Intelligent monitoring system for fracture image of transversely-drawn film

Abstract

The invention discloses an intelligent monitoring system for a transverse-drawing film fracture image, which comprises a film image module, an image processing module and a feedback output module, and is characterized in that the film image module is used for acquiring and transmitting an original film image in uniform-speed processing by shooting equipment; the image processing module is used for calculating the fracture condition of the monitoring film and generating an alarm signal; the feedback output module is used for receiving the alarm signal and carrying out sound-light alarm; the film image module is electrically connected with the image processing module, and the image processing module is electrically connected with the feedback output module. The system disclosed by the invention can be used for accurately detecting film breakage, so that excessive dependence on manual detection is reduced, and the problem of production loss caused by film breakage is effectively reduced and even solved, so that field workers are helped to efficiently solve the problem of film breakage.

Description

Technical Field

[0001] This invention relates to the field of thin film rupture detection technology, and specifically to an intelligent monitoring system for transverse tensile film rupture images. Background Technology

[0002] Currently, the film industry is one of the largest and most diverse polymer material industries in my country. With the rapid development of my country's economy, the demand and output of the plastic film industry have been increasing year by year, with an average annual growth rate exceeding 15%. Biaxially oriented film, or biaxially stretched film for short, is a film product obtained by stretching the raw film material longitudinally and laterally under tension at a temperature below its melting point and above its glass transition temperature using chemical, physical, and mechanical methods, followed by heat setting under tension.

[0003] Biaxially stretched film is divided into biaxial synchronous stretching and biaxial asynchronous stretching. Synchronous stretching is performed in the same stretching machine, where the film does not contact the roller surface, and both longitudinal and transverse stretching are completed simultaneously, followed by heat and cold treatment. Synchronously stretched films have uniform physical properties. Asynchronous stretching requires two stretching processes: first, longitudinal stretching, then clamping the longitudinally stretched film in a clamping zone for transverse stretching. The stretch ratio in the transverse stretching zone is adjustable, making asynchronous stretching more flexible than synchronous stretching. Asynchronous stretched films also have better mechanical strength, tensile strength, and a lower breakage rate than synchronously stretched films. However, because the transverse stretching zone is located in a closed, insulated chamber, its length can reach tens to hundreds of meters. This long, enclosed production zone makes it difficult for workers to observe the film rupture process in a timely manner and determine the cause of the rupture. Therefore, in traditional production, workers can only solve the problem of film rupture in the transverse stretching zone by constantly adjusting process parameters, which is time-consuming, material-intensive, and reduces production efficiency. Summary of the Invention

[0004] The present invention aims to solve at least one of the aforementioned technical problems by providing an intelligent monitoring system for transverse film rupture images. This system can accurately detect film rupture, reduce excessive reliance on manual inspection, and effectively reduce or even solve the problem of production losses caused by film rupture, thereby helping on-site personnel to efficiently solve the problem of film rupture.

[0005] To solve the above-mentioned technical problems, the present invention adopts the following technical solution:

[0006] A transverse tensile film rupture image intelligent monitoring system includes a film image module, an image processing module, and a feedback output module, characterized in that...

[0007] The thin film image module is used to capture and transmit the original thin film images during uniform processing; the image processing module is used to calculate and monitor the film's rupture status and generate alarm signals; the feedback output module is used to receive alarm signals and provide audible and visual alarms; the thin film image module and the image processing module are electrically connected, and the image processing module and the feedback output module are electrically connected.

[0008] The thin-film image module includes an image acquisition module, an illumination source module, and a raw image transmission module. The image acquisition module is used to acquire images of a thin film passing at a constant speed. The illumination source module is used to increase illumination during the acquisition process of the imaging device to make the acquired images clearer. The raw image transmission module is used to output the images acquired by the imaging device. The image acquisition module is electrically connected to the illumination source module and the raw image transmission module.

