Mushroom Detection and Positioning Method Based on Neural Network

In order to realize the mechanized mushroom sorting process, a detection method based on the YOLO algorithm to detect the mushrooms on the conveyor belt in real time is proposed, and then the Kalman filter is used to track the position of the mushrooms at each moment, and the mushrooms are obtained while detecting and tracking the mushrooms. The pixel coordinates of the center can be converted to obtain the world coordinates of the mushrooms according to the projection imaging principle of the monocular camera. Finally, the long-short-term memory (LSTM) network is used to predict the position of the mushrooms that may be outside the camera view. Experimental results show that this system has the advantages of good real-time performance, high positioning accuracy, and good prediction results, and can realize real-time detection and positioning of mushrooms.