Research on Application Technology of Target Tracking System Based on Multi-rotor

Aiming at the rapidity and accuracy of target tracking in multi-rotor low-altitude motion process, the application of moving target detection system in video sequence on multi-rotor platform is proposed. In this study, template matching and feature point detection in visual sensing were used to track image targets. A Kalman filtering, feature point detection and multi-scale Harris angle matching tracking method are applied. GPS real-time positioning of multi-rotor, high-speed long-distance WIFI transmits OPENMV3 data and multi-rotor position to ground station for monitoring. The multi-rotor airborne servo PTZ used an adaptive fuzzy PID control algorithm to automatically follow the visual sensing device target tracking center coordinate point.

I. Introduction

Multi-rotor target tracking of airborne vision sensing devices has strong application requirements in many fields, such as military, transportation, and natural disaster prevention. Therefore, An application technology research based on multi-rotor target tracking system. Multi-rotor master control uses Pixhawk open source flight control system, peripheral GPS module, SBUS receiver, power supply, driver and other modules. Openmv3 comes with a high-speed processor of the STM32F7 system, directly driving the PTZ, Image processing realized identification and tracking of specified targets, and adopted adaptive fuzzy PID algorithm to make the PTZ follow the tracking target of the camera.