Inspection On The Appearance Quality Of Lattice Apron Based On Image Processing

Currently the compact spinning equipments have a wide domestic application, among which the lattice apron accounts for a relatively large proportion. The lattice apron, as one of the most important components, its quality and service condition have a significant influence on the quality of yarn. The textile industry standard "The Lattice Apron for Compact Spinning" puts forward the main evaluating indicators on the quality of the lattice aprons, including porosity, cutting edge, appearance quality, and color and luster etc., providing a basis for the assessment of the quality of the lattice aprons. The factor of porosity is one of the key parameters for the evaluation of the lattice apron quality, however, the standard does not mention the method of measuring the porosity, and manual test or projection amplification method are two main testing methods of porosity at present, of which the testing results are substantially affected by subjective factors, consequently, the data is not accurate enough and it is difficult to achieve batch testing. Therefore, not only the production quality control of the lattice aprons is influenced, but also the stable good quality of the yarns produced in the textile factories.The development of digital image processing technology makes it possible to use machine vision instead of traditional manual test in detecting textiles. This paper deals with the detection for the lattice aprons appearance quality indicators (porosity, pore distribution etc.), extracting relevant information about quality parameters using the computer image processing technique. The method not only has a good reproducibility but also is intuitive, simple, highly accurate, and it is possible to rapidly detect a large number of samples and provide a quantitative index of the lattice apron appearance quality, which provides a quick, objective detection method for the lattice apron manufacturing enterprises as well as the enterprises using lattice aprons.This paper deals with the lattice apron image collection, tilt correction, threshold segmentation, yarn centerline calibration, and edge cutting as well as pore identification, fabric structure parameters and the actual pore characteristics extraction, then compares the yarn densities and diameters extracted by the images with those obtained from manual tests and analyses the results in order to illustrate the feasibility and accuracy of the image method, in the meanwhile, the pore uneven distribution and its shapes are hardly measured through the manual test method, which reflects the superiority of the image method. The main works and results of the paper are as follows:(1) Obtain the lattice apron images by using a high resolution scanner, and then analyze the corresponding relationship of the image sizes and the actual fabric sizes. Use the Radon transformation to analyze the tilts in the images and correct them, thus to guarantee the consistent of the fabric warp direction with the image column, providing the foundation for subsequent image processing. (2) During the image segmentation, analyze the pore sections and the sectional distribution of yarn gray levels, using the bottom grey value obtained from the image histogram bimodal method combined with modification value, the grayscale images are transformed into a binary images, which can fully display the segmented pore sections.(3) Calibrate the yarn central line position and remove the sections outside the four side yarn central lines based on the projection of warp and weft on binary image, thus the yarn number is an integer, making it more accurate of information about the subsequent extracted warp and weft densities.(4) Identify each pore in the lattice apron binary image by either removing the false pore method or filling the real pore method, we can obtain the pores of whose shape and size can reflect the real situation.(5) Extract the basic structural parameters including the fabric density, the diameter of warp and weft yarns, and calculate the theoretical porosity, which is consistent with manual one by comparison. At the same time, we can gain the parameters of the maximum pore area, the minimum pore area, the mean pore area, the irregularly pore size distribution, the actual porosity etc., which cannot be obtained by manual method.(6) Analyze the influence of abrasion on the appearance quality of the lattice apron, through the experiments with different abrasion times, we can discover that the unevenness of pore areas increases gradually with the increasing abrasion times, while the average pore areas and rectangularity decrease steadily as well as the real porosity.