Image Retrieval Technology Based on Imagery Cognition Model

TAN Zheng-yu,SUN Shou-qian
DOI: https://doi.org/10.3785/j.issn.1008-973x.2008.05.008
2008-01-01
Abstract:An imagery cognition model of plastic products was advanced to satisfy the industrial designers' demands for image retrieval in product design,based on which an E-image board system was designed.The shape and material imagery models of plastic chairs were built by semantic differential method.The distance distribution histogram(DDH) and Hu fixedness matrix were used to extract and express the products' image shape characteristics.And the texture characteristics were obtained and described by texture chart.Comparability measurement and imagery correlation label techniques were combined to expand and label the system database.This technique satisfies the designers' requirements to retrieve product image by imagery words in consistence with users' imagery cognition model, and it conquers the shortcoming in the traditional retrieval model which cannot describe users' imagery.
What problem does this paper attempt to address?