[0009] The image processing module includes a receiving improvement module, a fine processing module, and a fracture mode determination module. The receiving improvement module receives the original thin film image acquired by the imaging device and performs grayscale processing. The fine processing module performs fine processing on the grayscale processed thin film image. The fracture mode determination module determines the fracture mode of the thin film based on the feature extraction of the finely processed thin film image and generates an alarm signal. The fine processing module includes an image preprocessing submodule, an image segmentation submodule, and a fracture feature extraction submodule. The thin film image preprocessing submodule enhances and denoises the improved thin film image. The image segmentation submodule segments the target and background and extracts useful information. The fracture feature extraction submodule extracts features based on key information about the thin film fracture. The image preprocessing submodule is electrically connected to the image segmentation submodule and the fracture feature extraction submodule. The receiving improvement module is electrically connected to the fine processing module and the fracture mode determination module.

[0010] Furthermore, when installing the imaging device used to acquire images of the thin film in the image acquisition module, the lens of the imaging device is kept at an angle of 10° to 20° with the horizontal plane where the thin film is located, more preferably at an angle of 15°.

[0011] This invention also provides a method for operating the above-mentioned intelligent monitoring system for transverse film rupture images, which mainly includes the following steps:

[0012] S1. The film is stretched laterally in the equipment for the transverse stretching process and passes at a constant speed between the imaging equipment and the lighting equipment;

[0013] S2. The imaging device acquires the original image of the ruptured film and transmits the unprocessed original ruptured image to the image processing module;

[0014] S3. The image processing module performs grayscale processing on the received original image of the ruptured film, and then performs fine processing on the grayscale image;

[0015] S4. Analyze and determine the rupture mode of the ruptured film by extracting the rupture features, and generate an alarm signal;

[0016] S5. During the film production and output process, adjust the process parameters of different types of films according to the warning of film rupture.

[0017] In the above operating method, step S3 further includes the following steps:

[0018] S31. The image processing module performs grayscale processing on the received original image of the ruptured film;

[0019] S32. Linear transformation is used to enhance the grayscale image, and adaptive median filtering is further used to denoise the enhanced image;

[0020] S33. Perform Otsu thresholding and edge segmentation on the preprocessed image.

[0021] In the above operating method, step S4 further includes the following steps:

[0022] S41. Based on the segmented image, feature extraction and identification of film rupture are performed;

[0023] S42. Based on the area and perimeter values ​​of the transverse membrane rupture, distinguish whether the membrane rupture is a small-scale rupture or a longitudinal rupture, and generate an alarm signal.

[0024] Specifically, step S42 above further includes the following steps:

[0025] S421. Scan the segmented image from left to right and from top to bottom, and label the unconnected scan strokes in the same row or column with different numbers;

[0026] S422. Then scan the image from top left to bottom right and from bottom left to top right respectively. If adjacent rows or columns have a continuous path, mark adjacent columns with the same number. Finally, arrange the numbers.

[0027] S423. By counting the number of pixels in the target area of ​​the image, the target area A is the sum of the number of pixels with the same sign (grayscale value "1"). The formula for calculating the area A of the transversely stretched film rupture is:

[0028] ;

[0029] In the formula, f(i,j) is the pixel input point, and R is the target region in the image;

[0030] S424. The perimeter of the target region in the image is the length of the region boundary, and the perimeter P is represented by the sum of the pixels at the edge of the target.

[0031] S425. Define an initial matrix P(x,y) = 0. Similarly, scan the image and compare the pixel values ​​of two adjacent points. If one is 1 and the other is 0, then P(x,y) = 1, where x and y represent the edge coordinates.

[0032] S426. The sum of P(x,y)=1 is the desired perimeter P.

[0033] Compared with existing technologies, the system described in this invention enhances the contrast of transversely stretched film rupture images by employing linear transformation, denoises the images using adaptive median filtering, provides a basis for feature extraction by selecting the Otsu threshold segmentation method, and establishes a transversely stretched film rupture information database by using Canny operator edge detection. Furthermore, it selects the rupture area A and perimeter P to form a feature vector. Multiple experiments on samples show an accuracy rate of approximately 91%, achieving a preliminary design for an alarm system scheme for transversely stretched film rupture. This provides a theoretical basis for realizing an online monitoring system for biaxially stretched film production lines and a theoretical methodological research foundation for feedback control in actual production lines. Attached Figure Description

[0034] Figure 1 This is a schematic diagram of the module composition of the system described in this invention.

