摘要
针对传统漏水检测方法效率低、成本高、难以实时感知的问题,设计了一种基于环境声音识别的厕所漏水检测系统。该系统基于树莓派硬件,通过声音传感器进行声音采集,并调用本地声音分类模型完成漏水声音识别。之后系统将漏水识别情况上传至物联网平台进行统一监控和管理,方便工作人员通过平台查看厕所漏水情况。实验结果表明,在实际应用中该系统对厕所漏水现象的识别准确率达到93%,具有较好的泛化能力和鲁棒性,可有效用于精准及时监测厕所漏水情况的发生。该系统成本低,具有一定实用性。
Aiming at the problems of low efficiency,high cost and difficult real-time perception of traditional water leakage detection methods,We designed a toilet leakage detection system based on environmental sound recognition.The system is based on raspberry pie hardware,collects sound through sound sensor,and calls local sound classification model to complete water leakage sound recognition.After that,the system uploads the water leakage identification to the Internet of things platform for unified monitoring and management,so that the staff can view the water leakage of the toilet through the platform.The experimental results show that in practical application,the recognition accuracy of the system for toilet water leakage is 93%,which has good generalization ability and robustness,and can be effectively used to accurately and timely monitor the occurrence of toilet water leakage.The system has low cost and certain practicability.
作者
柳心敬
陈俊杰
张伟
修宇
LIU Xin-jing;CHEN Jun-jie;ZHANG Wei;XIU Yu(School of Artificial Intelligence,Anhui Engineering University,Wuhu 241000,China;School of computer and information,Anhui Engineering University,Wuhu 241000,China)
出处
《电脑与信息技术》
2023年第4期11-14,共4页
Computer and Information Technology
基金
高校国家级大学生创新创业训练计划项目(项目编号:202110363093)
安徽省大学生创新创业训练计划项目(项目编号:S202110363313)。
关键词
环境声音识别
漏水检测
树莓派
物联网
ambient sound recognition
water leakage detection
raspberry pie
internet of things
