多种识别方式组合的智能分类垃圾桶  

Intelligent Sorting Trash Bin with a Combination of Multiple Identification Methods

作　　者：马新玲[1] 郭兆阳 乐祺中 杨旭东 MA Xinling;GUO Zhaoyang;YUE Qizhong;YANG Xudong(School of Mechanical and Power Engineering,East China University of Science and Technology, Shanghai 200237, China)

机构地区：[1]华东理工大学机械与动力工程学院,上海200237

出　　处：《机械与电子》2020年第12期33-36,41,共5页Machinery & Electronics

基　　金：大学生创新创业训练计划项目经费支持(201910251062)。

摘　　要：设计了一款智能垃圾桶,采用高清摄像头、红外热成像仪、物相分析仪与数据处理设备,并利用机器视觉、热成像分析、物相分析等多种方式进行识别判断,与数据集对比得到垃圾分类结果。然后对多种识别方式的分类结果应用投票算法,利用控制矩阵控制各识别方式判断结果的可信度,从而计算比较得出最终判断结论。这种综合考虑多种识别方式的方法,能够有效提升垃圾分类判断的准确度。A smart trash bin is designed,which uses a camera,infrared thermal imager,phase analyzer and data processing equipment,and uses machine vision,thermal imaging analysis,phase analysis and other methods to identify and judge,and compare with the data set to obtain the garbage classification results.Then the voting algorithm is applied to the classification results of multiple identification methods,using the control matrix to control the credibility of the judgment results of each identification method,so as to calculate and compare to reach the final judgment conclusion.This comprehensive consideration of multiple identification methods can effectively improve the accuracy of garbage classification.

关 键 词：智能垃圾桶 垃圾分类 识别方式 投票算法 判断准确度 

分 类 号：TP391.41[自动化与计算机技术—计算机应用技术]