Font Size: a A A

Research On Classifying Methods And Development Of System Based On Leather Color And Texture

Posted on:2007-08-07Degree:MasterType:Thesis
Country:ChinaCandidate:C G SongFull Text:PDF
GTID:2121360185987877Subject:Mechanical Manufacturing and Automation
Abstract/Summary:PDF Full Text Request
With the spread application of computerized vision technology in various fields, these years have witnessed great development of leather industry in this aspect. To improve productivity of leather product, and to achieve automatic and intelligent production, it is important to classify leather products images with computer. The paper focused mainly on classification methods of leather colors and texture. Moreover, the author developed computerized inspection and classification system for application. The contents of the paper include:1. Analysis of leather product color and texture. This research generalized characteristic of color and texture on leather goods, based on the features of leather surface, which provided basis for the next study.2. To classify leather colors and texture quickly and steadily, the paper put forward the classification method combing main color clustering with color histogram. Meanwhile, the paper made some improvement in the methods of color histogram and used accumulative histogram to eliminate the digitized errors of color, which was verified by experiment.3. According to characters of leather texture, the paper applied a method of scanning both vertically and horizontally to derive eigenvalue of texture to define coarseness and profundity for measuring leather texture.4. Based on the identification technology of leather color and texture, the paper developed images inspection and classification system. It mainly included the design of hardware equipment and software system. In the rest of the section, elaborated design and principle of each module were given.5. The paper introduced practical application of leather inspection and classification system, and the effect was represented as well.
Keywords/Search Tags:leather goods, image processing, color recognition, texture recognition, classification system, classification and recognition, computer vision
PDF Full Text Request
Related items