Font Size: a A A

Image Retrieval Based On The Computing Model Of Plastic Image Recognition

Posted on:2007-05-08Degree:MasterType:Thesis
Country:ChinaCandidate:C Y YangFull Text:PDF
GTID:2178360182966731Subject:Computer application technology
Abstract/Summary:PDF Full Text Request
By analyzing the domain status quo of the plastic making industry, we concluded that it's necessary to introduce industrial design to promote this domain. So, this article aims to develop the aided design tool, image retrieval system based on the computing model of plastic image recognition, at the stage of computer-aided conceptual design which is the early phase in industrial design. This system enables designers search in terms of the required image description of consumers, and the result will be the design fodder set of the image description, then designers can refer to the useful materials to design. For this purpose, the main content of our research is to build the computing model of plastic image recognition, the image retrieval system based on this model and Web information extraction which provides the database for the whole system. It's introduced as follows:In the 1st part, the aim of our research is depicted from the view of domain requirement. Furthermore, the content and status in quo of this research are analyzed and the train of thought is brought forward.In the 2nd part, the key research is on the computing model of plastic image recognition. Plastic chairs were taken as the research objects. With semantic differential for experiment employed, the similarity matrix was built between form/material feature and product image. After that, computing of composite image similarity was implemented.In the 3rd part, the key research is on the image similarity labeling to the unlabeled pictures. This technique was implemented by the pictures' feature matching based on the similarity matrix, the result acquired in the 2nd part.In the 4th part, the design and implementation of the plastic image recognition model based image retrieval system is introduced, including product information extraction and image retrieval subsystem. Then verification is presented.In the end, all of the contributions are summed up and the conclusions about the works are drawn. Moreover, it discusses some unsolved problems and future research works.
Keywords/Search Tags:Product Image Recognition, Composite Image, Content-based Image Retrieval, Image Similarity Labeling, Web Information Extraction
PDF Full Text Request
Related items