建筑垃圾的颜色特征提取与分类研究  

Research of Construction Waste Color Feature Extraction and Classification

作　　者：黄惠玲 韩军[1] 吴飞斌[1] 戴厚德 

机构地区：[1]中国科学院海西研究院泉州装备制造研究所,福建泉州362200

出　　处：《光学与光电技术》2018年第1期53-57,共5页Optics & Optoelectronic Technology

基　　金：福建省泉州市科技重大项目资助(2016G006)

摘　　要：建筑垃圾的处理事关环保建设,是当前建筑领域亟待解决的问题之一,而目前建筑垃圾分拣设备的主要技术瓶颈是相近比重物质的分选。结合建筑垃圾的特点,采用颜色识别方法来解决相近密度物质分选这一技术难题。通过基于HSV的阈值分割算法和K均值聚类算法对建筑垃圾图像进行实时采集和处理,对不同颜色的建筑垃圾进行识别。实验结果表明,采用的建筑垃圾颜色识别方法结果与人类视觉系统判断一致,能够辨识目标砖块并确定其位置。The disposal of construction waste is related to the construction of environmental protection.The urgent need to solve the social problems needs to be solved urgently,and the primary technical bottlenecks of current construction waste resource recycling equipment is the sorting of similar substances.Combined with the characteristics of construction waste,color recognition method is adopted to solve this technical problem.Using the threshold segmentation algorithm based on HSV color space and K-means clustering algorithm for real-time acquisition and processing of construction waste image,the recognition of different colors of construction waste can be realized.The results show that the color recognition method of construction waste color is consistent with the human visual system,and it can recognize the target objects and determine its position.

关 键 词：颜色识别 彩色图像分割 阈值分割 K均值聚类算法 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]