基于計算機(jī)圖像處理技術(shù)的紗線質(zhì)量檢測的研究.pdf

1、天津工業(yè)大學(xué)碩士學(xué)位論文基于計算機(jī)圖像處理技術(shù)的紗線質(zhì)量檢測的研究姓名：劉麗杰申請學(xué)位級別：碩士專業(yè)：信號與信息處理指導(dǎo)教師：王躍存20070101AbstractTheindexofyamfilamentirregularityisoneofmainindicestomeasllreyamqualityAtpreseat，therearemainlythreemethodstodetectyamirregularity：lengthm

2、easl】rememandweightmeasllrementmethod，instrumentsmeasurementmethodandvisualmeasllrementmethodThesemethodseachhaveadvantagesanddisadvantagesWiththedevelopmentofcompaterimageprocessingtechnique，takingfulladvantageofdigital

3、imageprocessingtechniquetoovercomedisadvantagesofotherdetectmethodsandimprovingthelevelofyamqualitydetectionhasbecomeanimportantresearchfieldandhasimportantsignificancetoenhanceyamqualityThispapermainlystudiesappearanceq

4、ualitydetectioninyamdetectionThepaperproposesamethodthatmakesuseofthecompaterimageprocessingandanalysistechniqueinqualitydetectionofyamblackboardandyamfilamentirregularityanalysisInthismethod，yarnimagesareobtainedbyscann

5、ingblackboardwhichistwistedbytheyamBydesigningamulti—structuralelementautoadaptivedeterminationweight’Smorphologyfiltertopreprocesstheimage，animagewithminimaldistortionisgotandtargetsignaliscompletelydetachedfrombackgrou

6、ndsignal，iewipingofftheyawpandstandingoatthetargetThenmaximum—squareerrormethodisusedtodeterminethethresholdvalueautomatically，andbinarizetheimageaccordingtotheadvisedthresholdAfterward，thefeatureoftheimageisextractedand

7、analyzed，ieextractingeverytargetsignal7Sfeaturevaluesfrompostpreprocessingimageandthencalculatingthediscriminantindex，suchasyamaveragediameter、CVvalue、thinandthicknode、severethinandthicknode、ring、nepnumbersandSOonbyextra

8、ctedfeaturevaluesHere，themethodbasedonabutyamfragmentsareatodetectandcountthenepnumbersisadoptedAtlasttheexperimentresultsareanalyzedanddiscussedTheexperimentsindicatethatusingcomputerimageprocessingmethodtodetectyarnqua