Font Size: a A A

Image Analysis Method Of Slub Yarn's Size

Posted on:2010-07-11Degree:MasterType:Thesis
Country:ChinaCandidate:Y F WuFull Text:PDF
GTID:2121360278974974Subject:Textile Engineering
Abstract/Summary:PDF Full Text Request
Fineness alternation apparent morphology of slub yarn can form special sight effect in fabric and give people boorish and fluent fashion sense, it was liked by consumer, now have become a big kind of garment material products. At the same time, slub yarn sizes's recognition need to research in textile industry as an important project.Acquisition system of yarn diameters primary included Yarn Eveness Tester, Scanner, PC machine with MatLab software in this paper. When slub yarn's diameters were acquired, the slub yarn sample which was spun was wound uniformly on the blackboard base on using yarn eveness tester, yarn blackboards were scanned with two sides though scanner, then obtained two grayscale and binary images; MatLab 7.0 software read binary images, through image rotation, yarn diameters acquisition, image connection to compose a continuous slub yarn image using computer digital image processing technology, thus slub yarn diameters signals's acquisition was finished.For regular cycle of slub yarn, using two different methods determined one cycle of slub yarn's sizes. One method was that using auto-correlation analysis to slub yarn diameters variance signals, we could obtained auto-correlation function peak value which correspond yarn diameters signal various period and bone equal to one period of yarn's length, Based on several complete periods slub yarn diameters variance's frequency statistics and determined basics's fineness and slubs's kinds and fineness, we could obtained slub yarn's cycle rule through statistics to several periods lengths and fineness of slubs and basics. Result show that auto-correlation analysis can obtain accurately slub yarn morphology parameter variance's period and determine one period's morphology parameters, moreover the process of slub yarn is produced, there is existed transition regions between slub and basic parts, thus slub yarn theory and practice parameters have big differences, comparison theory value to nominal, basic parts have less lengths and fineness, slub parts have larger lengths and fineness which is compared by calculation according to slub amplitude's definition. Another method was that based on parameters's statistics, slub yarn segments were carried out by cluster analysis and obtained one period of regular slub yarn parameter's concrete value. Research show that, we can obtained slub yarn's cycle rule through method of cluster analysis to slub yarn composition component.Regard to regular cycle slub yarn with having different slub amplitudes, slub amplitude were defined as the ratio of slub and basic fineness, Because of the result of yarn twist shift, slub yarn which produced in practice, the practice value is usually more bigger than the theory. Based on several period diameters of slub and basic parts, then calculated slub practice amplitudes base on slub amplitude's definition and used regress analysis to set up the relationship between practice and theory amplitudes. The result show that slub practice amplitude and theory amplitude was correlated highly.As fuzzy cycle slub yarn in two different modes in this paper and calculating each segment concrete yarn's sizes which contained slub lengths and basic lengths. Based on fuzzy mode slub yarn's spinning principle, slub yarn in proporopm fuzzy mode that based on kinds of slub lengths and basic lengths parameters formed segments group according to a certain proportion, each time selected randomly a kind of segments become slub yarn. Slub yarn in random fuzzy mode that slub lengths and basic lengths parameters equiprobable appeared in a certain range. Calculation slub yarn's sizes in two different fuzzy mode and the same slub and basic segments's proportion in proportion fuzzy mode.
Keywords/Search Tags:slub yarn, binary image, proportion fuzzy, random fuzzy, auto-correlation analysis, cluster analysis, regress analysis
PDF Full Text Request
Related items