[0035] Figure 2 The image shows a schematic diagram of a ruptured membrane; where (a) is a small-scale rupture and (b) is a longitudinal rupture. Detailed Implementation

[0036] To better explain the technical solution of the present invention, the present invention will be further described in detail below with reference to the embodiments, but the implementation of the present invention is not limited thereto.

[0037] like Figure 1 As shown, the intelligent monitoring system for transverse film rupture images of the present invention includes a film image module, an image processing module, and a feedback output module. The film image module is used to capture and transmit the original film image during uniform processing by the imaging device. The image processing module is used to calculate and monitor the rupture status of the film and generate an alarm signal. The feedback output module is used to receive the alarm signal and perform an audible and visual alarm. The film image module and the image processing module are electrically connected, and the image processing module and the feedback output module are electrically connected.

[0038] The thin-film image module includes an image acquisition module, an illumination light source module, and an original image transmission module. The image acquisition module is used to acquire the images of the thin film passing at a constant speed. The illumination light source module is used to increase the light during the acquisition process of the imaging device to make the acquired images clearer. The original image transmission module is used to output the images acquired by the imaging device. The image acquisition module is electrically connected to the illumination light source module and the original image transmission module.

[0039] Specifically, the imaging device can be a camera. The parameters to be considered when selecting a camera include optical size, resolution, frame rate, shutter speed, and output interface. After comprehensively considering the actual requirements, time, cost factors, etc., it is preferably a Phantom v1612 high-speed camera. This camera can provide ultra-high throughput, excellent sensitivity, and the function of easily managing data. The maximum resolution it supports is 1280×800. For the selection of the camera lens, a Nikon AF-S 18-105mm f lens can be selected to meet the shooting requirements.

[0040] For the device providing the illumination light source, considering the complexity of the actual production line's horizontal stretching interval structure, in order to provide selectivity for arranging the light source in the actual production line and conduct different illuminations of point light sources and line light sources, two lighting devices are selected: one is a VISICO-LED photographic lamp supporting the high-speed camera, and the other is a common strip-shaped LED lamp. Features of the VISICO photographic lamp: high-brightness light supplement, uniform brightness, can be used for shooting and supplementing light for products under complex light sources, continuous light source without flicker, stable color temperature, stable brightness, stable power, etc. Features of the strip-shaped LED: can form a line light source, with a wide and uniform illumination range. At the same time, the lighting device is installed in a low-angle lighting manner, so as to obtain images with uniform light and clear thin-film rupture, and it is convenient for installation.

[0041] For the installation angle of the imaging device, since the reflection is the strongest when the installation angle of the lighting device is 10°, 15°, or 20°, and when it is 15°, total reflection can be achieved. At this time, the background of the captured image is simple, the light is uniform, and the contrast between the target and the background is the strongest. Therefore, when installing the imaging device for acquiring the thin-film images, the lens of the imaging device is kept at an angle of 10° to 20° with the horizontal plane where the thin film is located, and more preferably at an angle of 15°.

[0042] The image processing module includes a receiving improvement module, a fine processing module, and a fracture mode determination module. The receiving improvement module receives the original thin film image acquired by the imaging device and performs grayscale processing. The fine processing module performs fine processing on the grayscale processed thin film image. The fracture mode determination module determines the fracture mode of the thin film based on the feature extraction of the finely processed thin film image and generates an alarm signal. The fine processing module includes an image preprocessing submodule, an image segmentation submodule, and a fracture feature extraction submodule. The thin film image preprocessing submodule enhances and denoises the improved thin film image. The image segmentation submodule segments the target and background and extracts useful information. The fracture feature extraction submodule extracts features based on key information about the thin film fracture. The image preprocessing submodule is electrically connected to the image segmentation submodule and the fracture feature extraction submodule. The receiving improvement module is electrically connected to the fine processing module and the fracture mode determination module.

[0043] The operation method of the above-mentioned intelligent monitoring system for transverse film rupture images mainly includes the following steps:

[0044] S1. The film is stretched laterally in the equipment for the transverse stretching process and passes at a constant speed between the imaging equipment and the lighting equipment;

[0045] S2. The imaging device acquires the original image of the ruptured film and transmits the unprocessed original ruptured image to the image processing module;

[0046] S3. The image processing module performs grayscale processing on the received original ruptured film image, and then performs fine processing on the grayscale image. The fine processing includes preprocessing, segmentation and feature extraction of the grayscale image. Preprocessing mainly provides good ruptured image samples for the next segmentation step. Image segmentation can better highlight the rupture area and edges. Image feature extraction mainly extracts useful data information from the image to obtain a "non-image" representation or description.

