Based On YOLOV8 Intelligent Trash Can Garbage Classification Detection Algorithm
Songzhe Pan,Ning Wang,Yuyun Lin,Jianhao Tang
DOI: https://doi.org/10.54097/67mska34
2024-05-08
Abstract:Since 2019, China has begun to vigorously implement a garbage sorting system across the country, with Shanghai being the first city to fully implement garbage sorting. For example, residents do not have a strong sense of classification and lack of classification knowledge, resulting in low classification accuracy. The management of the wharf is relatively loose, and the garbage collection, transportation and treatment links sometimes fail to strictly distinguish between various types of garbage, which affects the classification effect. Waste collection facilities are rudimentary, and sorting labels may not be uniform and oversimplified, which is not conducive to fine sorting. Therefore, it is urgent to solve the problem of garbage classification and become an urgent problem for all countries to study. The problem of sorting valuable garbage after garbage collection. The rapid development of deep learning provides the technical demand for intelligence. In this paper, the YOLOV8 algorithm is used to train the garbage classification model, and about 10,000 custom garbage datasets with 4 categories are used to classify food waste, hazardous waste, recyclable waste and other waste. In this paper, methods such as 0s, pandas, and numpy are used in Python programming. For model deployment, models can be deployed using trained weight files and configuration files, and models can be integrated into tools such as OpenCV, TensorFlow, etc., for real-time detection or integration into other applications.