TW380180B - On-line uniformity inspection method and apparatus for web applying digital image processing - Google Patents

Abstract

A kind of on-line uniformity inspection method and apparatus for web applying digital image processing which employs digital image processing technique to proceed on-line uniformity inspection for the web. The method employs optical theory and CCD camera to calculate theoretical equation of the web transparency and weight per unit area and employs numerical analysis to get the best approximate theoretical equation based on the inspected transparency and weight per unit area. When executing on-line inspection, the web transparency measured by the computer visualization is transformed by approximate theoretical curve to calculate various relevant data for uniformity variation of web weight per unit area and applies Fast Fourier Transform on the uniformity variation to inspect whether the machine has irregular operation and periodical variation during the manufacturing process. The invention can be applied for on-line uniformity inspection system for manufacturing of web or using the web as base material, such as non-woven fabric and papermaking.

Description

A7 -B7 V. Description of the invention (When the basis weight is small, the uniformity of the cotton web is good, and inspection 丨 Nan ^: Li ^ improves the quality. There are two main types of traditional detection techniques, namely inspection of the nets. Sentence and characteristics' Eli cotton net (-) Measure the thickness of the cotton web with a floating pressure roller. This is used to convert the weight of the cotton web into heavy contact _ # 测 4 broke the difference __ some physical cotton web (under 100g / m2) The measurement is especially obvious. 3 (two) Ϊί? ㈣ * 3 ray]. Because the radiation dose is harmful to human health and cost considerations, so early sampling values are generally used as the basis for measuring the uniformity of cotton webs. The average weight of a uniformed army pays for the weight of 4 standards, so relying on light weight as the self-measurement affects the credibility of the overall measurement accuracy. It represents 1 ± insufficient, so the test method for the uniformity of the test cotton net is -fine ^- Adjust the characteristics of the lion mesh material to break the lion net material. The rhyme on the cotton net can be adjusted by using the butterfly to adjust the ribs, and set the size of the sampling area to prevent uniformity. Changes can improve the measurement accuracy without harming human health and the overall response of the clan ^ t III-! I- I -II 1 — I II I -.1; I 1 Read the notes on the back before filling out this page) The relevant graphics printed by the staff consumer cooperative of the Central Bureau of Standards of the Ministry of Economic Affairs are convenient for review by the reviewing committee. Description of the drawings: The first diagram is a diagram of the system device of the present invention; the second diagram is a diagram of the overall architecture of the system of the present invention; the third diagram is a flowchart of the light source compensation of the present invention; the fourth diagram is a diagram of the light source compensation of the present invention; Theoretical relationship diagram of light transmittance; the flowchart of calculating the basis weight and gray level M curve of the present invention; the eighth diagram is the approximate curve of the least square method of the present invention to obtain the approximate curve. The line diagram; the seventh diagram It is the detection flow chart of the system of the present invention; the description of the drawing numbers: (1) the light source (2) the cotton net camera (4) the image mining card (5) the computer The details of the present invention are as follows: This invention: The method and device for self-examination and inspection of E-turn image processing are as follows: (Please refer to the appropriate paper size of this paper _ (—mm)

-1T ΘΙ. A7 B7 V. Description of the invention (i) See Figure 1 and Figure 2), it is composed of light source 1, CCD camera 3, image capture card 4 and personal computer 5, where CCD camera 3 is used to capture The image of the cotton measuring net, the image card 4 is used to store the image, and the personal computer 5 is used to analyze the image of the female computer, and the result is stored and displayed on the output device. With the combination of the above-mentioned components, the method developed by the present invention is applied to achieve the uniformity detection on the cotton net. The characteristic of this method is to use the optical theory (see @ 四) to obtain the theoretical equation of the relationship between the light transmittance of the cotton web and the basis weight through the camera 3 and the light source 1 (see Figure 5), and then use the method of numerical analysis The best approximation theoretical equation (see Figure 6) is obtained by calculating the light transmittance (see Figures 1, 2, and 4) measured by computer vision and the basis weight data. When performing on-line inspection, the computer-aided light transmittance measured by computer vision is converted through the best approximation theoretical curve to obtain various relevant data of the basis weight change (see Figure 1), including the variation of the CV% dynamics Figure 2 · !, Basic Weight Statistics Figure 2-2, Basic Weight Monitoring Figure 2-3, and Basic Weight Movement Figure 2-5. And use the uniformity change ’to make the fast Fourier transform spectrum diagram 2-4, and check that the machine has abnormal operation and periodic variation during the manufacturing process. The theory and purpose of this method are described in detail as follows: '. (1) Light source compensation (see Figure 3). The purpose is to compensate for the unevenness of the light or the CCD camera itself and cause the phenomenon of uniform image intake. The present invention uses a standard uniform grayscale plate to capture its image as a standard image (the grayscale value of the ax ^ point is h ( x'y), and find the standard image average GrayAVE. With the standard image, the compensation image S (x, y) can be obtained, and then extract the image g (x, y) we want to analyze, plus' ^: processing More accurate. The method is as follows: y) Ding image 1 y ^ b GrayAVE = — Σ Σ S (x, y) = h (x, y) -GrayAVE G (x, y) = g (x, y) + S (x, y) where a. B is the length and width of the image acquisition range (in pixels) ⑴ Read the precautions on the back before filling this page]

