Heterosexual fibers in cotton refer to non-natural fibers and non-natural fibers, such as chemical fibers, hair, silk, hemp, plastic film and dyed thread which are mixed with cotton to seriously affect the quality of cotton and its products. Although the content of foreign fiber in lint is small, it has a serious impact on the quality of the textile. Once it is involved in the textile, it not only affects the spinning ability, but also causes various color spots on the cloth after dyeing, seriously affecting the appearance quality of the cloth, Textile industry caused significant economic losses. The establishment of cotton heterosexual fiber classification and identification system can effectively reduce the damage of the heterosexual fiber, meanwhile, the quality of the cotton can be reflected by the weight of the heterosexual fiber. For heterosexual fiber processing, the domestic and international uniform use of two methods, one optical detection, mainly the use of industrial high-speed CCD camera, with light-sensitive components of the cotton layer in the process of cotton layer imaging, after the image processing results; Is the use of sensors to detect, the use of capacitive sensors and micro-electronic technology for information detection, processing of the cotton layer to be the result of the test. Both of these methods have produced high-speed, effective, heterosexual fiber detection equipment that increases business productivity. Such as Germany TrÃ¼tzschler company SCFO heterosexual fiber detection and separation devices, Shanghai Textile Co., Ltd. Tatsu Zhaoyuan ZYG a 048 cotton fiber automatic detection devices. However, the existing foreign or domestic foreign fiber online cleaning equipment is generally expensive, poor cleaning effect, and does not have the function of classification and identification of foreign fibers, can not meet the requirements of accurate measurement of foreign fiber in cotton quality assessment test , The research in this field is still a blank. This paper presents a more adaptable SIW system. Combining with machine vision technology and image processing technology, this system uses the classification theory of decision tree to classify the opposite-sex fibers. Combining with the weight model, the weight of each type of heterosexual fibers statistics. 1 SIW system structure and working principle The system consists of image acquisition system, image processing system, heterosexual fiber classification and recognition system, eliminating system and weighing system. System workflow shown in Figure 1. After the system is initialized, the original lint image is extracted by the image acquisition system. The control of image acquisition is controlled by the host computer, and the original image information collected is sent to the image processing system. The system is selected by DSP system to enhance and filter the image , Segmentation and feature extraction. After the characteristics of the heterosexual fiber are extracted, the processing result is transmitted to the host computer, and the host computer sends the characteristic information to the heterogeneous fiber recognition system for recognition, and the recognition system sends the recognition result to the host computer. The host computer calls the corresponding heterosexual fiber weight model according to the identified type, calculates the weight of the heterosexual fiber, and saves it to the host computer. At the same time, the host computer starts the peripheral machinery to clear the foreign fibers. 2.1 Image Acquisition Equipment The system chosen to choose Canada DALSA P2-4X-04k40 industrial linear array of black and white CCD camera with UV light source, mainly for the acquisition of fluorescent white polypropylene filament and white plastic film. Image capture card with camera configuration capture card, the card is based on Camera Link bus. Choose a color camera, choose Canada DALAS PC-30 industrial CCD, equipped with LED linear white light source, mainly for the collection of colored heterosexual fibers, the sensor is three linear CCD, respectively, red, green and blue colors. Image capture card with camera configuration capture card, the card is based on Camera Link bus. 2.2 image processing equipment selection TIDM642 dedicated image processing DSP system. Combined with the function library provided by the system, a function suitable for cotton heterosexual fiber processing is developed, including enhancement processing, filter processing, grayscale processing, segmentation processing, contour extraction processing, area and perimeter processing. For the feature extraction part, simple color feature extraction can also be done on the DSP, for complex features need to be done on the host computer, such as morphological features. 2.3 foreign fiber rejection equipment The removal device is controlled by a pneumatic solenoid valve to complete, according to the width and flow rate of the cotton stratosphere set a row of high pressure nozzle, according to the host computer sent over the position of heterosexual fibers to control the action of a high pressure nozzle, Used to clear the area of â€‹â€‹heterosexual fibers. 3 SIW software design 3.1 image acquisition module image acquisition process goes through the process are generally: Start (initialization), parameter settings, capture color images to the screen or memory, capture black and white UV light source image to the screen or memory, the end (release resources). Before creating a program, first run the image capture card's CD-ROM installer to get the required library files. DLL, call the relevant include file (* .h) in the program and add the static link library (* .lib) file To the project file for the compiler to use when linking. After adding the library file provided by the image acquisition card into the VC program, the image acquisition system must first open the specified image acquisition card and read out the parameters set before the last shutdown to initialize and obtain The required system settings, create a reference handle for the card for reference by each function. CCD camera calibration method using linear calibration method, through the image pixel coordinates to image plane coordinate