannot actively track it in real time. Target

Images