Research On Automatic Identification Method Of National Wool And Australian Wool Fiber

Wool is an important textile material.With the improvement of people's living standards,high-end wool products have gradually become popular.Because of its super high-quality characteristics,Australian wool has always been the first choice for the production of high-end wool products.However,due to the high price of Australian wool,under the temptation of high profits,some manufacturers adulterated counterfeiting or illegally used pure wool marks,which seriously harmed the interests of consumers.It is necessary to find a way to distinguish Chinese and Australian wool.Strengthen the construction of textile standard testing,provide better technical support for product quality,regulate the market,and safeguard the rights and interests of consumers.For the identification of similar textile fibers,many identification methods have been studied for a long time,but the manual identification method based on microscope is still the most important detection method at present.This method is time-consuming,laborious,inefficient and easy to misjudge.A fast and accurate automatic identification method is needed.To strengthen the research of similar fiber identification methods,it is necessary to keep up with the latest technology at home and abroad.In recent years,computer image processing and recognition technology has developed very well,and it has gradually been used in the detection of textiles.The main work of this paper is as follows:(1)Based on the experience of artificially identifying textile fibers,a variety of microscopes were used to collect fiber images.The quality,efficiency and cost of fiber images were compared and analyzed,and the equipment for collecting fiber images was selected.The images of Chinese and Australian wool were collected by this device as the follow-up experimental data.(2)Process the captured images.In the research of fiber image processing scheme,image processing technology is adopted,and MATLAB software is used for image processing experiment.Combining the condition of the sample and the processing result,the fiber image processing scheme suitable for the environment studied in this paper is determined.(3)When extracting the characteristic parameters of the two kinds of wool fibers,the texture characteristic parameters are extracted.The gray level co-occurrence matrix method(GLCM)was used to analyze the surface texture of Chinese and Australian wool fibers,and 13 fiber texture features for subsequent automatic identification were selected and extracted.(4)Based on the small sample,the support vector machine(SVM)is used as the classifier.Through simulation experiments,the kernel function and parameters are determined according to the actual experimental results of the samples.In this paper,a polynomial kernel function is used to train and obtain the fiber recognition and classification model,so as to realize the automatic recognition of Chinese and Australian wool fibers.The results showed that the recognition rate of Chinese wool and Australian wool fiber was 92.68%.