transformation, image plane coordinates to camera coordinates, so that the image of the center pixel coordinates correspond to the real space cotton layer of the center coordinates. Image acquisition module will get the image information directly into the DSP module through the bus. 3.2 DSP image processing module DSP image processing module to complete the color and black and white image enhancement, filtering, grayscale processing, the target segmentation extraction processing, contour extraction and color information extraction processing. Image enhancement techniques using Gaussian convolution smoothing techniques. Image filtering is performed using median filtering techniques. Both of these processing techniques are conventional. The grayscale image processing is the process of converting a color image to 256 grayscale images using the algorithm of formula (1). Y = 0.212671 Ã— R + 0.715160 Ã— G + 0.072169 Ã— B + 0 Ã— A (1) where Y represents the gray value of the corresponding pixel. R, G, and B represent the three primary color sizes of the color image pixel points. A representative. Image target segmentation processing using local Mean-shift technology. Firstly, the gray image was studied and found that the gray values â€‹â€‹of the cotton layer and the heterosexual fiber were mixed together. The background of the cotton layer occupied the absolute superiority in the pixel distribution of the entire image, while the opposite-shaped fiber image accounted for only a very small part. The range of gray values â€‹â€‹is roughly between 230 and 255, and the majority of the gray values â€‹â€‹of the heterosexual fibers are around 230. The traditional method of domain segmentation can not be used to complete the image segmentation. Considering the characteristics of the gray distribution of the heterosexual fibers, the local Mean-shift algorithm is adopted to segment the image. The Mean shift process is a nuclear density estimation method based on the Parzen window method for estimating the probability density function in pattern recognition. Corresponding to the window function in the Parzen window method, a kernel function K (x) is defined. In most cases, the kernel of symmetry is concerned, so it can be expressed as follows: where is the normalization constant that makes K (x) integral. Here, Gaussian kernel function is taken as an example to derive a recursive formula and Mean-shift vector with convergence: The grayscale and segmentation results are shown in Fig.3 and Fig.4. Take an image of a layer of cotton that contains feathers and colored threads of the opposite sex as an example. Figure 3 grayscale renderings Figure 4 Segmented renderings Contour extraction processing and color information extraction processing heterosexual fiber feature extraction is the two main steps. The feature extraction of the heterosexual fiber includes two aspects, the shape feature and the color feature. The shape feature has been found through the study, you can extract the corresponding contour torque as the identification of the main features of the opposite sex fibers, appearance ratio, duty cycle and roundness as the second classification features, and color information can be identified as an auxiliary feature. Contour extraction processing using hollowing point method. Since the image is converted into a binary image after the image segmentation process, it is effective and quick to hollow out the dots. Contour extraction results shown in Figure 5. With the help of the completed contour extraction, combined with the original color image, the method of line scan and column scan can mark the color target area and extract the corresponding color target. At the same time, the characteristic torque and the position coordinates of the corresponding fibers can be obtained. The effect of the mark and the color target are shown in FIG. 6 and FIG. 7. Figure 5 contour extraction results Figure 6 color target calibration graph Figure 7 color target extraction effect Figure 3.3 heterogenous fiber recognition module 3.3.1 heterosexual feature selection Rough Sets (Rough Sets) theory is a characterization of incompleteness and uncertainty Mathematical tools, can effectively analyze and deal with all kinds of incomplete information such as inaccuracies, inconsistencies, incompleteness, and find out the hidden knowledge and reveal the underlying law. At present, it has been successfully applied in artificial intelligence, knowledge and data discovery, pattern recognition and classification, fault detection and so on. The rough set theory is introduced into the feature extraction method of cotton heterosexual fiber, and the effective image feature vector is extracted. The basic rough set theory includes two aspects, that is rough set knowledge representation and rough set reduction theory. 1) The knowledge representation method of the heterosexual fiber rough set Knowledge based on rough set theory is mainly based on the information table, an effective data table knowledge representation system. Data table The basic components of knowledge representation system is the collection of research objects. Knowledge about these objects is described by the characteristics of the specified objects and their eigenvalues. For the contour images of cotton foreign bodies, taking a sample of each foreign fiber as an example, the characteristic table of the properties of the foreign fibers obtained is shown in Table 1. Table 1 Characteristics of Heterogeneous Fibers Feature Heterofibers Sample Profile Torque Appearance Ratio Duty Cycle Circularity R Mean G Mean B Mean Corner Mean Hair 0.020 2.921 0.060 192 16 Twine 0.115 3.987 0.189 203 27 Feather 0.075 5.635 0.158 98 24 Red Polypropylene 0.075 3.489 0.456 198 32 Blue Silk 0.098 5.980 0.198 225 39 Red Silk 0.101 3.370 0.175 227 28 Black Plastic Cloth 0.168 4.509 0.780 226 36 2) Reduction method of heterosexual fiber rough set After the reduction of the knowledge table of the heterosexual fiber, the attribute column of the corner point is deleted, and the detection of the corner point belongs to the low-level image processing. Because of the complexity of the heterogeneous fiber image, the mean value of the corner point can not be used as an effective feature So remove the corner of a column, retaining the outline of torque, appearance ratio, duty cycle, roundness and RGB average attribute column, in which the contour torque and area and circumference are directly related to the first level of classification criteria, you can Fibers are divided into silk, bar and film categories; appearance ratio reflects the degree of the rectangle of the target contour, the duty cycle reflects the target area Degree of fullness, roundness reflects the degree of circularity of the target, these three characteristics as a second level of classification criteria; RGB mean value of the color information of the opposite sex fiber, which does not have the classification of the global attribute, but in a certain part of the distinction Degree, can only be classified as a low attribute. 