.C-° Γ (2) Theoretical relationship between basis weight and transmittance (see Figure 4). Firstly, through the light absorption theory (the exponential law 〇f absorpti〇n), diarrhea took the lead in closing the blanket in Texi. 1 Obtained by JW = e ~ ahvdn printed by Employee Consumer Cooperatives of the Central Bureau of Standards and Economics of the Ministry of Economic Affairs where (2)

The intensity of the light received by Iu at the pixel (i, j) position on the image is 10 = the intensity of the projected light α = the light absorption coefficient of the material. The degree applies to the CNS Α4 specification (210X29 ^ 5 " A7 B7, invention description (") hi j = thickness of the material to be measured du = density of the material to be measured and the image of the area captured by the camera, the unit pixel area (Pixel) weight π ′ · is = dij htj Aij (3) we The function of converting the unit pixel area and the weight of (ριχει) into i .. can be obtained by equations (3) and (4).

Wy = aHf) A ^ Therefore, the weight of the image captured by the camera is ⑷

W: ΣΣ1πΦη = f / «iy * i on yu xij (5) The average area area weight is one > Λ ab T a 〇ir Ψ = ~ Έ / Σ Σ Λ = ~ ^ (y-) ^ / = iy» i iij / = iy = ia jj where ab is the length and width of the imaging range (unit: Pixei) ⑻ (3) The best approximation equation for the relationship between the light transmittance and the basis weight using the least square method (see Figure 6). When measuring cotton webs of the same material, α is constant; assuming that the light source projection intensity is constant /, it is constant, so we can know from the formula (6) that the basis weight is only related to the light intensity (or gray level value) received by the sensor. ), And the relationship between the light intensity and the grayscale value received by the sensor is 'experimentally known that there is a positive correlation between the two <' ° Here we can approximate the formula (6) by using the least square method of numerical analysis to replace f, 'cause

W Printed by the Consumer Cooperatives of the Central Bureau of Standards, Ministry of Economic Affairs

Cifi (X) = C1 + C2X + C3X2 + ... + CnXn-1 ⑺ i: is the coefficient to be measured. [: The average grayscale value of the image transmittance in the area captured by the camera, X is inversely proportional to the ln (^) constant. ‘With / 0 as the paper size Applies to Chinese National Standard (CNS) A4 specifications (210X297 mm) (Please read the precautions on the back before filling this page)

A7 B7 V. Description of the invention (y) fi (X): Approximation equations can be combined by any complete orthogonal basis equations. This article uses polynomials to describe mi ^ \ m τχ)% χ …… txr / = ι '. = 1 MP 'pj · ί > / ^ -1; = 1 Σχ; ... Σ ^ + Ι y = i txrr ixf Σ ^ + 1. Bu 1 / »1 / = 1-ρ 一 / _el c2 β. Λ — 2 toilets / »1 C3 = m 2_ iM Cn JJ — Y.xpWi a; * l (8) Therefore, by measuring m standard cotton web samples, the average grayscale value χ and the average basis weight $ Then, the moment _ equation of mathematical formula (8) can be obtained by approximating the minimum flat green with n ΐίίϊ, and the coefficient Ci of the equation can be evaluated. Then, the best approximation of the order η of the equation is evaluated by the least square error equation (8). value. .ζ = Σ ^ = Σ (^-^) 2 pK-qc ^ -c ^ 2-. ···· ~ cnxr1) 2 (9) Among them, printed by the Consumer Cooperatives of the Central Standards Bureau of the Ministry of Economic Affairs w >: The average basis weight measured for the i-th sample. %. Is the actual average basis weight of the i-th sample. From the error of L when n = 1, 2, ..., the following two methods are used to find the optimal 11 and [10,000 when 1. s L is the minimum order 0, and c, .. Method 2: When L is less than the allowed error, the order ^ and c. Substitute Fangqian (7) as the system's best approximation equation. 1 (4) Find the basis weight of the cotton web (see Figure 7). This paper size is applicable to Chinese national broadcast, (Please read the precautions on the back before filling this page)

