A study of real-time path planning for mobile robots based on environment recognition using omnidirectional images

Abstract

The successful combination of robust environment detection and appropriate path planning is a key challenge in developing intelligent mobile robots. The main problem is the time-consuming process of environment recognition using a sequence of images. We introduce a new path planning algorithm that reduces the processing time and achieves robust and stable navigation, and a PSoC microcontroller for stable robot control. The proposed smart image recognition algorithm for mobile robots provides intelligent environment detection and path planning using omnidirectional images and a heuristic search. Based on the detected environment, an appropriate path is generated by applying the proposed heuristic search algorithm. The validity of the proposed algorithm is verified using an actual mobile robot in a real-world environment.