Design and Implementation of a Lane Detection-Based Steering Control System for an Autonomous Mobile Robot

The development of autonomous mobile robots (AMRs) has advanced rapidly, driven by increasing demands for intelligent and reliable navigation systems. This paper presents the design and implementation of a lane detection-based steering control system to enable an AMR to follow lanes autonomously. The system employs RoboRealm software for real-time image processing, allowing lane markers to be detected from camera input and translated into steering commands for the robot car. A modular vision-processing architecture, integrating a camera, Mini PC, microcontroller, servo motor, and DC motors, is utilized to simplify development while maintaining sufficient accuracy and responsiveness for controlled environments. Lane detection is achieved through image pre-processing techniques such as color thresholding, blob filtering, and Hough Transform, with steering adjustments dynamically based on the computed Center of Gravity (COG) between lane boundaries. Experimental results demonstrate that the system successfully maintains lane adherence during both straight and turning maneuvers with minimal deviation, validating the effectiveness of the RoboRealm-based approach for small-scale autonomous navigation. Future work will focus on enhancing system robustness under complex lighting conditions and dynamic environments through adaptive or learning-based vision techniques.