[0047] S4. Analyze and judge the rupture mode of the ruptured film by extracting the rupture features, and generate an alarm signal; The rupture mode of transverse stretching of film is mainly divided into two types: small-scale rupture and longitudinal rupture. Both are caused by temperature changes. Small-scale rupture is mainly caused by the low processing temperature during stretching, resulting in holes. Longitudinal rupture is caused by the high temperature during stretching, resulting in the longitudinal rupture of the entire film along the stretching direction, with the "curling edge" phenomenon appearing at the rupture point.

[0048] S5. During the film production and output process, adjust the process parameters of different types of films according to the warning of film rupture; classify the rupture mode into two types according to the rupture form: small-scale rupture and longitudinal rupture, and use the identification characteristic parameters of area and perimeter to distinguish the two rupture modes, and further use the characteristic parameters to adjust the processing temperature parameter range of the film.

[0049] Step S3 above further includes the following steps:

[0050] S31. The image processing module performs grayscale processing on the received original image of the ruptured film;

[0051] Image grayscale conversion aims to remove the color effects from color images, converting them into grayscale images that better highlight important details. While grayscale conversion results in the loss of color information and a reduction in matrix dimension, it significantly improves computational speed and preserves gradient information. Converting an image to grayscale reduces the computational burden on subsequent image processing, preparing the image for later processing.

[0052] S32. Linear transformation is used to enhance the grayscale image, and adaptive median filtering is further used to denoise the enhanced image;

[0053] When performing contrast enhancement on an image, a linear transformation method is used. Specifically, the linear transformation function... It can be represented as:

[0054]

[0055] in, , These are the slope and intercept of the linear transformation function formula, respectively, which control the contrast and brightness of the image; , These represent the gray levels of the input and output images, respectively.

[0056] And when When the contrast of the output image increases; when At this time, the contrast of the output image decreases;

[0057] when and At this time, the grayscale value of an image pixel shifts up or down, making the entire image darker or brighter; for example Dark areas will become brighter, and bright areas will become darker;

[0058] when , When the output image is the same as the input image, when... , When this happens, the grayscale of the output image is inverted.

[0059] Perform a linear transformation The grayscale range of the processed image is expanded, and the contrast of the resulting image is increased. Therefore, linear transformation has the effect of enhancing the overall contrast of the image of the transversely stretched film rupture.

[0060] When performing noise reduction analysis on images, median filtering can be used. Median filtering is a non-linear filter that can be used to handle impulse and salt-and-pepper noise, effectively preserving the image's edge information during processing. Median filtering takes the median value of the pixels covered by the convolution kernel as the anchor pixel value. The expression for median filtering is:

[0061]

[0062] In the formula, These are the pixel coordinates currently being processed. The center pixel coordinates, Due to image size limitations, a 3x3 window corresponds to A 5x5 window corresponds to .

[0063] However, if the median filtering window is too small or too large, it will affect the image processing effect. Therefore, adaptive median filtering is considered. Adaptive median filtering, based on median filtering, flexibly changes the filtering window according to set conditions and autonomously judges noise. Therefore, adaptive median filtering has a better processing effect on images with randomness and locality.

[0064] S33. Perform Otsu thresholding (maximum inter-class variance thresholding) and edge segmentation on the preprocessed image. An image consists of a target, background, and noise. Image segmentation separates the target region, background region, and noise portion of the image to facilitate feature extraction. Considering the actual situation of the thin film and the high contrast of the image, thresholding provides a basis for feature extraction. Edge segmentation (using the Canny edge detection operator) facilitates the study of edge breakage causes and provides a foundation for establishing a thin film breakage information database.

[0065] The Otsu's algorithm, also known as the maximum inter-class variance method, aims to maximize the average gray-level difference between the foreground and background of the selected optimal segmentation threshold image. This maximized difference is represented by the variance of each region, and the optimal region segmentation threshold can be calculated automatically. The algorithm principle is as follows:

[0066] Let Nr be the number of pixels with gray level r in the image, and let the gray level range be [0, L-1]. The total number of pixels is:

[0067]

[0068] In the formula, N represents the order of magnitude of r, and L represents the total number of gray levels in the image.