3.3.2 Classification of foreign fibers According to the definition of the contour moment , the contour moment D describes the ratio of the longest diagonal dm to the vertical diagonal dp in the contour, as shown in Fig.8. Figure 8 illustrates the outline of the moment of torque The study found that the size of the contour torque and the area and the circumference is directly related to the ratio, so here proposed the use of the area S and perimeter L ratio as a new feature of the outline of the contour classification. It is defined as follows: Eight kinds of heterosexual fibers of hair, polypropylene yarn, hemp rope, colored thread, feather, cloth, tape and plastic sheet are taken as research objects by using outline moment as a rough classification standard, Sample analysis, the resulting contour of the heterofibrous fiber profile shown in Figure 9: Figure 9 contour torque distribution diagram tf said hair, bls said polypropylene filament, ms said hemp rope, ysx said colored line, jm feathers, bt said cloth, cd said tape, slb said plastic film. (1) The first classification criteria are based on the contour moment. It can be seen from Fig. 9 that the moment of hair is distributed between 0.04 and 0.06; the moment of polypropylene yarn, hemp rope, colored thread, chicken feather and cloth is distributed between 0.06 and 0.15; the moment of tape and plastic piece Distributed between 0.15 and 0.225; (4) The distribution shows a three-stage distribution, corresponding to the three categories of silk category, category and category. (2) For the classification of blue silk, red silk, hemp rope, feather and red polypropylene filament in the strip, combined with the property list of the opposite sex fiber, through appearance ratio, duty ratio, roundness and RGB mean Features can be effectively identified. Based on this, the classification idea based on decision tree is proposed, and the optimal combination of experiments is obtained through a large number of experiments to get the flow chart of decision tree classification shown in Fig.10. "Other" is a foreign body that does not conform to the characteristics of the opposite sex fiber. Figure 10 Decision Tree Recognition Figure experimental classification according to the above method, the results shown in Figure 11. (1) Hair (2) Feather (3) Black plastic sheet (4) Hemp rope (5) Red silk (6) Blue silk (7) Red polypropylene filament (8) Fiber weight, type under the premise of about 20000 about all the images were all image processing, to get the area of â€‹â€‹its foreign fibers, select the effective sample, remove the largest sample and the smallest sample, calculated for each foreign fiber The average weight per unit area, as shown in Table 3. Table 3 Average basis weight Weight Heterosexual Sample Hair Hemp Feather Black Plastic Sheet Red Polypropylene Silk Blue Silk Red Silk Fluorescent White Polypropylene Ribbon Fluorescent White Plastic Sheet Average Unit Weight (mg / mm2) 4.00 6.00 1.65 2.35 0.73 1.17 5.16 0.86 0.62 4 Conclusions In this paper, a system of identifying and weighing (CMS) system for identifying and weighing cotton heterosexual fibers has been established to realize the full automation of testing and weighting of the heterosexual fibers. Firstly, the composition and working principle of the system are put forward. The necessary hardware and software implementation process of the system is explained. Then based on the rough set theory, the data of the different fiber characteristic parameters are analyzed and the effective differences Fiber eigenvalue; and research on classification and identification of different fibers and weight model. Finally, the SIW system is tested, the test results show that the system can effectively identify the heterosexual fibers in cotton, and accurately get the weight of all kinds of heterosexual fibers.
Women's Cashmere Sweater uses cashmere as raw materials, which is suede, soft feeling warmth retention property good, and has a strong hygroscopicity natural luster downy, wearing comfortable, elegant and showily, unique style, feels showily elegance, double charm, and it can enjoy a warm and elegant demeanor, which is ideal product to pursuit of high-grade and perfect clothing for women.
Our company's women cashmere sweater is fashionable with rich colors, fashionable. Women's sweater has upper collar, double lapel, big V collar, square collar, round collar and collar. Deformed big V is fashionable implicative, more beautiful. Because of the physical properties of cashmere sweater, cashmere sweaters are easy to piling and static.
Please note that the inside bag should not be loaded with hard objects, and don't insert pens, etc., so as to avoid local friction. And Wear time not too long, ten days then make it bounce back to avoid fiber fatigue. Cashmere sweaters are usually dry-cleaned and hand-washed; put special lotion in 35 â„ƒ water and then put soaked cashmere sweater in it and soak 15-30 minutes, in key smudgy place and neckline cuffs can use a high concentration of detergent, washing methods take knead crowded, the rest parts knead gently. Jacquard or multicolored cashmere sweaters should not be soaked, and different colored cashmere sweaters should not be washed together to avoid color.
Women`s Cashmere Sweater
Women`S Cashmere Sweater,Women'S Grey Cashmere Sweater,Ladies Cashmere Sweater,Long Sleeve Cashmere Sweater
Dehong International Cashmere CO., LTD. , http://www.dehongrongye.com