Developing an Autonomous Robotics System Utilizing Camera and IMU Fusion with PID-Based Path Correction

摘要

This paper proposes a robot equipped with an autonomous navigation system that integrates visual data from a camera and dynamic data from an Inertial Measurement Unit (IMU) to ensure precise lane following. The system consists of the following components: 1) A Logitech web camera that continuously captures and provides visual feedback about the surrounding environment, 2) An IMU that provides real- time information on the robot's position, orientation, and movement, 3) A Jetson NX Platform serving as the central processing unit to process the fused data in real-time, interpreting the camera feed to extract relevant features such as lane lines and processing the IMU data to determine the robot's current orientation, acceleration, and velocity, 4) and Proportional-Integral-Derivative (PID) Controller which receives the integrated data and continuously evaluates the robot's current position and movement against the desired path. Based on these evaluations, the PID controller dynamically adjusts the robot's steering angle to ensure it remains within the lane, correcting its path in realtime. This integration of IMU and camera data with PID control results in a robust and adaptive autonomous navigation system capable of reliable lane detection and navigation, providing the robot with effective autonomous driving capabilities. The system adapts to environmental changes, such as varying lighting conditions and road textures, by continuously updating its parameters to maintain optimal performance. Furthermore, the real-time feedback loop enables the robot to iteratively improve its navigation capabilities, ensuring stable and accurate lane following.