[0069] The probability of each grayscale value appearing is:

[0070]

[0071] again Taking a threshold T0 in the range [0, L-1], the image is divided into two classes, G0 and G1, where... , The probabilities are obtained as follows:

[0072]

[0073]

[0074] Let the average gray level of the image be μ, and the average gray levels of G0 and G1 be:

[0075]

[0076]

[0077] The average gray level of the image can then be written as:

[0078]

[0079] The total variance of [0, L-1] for:

[0080]

[0081] If we take another threshold in [0,L-1], then the maximum threshold T is the optimal segmentation threshold.

[0082] The Otsu algorithm is the most autonomous, requiring no human intervention. It can automatically calculate the optimal region segmentation threshold, exhibiting good stability and strong versatility. Therefore, the Otsu method was chosen for threshold segmentation of the transversely stretched film rupture image.

[0083] Step S4 above further includes the following steps:

[0084] S41. Based on the segmented image, feature extraction and identification of film rupture are performed;

[0085] S42. Based on the area and perimeter values ​​of the transverse membrane rupture, distinguish whether the membrane rupture is a small-scale rupture or a longitudinal rupture, and generate an alarm signal.

[0086] Based on the characteristics of transverse film rupture, which can be categorized into small-scale rupture and longitudinal rupture, and considering the practical requirements for rapid processing and the need for easily extractable features, the most discriminative features were selected. Therefore, based on the grayscale value characteristics of the film rupture, geometric features were chosen for measurement and analysis.

[0087] Because both small-scale and longitudinal fractures are elliptical in shape, roundness and slenderness are unsuitable as characteristic parameters for transverse tensile film fractures. The most significant difference between small-scale and longitudinal fractures is their size. For transverse tensile film fractures, the area A and perimeter P of a small-scale fracture are much smaller than those of a longitudinal fracture. Therefore, small-scale and longitudinal fractures can be distinguished based on their area and perimeter values. In the experiment, the samples were first processed, and threshold values ​​A and P for the area and perimeter of small-scale and longitudinal fractures were selected respectively. After obtaining their threshold ranges, the transverse tensile film fractures were then classified.

[0088] The shape of geometric features is usually considered as a region enclosed by a closed contour curve, and can be divided into two categories: boundary-based and region-based. Before extracting the geometric features of an image, image segmentation is required. After image segmentation, an image of a transversely stretched thin film rupture is obtained, such as... Figure 2 As shown. By Figure 2 It can be seen that there are significant differences in the area and perimeter between small-scale fractures and longitudinal fractures. Furthermore, because the fracture site is completely separated from the background after threshold segmentation, the area and perimeter are easy to calculate. Therefore, this study selects the area and perimeter of the fracture as the measurement characteristic parameters.

[0089] Specifically, step S42 above further includes the following steps:

[0090] S421. Scan the segmented image from left to right and from top to bottom, and label the unconnected scan strokes in the same row or column with different numbers;

[0091] S422. Then scan the image from top left to bottom right and from bottom left to top right respectively. If adjacent rows or columns have a continuous path, mark adjacent columns with the same number. Finally, arrange the numbers.

[0092] S423. By counting the number of pixels in the target area of ​​the image, the target area A is the sum of the number of pixels with the same sign (grayscale value "1"). The formula for calculating the area A of the transversely stretched film rupture is:

[0093] ;

[0094] In the formula, f(i,j) is the pixel input point, and R is the target region in the image;

[0095] S424. The perimeter of the target region in the image is the length of the region boundary, and the perimeter P is represented by the sum of the pixels at the edge of the target.

[0096] S425. Define an initial matrix P(x,y) = 0. Similarly, scan the image and compare the pixel values ​​of two adjacent points. If one is 1 and the other is 0, then P(x,y) = 1, where x and y represent the edge coordinates.

[0097] S426. The sum of P(x,y)=1 is the desired perimeter P.

