| In the process of silk making and silk weaving,the degree of difficulty in splitting the raw silk made of a plurality of silk filigrees is the raw silk cohesive property.It is an important indicator of the quality of raw silk and plays a very important role in the raw silk rating.Due to the low degree of automation,poor accuracy of manual detection,no objective quantitative indicators of the judgment standard,and non-retroactivity of the detection process,the raw silk cohesion detection device and method have problems.The detection of raw silk cohesive performance has become the most controversial item in raw silk quality testing.Therefore,this paper proposes a machine vision-based detection method for raw silk cohesion performance,and designs a set of electronic detection devices for raw silk cohesion performance that can be traced back.Firstly,for the existing raw silk cohesion detection device,only the single wire ingot detection,the friction blade retention position is not fixed,the raw silk tension is changed with the reciprocating friction,and the automatic detection function of the raw silk cohesion performance is designed.The mechanical part is composed of a multi-wire ingot synchronous fixing mechanism,a constant tensioning mechanism of the yarn,and a reciprocating friction mechanism of the raw silk.It can simultaneously test multiple wire ingots,and ensures that the tension of each wire is consistent during the friction process,improving detection efficiency and stability;the control part is composed of a single chip microcomputer,a data acquisition card,a control program and the like.The principle is that the friction tool holder automatically stops to a fixed position by controlling the number of motor turns,and real-time communication with the host computer by means of the data acquisition card,and the raw silk image and the number of times of cohesion are obtained.Secondly,since the diameter of the raw silk is between 30 um and 100 um,it isjudged that the cracked area causes a large error by directly acquiring the image of the rubbed area.Therefore,this paper collects the segmentation of the raw silk by the friction zone,and proposes a SIFT-based image stitching algorithm to synthesize a small-size image with multiple overlapping areas of the same scene into a large-size and high-quality image.Then,based on the machine vision method,the spliced image is subjected to threshold segmentation,interference information removal,image filling and raw silk edge detection,and the raw silk edge and the skeleton are extracted for curve fitting.The relative change of the diameter of the raw silk is obtained by calculating the linear distance between the upper and lower edge points of the raw silk;the cracked area is marked according to the axial length of the raw silk diameter to assist the manual judgment;raw silk cohesion performance is characterized by the number of frictions corresponding to half or more of the raw silk cracking area greater than 6 mm.Finally,in order to make the raw silk cohesive performance detection process reproducible,the research and development of the raw silk cohesive performance detection information management system.It is used to display raw silk image and processed raw silk image,and record raw silk image and cohesive detection data information,and provide information detection single query and test report export function.The quality inspector can check the raw silk detection result and the raw silk image information by detecting the single number,and at the same time,the detection process and the traceability of the detection result can be realized.The research and analysis of this subject can effectively detect the silk cohesion performance.Compared with the existing detection methods,the detection efficiency is greatly improved,and the quality inspection process can be traced back.This is a great progress in the field of raw silk quality detection,which is beneficial to further improve the silk.Industry quality management level and competitiveness. |
PDF Full Text Request