卞 «4 public A7 V. Description of the invention (ζΓ) Although the details of the coffee are x, the best (5) calculates the various characteristics of the cotton net (see Figure 7). With the basis weight data obtained by dynamic detection, the average sentence degree variation number of the cotton net and the probability value of the basis weight statistical graph can be obtained. cv% HP ^) ΐ = ϊ P (^) =

N (10) (11) (12) where cv% is the uniformity variation number κ: the average value of the basis weight from dynamic sampling to time t N: the total number of samples from the moving edge to time t. η: The number of basis weights is i. P (K). Basis weight is the probability of i occurring. (6) Use the Fourier transform spectrum to monitor the normal operation of the machine (refer to Figure 7) The invention uses the fast leaf conversion to calculate the basis weight to detect the basis weight _ $ 1 ~ ',' The special leaf calculation formula is explained as follows: = 万 客 叩) Which [-2,7from] = Talk) + / To do) (13) ---------- (Please read the note on the back before filling in this page) -a Among them, (14) The iF (X) printed by the Consumer Cooperative of the Central Standards Bureau of the Ministry of Economic Affairs: the average weight of the basis weight obtained from the Xth image. Jealousy (w): For Fourier Transformation >) is the real part of j). (w) is the imaginary part of (w) 丨 F (w) 丨 is the Fourier spectrum Printed by the Consumer Cooperatives of the Central Standards Bureau of the Ministry of Economic Affairs A7 B7 V. Description of the invention (b) 'From the above invention' The detailed description of the monitoring system for the online uniformity of the cotton net by the camera and the monthly display show that the present invention is of great industrial use value. The advantages are summarized as follows: This system is non-contact detection and does not damage the detection. Any physical properties of the material. (2) · = The sampling template of the system can be set to two-dimensional area sampling, so the measured data can represent the area and domain characteristics at that time. It will not change due to the density of small areas, which will affect the overall measurement accuracy of the system. (3). The system can use the Fourier operation of the uniformity data package to monitor whether there is abnormal operation and periodic unevenness in the competition. ~ One (four). This system can increase the number of cameras horizontally to control the uniformity of horizontal and vertical cotton webs, so that supervisors can easily control the state of the system and assist in tracking down the causes of unevenness. (5) The system can use the data obtained by measuring the uniformity to make the system's overall uniformity distribution statistics and variation analysis. (6) The system can display the online system, the relationship between the measured uniformity and the standard value and the allowable value (upper and lower limits) in real time. When the uniformity exceeds the allowable value, the system will send a warning to improve the quality. Control, and make online real-time screen display continuous changes in basis weight or thickness. In summary, the "overall technology and technical idea of the present invention" is an unprecedented device and method, which can achieve the expected efficacy. It can correctly detect the uniformity of the cotton web and improve the production efficiency and quality. Value; in line with the provisions of the Patent Law. It is provided in accordance with the law, with a special & special shooting request, please review it, and pray for and approve the patent in this case. It is the Lord ’s prayer. The aforementioned invention is based on a more comprehensive example, but it is not intended to limit it. In the present invention, any skilled artisan of earthenware can deviate from the spirit and scope of this letter, and should make some modifications and retouching. This should be regarded as the scope of protection of the present invention, and attached to the present invention later The scope of patent application is defined. (Please read the notes on the back before filling this page) ------ Τ ______ ^ ---- 1 ------

