This product is a kind of intelligent classification device for domestic waste under visual guidance. The device first collects the picture of the garbage to be identified, uses the DNN deep neural network to train the collected picture, obtains the trained model and burns the model and program into the K210 chip. The camera collects the surface image of the conveyor belt in real time, and processes and analyzes the image through the vision system to identify the type of domestic waste on the conveyor belt. The classification identification signal simultaneously sends a signal to the motor and the steering gear (hereinafter abbreviated as steering gear) that controls the rotation of the cylindrical trash can. Then the motor is started and the steering gear quickly rotates the trash can of the correct category to the nearest position from the conveyor belt, and the garbage is poured into the designated trash can. The whole process adopts online automatic detection, advanced perception technology and deep learning algorithm to intelligently identify the four types of domestic waste and classify them, which greatly improves the efficiency of waste classification, reduces the labor cost in the later stage, and improves the correct rate of waste classification, thus creating an environmentally friendly, healthy, intelligent and fashionable lifestyle for the society. Innovations: 1. Innovatively apply visual technology to the field of domestic waste classification, and use image recognition and processing technology to realize the recognition and classification of domestic waste images in specific detection areas, which is conducive to improving detection accuracy and reducing detection errors. 2. Research and development of a new type of classification action, compact structure, smooth action, improve the efficiency of classification. 3. A mobile phone APP has been developed. The garbage information in the garbage can users' homes will be stored on the mobile phone. The garbage generation of garbage can users will be compared and analyzed, and the statistical curve will be given.