Intelligent classification system based on target detection algorithm and laser-SLAM algorithm

Currently, garbage classification and recycling have been a hot spot in society. The method of realizing garbage classification through manual manner is not only inefficient, high cost, but also in a harsh working environment. Therefore, through the research and application of technologies related to machine vision and autonomous driving, this paper designs an intelligent classification system based on YOLOv5 object detection algorithm and Gmapping laser-SLAM (Simultaneous Localization and Mapping) algorithm to realize full-automation of garbage classification. The system vision module adopts YOLOv5 algorithm and trains the model on self-made garbage classification data set. U sing the trained model, the algorithm can extract the features of different types of garbage images to realize the purpose of garbage detection and classification. The movement control module adopts Gmapping algorithm based on LIDAR (Light Laser Detection and Ranging) to realize the synchronous positioning and map construction in the working environment, thereafter to complete the transportation of garbage. The experimental results show that the classification system designed in this paper can accurately transport the garbage to the destination with good intelligence.