Target tracking and platform control system and method based on monocular vision mobile robot

Abstract

The invention discloses a target tracking and platform control system and method based on monocular vision mobile robot,and relates to the technical field of robots. The method comprises the steps that firstly,according to an estimated state of a target and a robot at the moment k,control components are generated,wherein the control components include an angular speed delta phi k and a linear speed delta x k; secondly,a monocular vision SLAM method is used for obtaining the state X<r>k and covariance P<r>k for estimation; the target predicts the self state X<t->k and a covariance matrix P<t->kaccording to a motion model f<t>,then,X<r>k and X<t->k are used for being combined with a target observing model for observing a current target observing value Z<t->k,and an observation residual matrix S<t->k is generated; finally,a target actual observation value Z<t>k is used for upgrading the target state and the variance matrix to obtain X<t>k and P<t>k. An image photographed by a video camera carried on the mobile robot is used for completing real-time online estimation on the mobile target state,the robot state and the environment characteristic state,it is ensured that the robot generates an enough observation parallax error for the target,target state convergence under the target pure direction observed quantity condition is ensured,and the application prospect is wide.

Description

technical field [0001] The invention relates to the technical field of robots, in particular to a monocular vision-based mobile robot target tracking and platform control method in an unknown environment. Background technique [0002] Mobile robot applications first need to solve the problem of robot cognition. Simultaneous robot positioning, map construction and target tracking (Simultaneous Localization, Mapping and Object Tracking, SLAMOT) can realize the simultaneous estimation of the robot's own state, environmental state and target state. The real-time cognitive ability has a wide range of application prospects. Most of the existing research on SLAMOT is based on active sensors or stereo vision sensors, and these sensors have defects such as large size, heavy weight, high energy consumption, high equipment cost, and limited observation distance, which limit the practical application of SLAMOT. scope. As a passive sensor, the monocular vision sensor has played an imp...

AI Technical Summary

Problems solved by technology

Finally, the existing mobile platform moving target tracking methods in unknown environments generally use multi-sensor information fusion to achieve independent estimation of the mobile platform and target states. The use of multiple sensors limits the application range of research results and reduces the difficulty of the problem. However, the independent estimation method does not consider the coupling of different object states, thus affecting the accuracy of system state estimation.

Images