Combining Binocular Camera and Infrared Beacon: Enhancing Trajectory Planning and Motion Error Elimination for Floor Grinding Robots

To enable the floor grinding robot to operate continuously for extended periods, this study aims to design an intelligent pose error correction strategy based on a stereo camera and infrared beacons. An active infrared (IR) beacon is proposed as the tracking target, and a stereo camera equipped with a 360-degree rotating platform is designed to acquire the global position and orientation data of the floor polishing robot. To dynamically correct local positioning information, a new algorithm is introduced that periodically integrates global pose data with the cumulative measurements from the inertial measurement unit (IMU) and odometer, effectively eliminating cumulative errors. Field tests have demonstrated that this method successfully controls the robot to follow a Zig-Zag working path during long-duration operations, maintaining operational consistency and accuracy without deviation caused by sensor errors.