Research on Navigation Line Extraction of Garden Mobile Robot Based on Edge Detection
Chen Jiqing,Wang Zhikui,Long Teng,Wu Jiahua,Cai Ganwei,Zhang Hongdu
DOI: https://doi.org/10.1007/s10846-022-01648-7
2022-01-01
Journal of Intelligent & Robotic Systems
Abstract:The autonomous navigation of garden mobile robot plays an important role in garden monitoring and maintenance, in which the extraction of navigation line is the key technology. Due to the complexity of garden environment, the current algorithms of extracting navigation lines are easily affected by interference, relying on image preprocessing. In order to solve these problems, this paper proposed a cascade anti-interference algorithm based on edge detection. Firstly, the improved vegetation index combined with Otsu method is used to segment the road and background. Then, an edge detection algorithm is proposed to overcome the noise interference and get the road edge points. Lastly, the cascade anti-interference algorithm eliminates the interference points and extracts navigation lines. Experimental results show that the average processing time of the proposed algorithm is 36.25 ms, and the error is less than 0.8°, which is better than the traditional least square method and Hough transform method.