OPTICAL IDENTIFICATION OF CONSTRUCTION AND DEMOLITION WASTE BY USING IMAGE PROCESSING AND MACHINE LEARNING METHODS

摘要

This paper discusses the possibility of the optical identification of recycled aggregates of construction and demolition waste (CDW) as basis of an innovative sorting method on the field of processing of CDW. The first target was to find suitable identification attributes for the differentiation of aggregates, which are difficult to separate. For the investigations images of the given aggregate classes were captured and analysed by algorithms of image processing and machine learning. The interdependencies between dataset character, feature vector, type of the selected classifier and parameter settings of classifier are very complex and they were analyzed in this paper.