%0 Journal Article
%T 基于YOLOv8的可回收垃圾识别方法研究<br>Research on Recyclable Waste Identification Method Based on YOLOv8
%A 张传庆
%A 陈杨山
%A 刘西钉
%A 王建强
%A 赵曙光
%J Computer Science and Application
%P 1019-1025
%@ 2161-881X
%D 2023
%I Hans Publishing
%R 10.12677/CSA.2023.135099
%X <span style=\"font-size:14px;\">随着经济和社会的快速发展，如何更好地分类、清运和回收垃圾已变得日益重要和广受重视。为了让垃圾的分选更加智能化，减少对人工的依赖，可以采用基于深度学习的目标识别算法，对垃圾图像进行有效地识别和检测。本文分析了于2023年1月新提出的YOLOv8 (You Only Look Once)算法的改进之处，并将YOLOv8算法应用于可回收垃圾的识别。实验结果表明，与基于YOLOv5算法的识别结果相比，基于YOLOv8算法的可回收垃圾的识别精度显著提高，mAP (平均精度值)达到了96%，已能满足分拣的需要。</span><br />
<span style=\"font-size:14px;\">With the rapid development of the economy and society, how to better sort, remove and recycle garbage has become increasingly important and widely valued. In order to make the sorting of garbage more intelligent and reduce the dependence on labor, a target recognition algorithm based on deep learning can be used to effectively identify and detect garbage images. This paper analyzes the improvements of the newly proposed YOLOv8 (You Only Look Once) algorithm in January 2023, and applies the YOLOv8 algorithm to the identification of recyclable waste. The experimental results show that compared with the recognition results based on YOLOv5 algorithm, the recognition accuracy of recyclable garbage based on YOLOv8 algorithm is significantly improved, and the mAP (average accuracy value) reaches 96%, which can meet the needs of sorting.</span>
%K 垃圾识别，目标检测，YOLOv8，Ultralytics<br>Garbage Recognition
%K Object Detection
%K YOLOv8
%K Ultralytics
%U http://www.hanspub.org/journal/PaperInformation.aspx?PaperID=65924