Research On Cooperation Strategy Of Heterogeneous Robot System Tracking For Urban Scenes

In such scenarios as traffic monitoring,agricultural production and logistics distri-bution,robots play an irreplaceable role in the efficiency and reliability of tasks.Facing the increasingly complex task demands and increasing tasks,it will become the main-stream to use robots to perform some repetitive and dangerous tasks,and robots play an increasingly important roleAs a sensor platform,UAV and UGV have strong complementarity in perception,load,speed and other aspects:UAV provides a wide field of vision and fast moving speed,and UGV provides high-definition perception and reliable operation.Facing the complex urban scene task,only a single kind of robot can not complete the complex task.Heterogeneous robot cooperation system can play the complementary advantages of UAV and UGV.This research aims to propose a novel heterogeneous robot system composed of UAV and UGV,which can achieve the tracking task more effectively and reliably.The main research includes:(1)Based on the cooperation between UAV and UGV,a heterogeneous robot coop-eration system is proposed.The system is composed of three modules:a UAV,a UGV,and a ground station console.It can track target objects well from UAV perspective,send target localization to UGV,and expand the sensing range of UGV.(2)Because UAV has the characteristics of flexible motion,good mobility and high field of vision,it is very suitable for the target tracking and localizing task.This study proposes a method of tracking and locating targets by UAV,the method uses TLD tracking algorithm to track the target in the video stream,obtains the pixel coordinates of the target,and then estimates the target localization by camera imaging principle and UAV flight parameters,which provides a broader tracking range for the UGV.(3)In this study,the untraversable areas are divided into two categories:the obsta-cle type untraversable area and the cliff type untraversable area.This study proposes an effective and robust data fusion algorithm to solve the problem of detecting the cliff type untraversable area in the process of navigation and obstacle avoidance,and an edge point filter is proposed.Finally,the fused data is used for navigation and obstacle avoidance.This study proposes a heterogeneous robot cooperative tracking system,which uses UAV to track target objects,and uses fusion data for UGV to navigate and obstacle avoid,which can better accomplish the cooperative tracking task,and provides a new cooperative way for heterogeneous robot system.