Research on Target Recognition of Municipal Pipeline Dredging Robot Based on Visual SLAM
With the acceleration of the urbanization process,the demand of the municipal pipeline dredging work is increasing.In order to improve the dredging efficiency and reduce the labor cost,this study is committed to develop a municipal pipeline dredging robot based on visual SLAM technology.The robot can enable the automatic identification of obstacles,silt and other foreign bodies inside pipes through deep learning and computer vision technology.Visual SLAM technology enables robots to achieve autonomous positioning and map construction through visual sensors,providing a basis for subsequent target recognition and dredging operations.This paper details the composition of robotic systems,the implementation of visual SLAM,and the algorithmic framework for target recognition.The target recognition method proposed in this paper includes the steps of image preprocessing,feature extraction,target detection and recognition,which can accurately identify the obstacles in the pipeline and realize autonomous obstacle avoidance and dredging operations.The results show that the robot has good dredging effect and target recognition accuracy,which provides a new solution for the municipal pipeline dredging work.
NETL
NSTL
万方数据
