基于机器视觉的垃圾分类算法研究与应用
    点此下载全文
引用本文:王光清,李文拴,党佳琦,张愉.基于机器视觉的垃圾分类算法研究与应用[J].计算技术与自动化,2024,(1):78-83
摘要点击次数: 88
全文下载次数: 0
作者单位
王光清,李文拴,党佳琦,张愉 (延安大学 化学与化工学院陕西 延安 716000) 
中文摘要:垃圾分类识别算法是目前研究的热点问题,本文通过引入色块追踪模块Lab颜色模型对YOLOv3算法进行优化,利用优化后的算法搭建训练模型。并针对目前垃圾类别利用网络爬虫爬取日常生活中常见的垃圾图像并进行分类,形成数据集。其次通过优化的YOLOv3算法对处理好的数据集进行模型训练,将训练后的模型进行模型检测。最后通过实际测试,优化后的YOLOv3算法识别的平均准确率达到了94.33%,与原始算法相比,优化后的算法在稳定性和准确度上都有了明显的改善。
中文关键词:垃圾分类  色块追踪模块  模型训练  YOLOv3算法优化
 
Research and Application of Garbage Classification Algorithm Based on Machine Vision
Abstract:The garbage classification and recognition algorithm is a hot topic in the current research. This paper optimizes the YOLOv3 algorithm by introducing the color block tracking module Lab color model, and uses the optimized algorithm to build a training model According to the current garbage category, we use web crawlers to crawl the common garbage images in daily life and classify them to form a data set Secondly, the optimized YOLOv3 algorithm is used to train the model of the processed data set, and the trained model is checked Finally, through practical testing, the average recognition accuracy of the optimized YOLOv3 algorithm reaches 94.33%. Compared with the original algorithm, the stability and accuracy of the optimized algorithm have been significantly improved.
keywords:refuse classification  color block tracking module  model training  YOLOv3 algorithm optimization
查看全文   查看/发表评论   下载pdf阅读器