期刊
WASTE MANAGEMENT
卷 32, 期 12, 页码 2229-2238出版社
PERGAMON-ELSEVIER SCIENCE LTD
DOI: 10.1016/j.wasman.2012.06.002
关键词
Solid waste monitoring and management; Bin level detection; GLAM; MLP; KNN
资金
- [LRGS/TD/2011/UKM/ICT/04/01]
An advanced image processing approach integrated with communication technologies and a camera for waste bin level detection has been presented. The proposed system is developed to address environmental concerns associated with waste bins and the variety of waste being disposed in them. A gray level aura matrix (CLAM) approach is proposed to extract the bin image texture. CLAM parameters, such as neighboring systems, are investigated to determine their optimal values. To evaluate the performance of the system, the extracted image is trained and tested using multi-layer perceptions (MLPs) and K-nearest neighbor (KNN) classifiers. The results have shown that the accuracy of bin level classification reach acceptable performance levels for class and grade classification with rates of 98.98% and 90.19% using the MLP classifier and 96.91% and 89.14% using the KNN classifier, respectively. The results demonstrated that the system performance is robust and can be applied to a variety of waste and waste bin level detection under various conditions. (c) 2012 Elsevier Ltd. All rights reserved.
作者
我是这篇论文的作者
点击您的名字以认领此论文并将其添加到您的个人资料中。
推荐
暂无数据