Unmanned sweeper charging docking method based on computer vision and machine learning

The invention provides an unmanned sweeper charging docking method based on computer vision and machine learning, and the method comprises a sweeper and a charging pile fixed in a park, and the method comprises the steps: receiving RTK data, carrying out the map dotting, enabling the sweeper to carry out the map dotting around a road needing to be swept in the park, and enabling the sweeping vehicle to carry out the map dotting; establishing a road map of the park, and setting position information of the charging piles on the map; in the sweeping operation process of the sweeper, autonomous charging is carried out when the electric quantity is insufficient, firstly, data of RTK and IMU are received, a Q-learning algorithm is operated according to position information of a charging pile on a map, path planning is carried out on the current position of the sweeper and the position of the charging pile, when the sweeper moves to the position of the charging pile, pedestrians and vehicles are detected in real tim