Obstacle Avoidance Strategy for Mobile Robot Based on Monocular Camera

This research paper proposes a real-time obstacle avoidance strategy for mobile robots with a monocular camera. The approach uses a binary semantic segmentation FCN-VGG-16 to extract features from images captured by the monocular camera and estimate the position and distance of obstacles in the robot's environment. Segmented images are used to create the frontal view of a mobile robot. Then, the optimized path planning based on the enhanced A* algorithm with a set of weighted factors, such as collision, path, and smooth cost improves the performance of a mobile robot's path. In addition, a collision-free and smooth obstacle avoidance strategy will be devised by optimizing the cost functions. Lastly, the results of our evaluation show that the approach successfully detects and avoids static and dynamic obstacles in real time with high accuracy, efficiency, and smooth steering with low angle changes. Our approach offers a potential solution for obstacle avoidance in both global and local path planning, addressing the challenges of complex environments while minimizing the need for expensive and complicated sensor systems.