An automatic tree trunk whitewashing robot control scheme based on the YOLOv5 visual model

In urban street trees, gardens, and other scenarios requiring trunk whitewashing, this paper addresses the challenges of manual whitewashing, such as low efficiency, uneven coating, and repetitive labor. We propose a control scheme for an automatic trunk whitewashing robot that integrates machine vision and an STM32 microcontroller, leveraging the YOLOv5 vision model. MaixCAM serves as the visual module to identify and locate tree trunks needing whitewashing. By integrating sensor ranging, we measure and calculate the tree trunk diameter, transmitting the data to the STM32 microcontroller for centralized control. Experimental results show that the relative error of the measured diameter remains within ±5%. The microcontroller dynamically adjusts the whitewashing mechanism’s diameter to adapt to different trunk sizes, ensuring efficient and uniform coating. This scheme integrates machine vision and embedded control technologies, providing valuable insights for mechanical engineering and mechatronics. It drives the technological transition of outdoor mobile robots from semi-automation to full automation.