Portable target tracking method and system based on quadcopter

Abstract

The invention discloses a portable target tracking method and system based on a four-axis air vehicle. The method includes the steps that (1) the four-axis air vehicle tracks and collects a target image in real time, and the image is output after being compressed; (2) video compression data are received, decompression is performed, target detecting and tracking are conducted through a TLD algorithm according to the decompressed image data, template matching is executed independently, and when a target is detected, the detected target location is output; (3) a control parameter is obtained through calculation by using a PID algorithm according to the target location, a PID coefficient in the PID algorithm is set according to the flight state, and the movement of the four-axis air vehicle is controlled. The device comprises the four-axis air vehicle, a target detecting and tracking module and a flight control unit, wherein the four-axis air vehicle is used for tracking and collecting the target image in real time, the target detecting and tracking module is used for conducting target detecting and tracking, and the flight control unit is used for calculating the control parameter. The device has the advantages of being high in execution speed, high in tracking accuracy and capable of achieving the purpose of automatic searching and tracking.

Description

Technical field [0001] The invention relates to the technical field of target detection and tracking, in particular to a portable target tracking method and system based on a quadcopter. Background technique [0002] In the process of performing some specific tasks, such as anti-terrorism, aerial photography, traffic monitoring, disaster relief, etc., it is often necessary to be able to autonomously track specific targets and transmit relevant information about the targets back to the command center, using drones and Computer vision processing technology to realize the autonomous tracking flight of the target and the information feedback of the target can not only effectively reduce the casualties, but also reduce the loss of financial and material resources. [0003] In the prior art, due to limitations of detection and tracking algorithms and control algorithms, tracking accuracy and tracking efficiency are not high, and the control accuracy of target tracking in complex situatio...

AI Technical Summary

Problems solved by technology

[0006] However, when the TLD algorithm detects moving targets, due to the large amount of calculation and low execution speed of the detection and tracking algorithm, especially the target detection part of the algorithm is the most time-consuming part, the template matching algorithm needs a lot of calculations, and every Each pixel of a detection image needs to be calculated with each pixel of the template to obtain a matching result, which makes the TLD algorithm run at a low speed and affects the accuracy and efficiency of target detection and tracking.

Images