FPGA-based real-time target tracking system and method

Abstract

The invention discloses an FPGA-based real-time target tracking system comprising an average filter, a motion compensator, a dynamic block detection module, a suggestion box search module and a targetmatching module. The target tracking method comprises the following steps: the average filter divides video data into multiple image blocks and calculates an average value of the image blocks, the motion compensator and the dynamic block detection module process the image blocks to obtain high probability image blocks, and the target matching module performs matching calculation on the high probability image blocks optimized by the suggested frame search module to obtain a final tracking target. The FPGA-based real-time target tracking system is optimized on the architecture of the system, and multiple parallel calculations are used in the whole calculation process to improve the processing speed and achieve the purpose of real-time tracking.

Description

technical field [0001] The invention relates to the field of vision applications, in particular to an FPGA-based real-time target tracking system and method. Background technique [0002] Real-time target tracking system is widely used in human-computer interaction, security monitoring, augmented reality and other fields. In order to obtain good tracking performance and strong robustness to complex external environments, the tracking system needs complex processing and huge Computing costs, and some specific platforms have high requirements for system power consumption, making it difficult to achieve real-time tracking. [0003] Although the existing target tracking schemes have jointly optimized the tracking algorithm and the specific architecture, the tracking performance has not yet reached real-time due to the slow processing speed. Therefore, a system with high performance and high energy efficiency is urgently needed to realize Real-time target tracking and high energ...

AI Technical Summary

Problems solved by technology

[0002] Real-time target tracking system is widely used in human-computer interaction, security monitoring, augmented reality and other fields. In order to obtain good tracking performance and strong robustness to complex external environments, the tracking system needs complex processing and huge Computing costs, and some specific platforms have high requirements for system power consumption, making it difficult to achieve real-time tracking

[0003] Although the existing target tracking schemes have jointly optimized the tracking algorithm and the specific architecture, the tracking performance has not yet reached real-time due to the slow processing speed. Therefore, a system with high performance and high energy efficiency is urgently needed to realize Real-time object tracking and high energy efficiency

Images