Printed by the Department of Industrial and Consumer Cooperatives 6. Scope of patent application-a device that uses digital transfer processing to resolve the average sentence on the Internet, a camera, an image capture card, and a personal electronic combination of t CQ) Photography & Used for & Take $ -W Electricity to calculate and analyze the image, and two methods for detecting the uniformity of ΐ 牡, 系, 系, 系, 光学, 光学, 曼, 曼, 麦, 麦, 机, 机, 机, 丰 ,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,, and value of the two images, and use them to determine the uniformity of the net. The theoretical equation of the relationship between the light transmittance and the basis weight, the numerical analysis method is used to measure the light transmittance measured by computer vision and the online detection, measured by computer vision = to use the package state diagram, basis weight statistics Map, basis weight monitoring chart and profit period = Fourier conversion is fine, to check whether the machine is in the manufacturing process "(3) The process of relying on the position image described in the special item 2 to measure the uniformity of the cotton web line For example, non-woven fabrics and papermaking (4) The application of digital image processing described in item 2 of the ϋSpecialized Circumvention Method for U-Line Uniformity Detection on Cotton Nets can be detected by the system CV% _SI. 3 Uniformity), the variation of the deviation is characterized by the monitoring of the quality. Basis weight (or the application of digital image processing as described in item 2 on the cotton mesh line range J (6) The application of digital image processing as described in item 2 on cotton mesh line detection uniformity method, can lay the basis weight monitoring chart When you know if the current basis weight changes the threshold set by the quality control system, the system can send a warning to notify the monitor C (the machine d, check the cause of the unevenness, to ensure the quality of the cotton net. The second item of the scope of patents and patents The application of digital image processing to the cotton web line 2 can be used to know the change of the base (8) of the cotton web at the current detection position through the dynamic image of the basis weight, so as to facilitate field personnel to control the quality. . Inspection = digital image processing as described in item 2 of the scope of patent application = where the Fourier transform is obtained and then used to obtain the Fourier = average = ⑶ When different peaks are generated, this can be used as a reward. ^ (9) 1 Please turn to the second item of Lin ’s method of image processing and m-measurement. It can be obtained by setting up multiple cameras in different Fang Jun. (1) (2) t Order: ——- Macao Standard (CNS) A4 Specification (210X297 mm

Claims (1)

(1 CM and longitudinal uniformity change are its characteristics. As described in item 2 of the scope of patent application, the method for checking the uniformity of cotton webs using digital image processing can be based on the physical characteristics of the cotton webs, with various cameras and Light sources of different lengths (1 1) are used as the best hardware to facilitate the measurement of various materials. ”As described in item 2 of the patent application scope, the method of applying digital image processing to the uniformity detection of cotton webs can be borrowed. The range of the image taken by the camera is set to sample points, lines, and planes in various azimuth ranges, and various data on the uniformity of the cotton web are measured to make the measurement data more representative and accurate. Q1? The method for detecting the uniformity of cotton webs using digital image processing as described in item 2 uses a standard uniform grayscale board to capture its image as a standard shadow ^ (axb) 'The grayscale value of each point is h (x , Y), and find the average value of the standard image GrayAvE. Subtract GrayAVE from the standard image points to obtain the compensated image S (x, y), and then capture the images g (x, y) 'g to analyze. (x, y) for more accurate image processing Among them, a.b is the length and width of the image range (unit Pixel), in order to achieve the compensation of the light source.] As described in the second patent application scope, the digital image processing is used to detect the uniformity of the cotton mesh line. The method 'first uses the light absorption theory to obtain the relationship between the basis weight and the light transmittance in the camera', and then calculates the pixel unit area, the weight of the area image, and then calculates the average weight of the area area. (1 4 ) As described in item 2 of the scope of the patent application, the method for detecting the uniformity of the cotton web by applying digital image processing is to find the relationship between the light transmittance and the basis weight of the cotton web by the least square method (1 5). The digital image processing described in item 2 is used to check the uniformity of the cotton web. The method 'when performing the detection of the basis weight of the cotton web' can be obtained by the average grayscale value measured by image processing through the best approximate curve equation. Conversion to obtain the basis weight of the cotton net. (Please read the precautions on the back before filling in this page) --I order --- ___ β ___ The paper size of the printed paper is applicable to China National Standards (CNS) ) 4 threat (Dong 21〇x297 male)