Claims

1. A transverse tensile film rupture image intelligent monitoring system, comprising a film image module, an image processing module, and a feedback output module, characterized in that, The thin film image module is used to capture and transmit the original thin film images during uniform processing; the image processing module is used to calculate and monitor the film's rupture status and generate alarm signals; the feedback output module is used to receive alarm signals and provide audible and visual alarms; the thin film image module and the image processing module are electrically connected, and the image processing module and the feedback output module are electrically connected. The thin-film image module includes an image acquisition module, an illumination source module, and a raw image transmission module. The image acquisition module is used to acquire images of a thin film passing at a constant speed. The illumination source module is used to increase illumination during the acquisition process of the imaging device to make the acquired images clearer. The raw image transmission module is used to output the images acquired by the imaging device. The image acquisition module is electrically connected to the illumination source module and the raw image transmission module. The image processing module includes a receiving improvement module, a fine processing module, and a fracture mode determination module. The receiving improvement module receives the original thin film image acquired by the imaging device and performs grayscale processing. The fine processing module performs fine processing on the grayscale processed thin film image. The fracture mode determination module determines the fracture mode of the thin film based on the feature extraction of the finely processed thin film image and generates an alarm signal. The fine processing module includes an image preprocessing submodule, an image segmentation submodule, and a fracture feature extraction submodule. The thin film image preprocessing submodule enhances and denoises the improved thin film image. The image segmentation submodule segments the target and background and extracts useful information. The fracture feature extraction submodule extracts features based on key information about the thin film fracture. The image preprocessing submodule is electrically connected to the image segmentation submodule and the fracture feature extraction submodule. The receiving improvement module is electrically connected to the fine processing module and the fracture mode determination module.

2. The intelligent monitoring system for transverse film rupture images according to claim 1, characterized in that, When installing the imaging device used to acquire images of the thin film in the image acquisition module, the lens of the imaging device should be kept at an angle of 10° to 20° to the horizontal plane where the thin film is located.

3. The operation method of the intelligent monitoring system for transverse film rupture images according to claim 1 or 2, comprising the following steps: S1. The film is stretched laterally in the equipment for the transverse stretching process and passes at a constant speed between the imaging equipment and the lighting equipment; S2. The imaging device acquires the original image of the ruptured film and transmits the unprocessed original ruptured image to the image processing module; S3. The image processing module performs grayscale processing on the received original image of the ruptured film, and then performs fine processing on the grayscale image; S4. Analyze and determine the rupture mode of the ruptured film by extracting the rupture features, and generate an alarm signal; S5. During the film production and output process, adjust the process parameters of different types of films according to the warning of film rupture.

4. The operating method according to claim 3, characterized in that, Step S3 further includes the following steps: S31. The image processing module performs grayscale processing on the received original image of the ruptured film; S32. Linear transformation is used to enhance the grayscale image, and adaptive median filtering is further used to denoise the enhanced image; S33. Perform threshold segmentation and edge segmentation using the Otsu's method on the preprocessed image.

5. The operating method according to claim 3, characterized in that, Step S4 further includes the following steps: S41. Based on the segmented image, feature extraction and identification of film rupture are performed; S42. Based on the area and perimeter values ​​of the transverse membrane rupture, distinguish whether the membrane rupture is a small-scale rupture or a longitudinal rupture, and generate an alarm signal.

6. The operating method according to claim 3, characterized in that, Step S42 further includes the following steps: S421. Scan the segmented image from left to right and from top to bottom, and label the unconnected scan strokes in the same row or column with different numbers; S422. Then scan the image from top left to bottom right and from bottom left to top right respectively. If adjacent rows or columns have a continuous path, mark adjacent columns with the same number. Finally, arrange the numbers. S423. By counting the number of pixels in the target area of ​​the image, the target area A is the sum of the number of pixels with the same sign (grayscale value "1"). The formula for calculating the area A of the transversely stretched film rupture is: ； In the formula, f(i,j) is the pixel input point, and R is the target region in the image; S424. The perimeter of the target region in the image is the length of the region boundary, and the perimeter P is represented by the sum of the pixels at the edge of the target. S425. Define an initial matrix P(x,y) = 0. Similarly, scan the image and compare the pixel values ​​of two adjacent points. If one is 1 and the other is 0, then P(x,y) = 1, where x and y represent the edge coordinates. S426. The sum of P(x,y)=1 is the desired perimeter P.

Images