Design and implementation of garbage classification and detection system based on YOLOv8

Garbage classification can be seen everywhere in today's society, but due to the changes of the times, the awareness that different types of waste should go into different types of bins is not widespread among most people. Even some people don't even know what kinds of garbage there are. Therefore, it is very important to identify garbage. In order to speed up training and recognition, the author improved the YOLOv8 algorithm newly released this year, and at the same time used crawlers to crawl thousands of pictures of different types of garbage on the Internet for training. The author obtained the model with the highest accuracy by increasing the number of epochs and adjusting the batch.