Cloth machine vision intelligent detection equipment

Abstract

The present application relates to the technical field of intelligent detection, in particular to a cloth machine vision intelligent detection equipment. It includes an intelligent camera and a non-woven fabric detection platform, the intelligent camera is used for collecting the image of the non-woven fabric surface, the non-woven fabric detection platform is used for obtaining the image and detecting the flaw on the image, and the flaw part is alarmed and marked. In the present application, the image of the non-woven fabric surface is collected by the intelligent camera, then the flaw part in the image is found, and the regularity of the pattern in the non-woven fabric is combined, and the authenticity of the flaw part is determined by comparing the shape feature and the color feature, so that the intelligent detection is realized, and the accuracy of the detection result is ensured.

Description

Technical Field

[0001] This invention relates to the field of intelligent inspection technology, and more specifically, to a machine vision intelligent inspection equipment for fabrics. Background Technology

[0002] Due to uncontrollable factors in the nonwoven fabric production process (such as equipment malfunctions, raw material contamination, or unexpected interference), defects can occur in the products, easily leading to batch-wide scrapping and potentially causing customer complaints, thus damaging the brand image. While existing technologies provide equipment for defect detection in nonwoven fabrics, certain technical challenges remain.

[0003] Chinese patent CN112102253A discloses an automated detection method and system for nonwoven fabric surface defects based on machine vision. It uses image processing technology to preprocess the acquired original images of the nonwoven fabric to improve image quality and extract feature quantities of the target image. Finally, it uses image processing technology to classify and organize the extracted feature quantities to complete the system's detection. This method can enhance the texture of the nonwoven fabric image and effectively remove interfering minimum points, making the texture clearly stand out. However, it does not consider the influence of the regular patterns of the nonwoven fabric itself and does not combine shape and color features to verify defects.

[0004] Therefore, a device capable of detecting pattern regularity is needed to verify defects, enhance detection accuracy and efficiency, and reduce the cost increase caused by rework. Summary of the Invention

[0005] The purpose of this invention is to provide a machine vision intelligent inspection equipment for fabrics to solve the problems mentioned in the background art.

[0006] To achieve the above objectives, a machine vision intelligent inspection equipment for fabrics is provided, comprising:

[0007] A smart camera, used to capture images of the surface of a nonwoven fabric;

[0008] A nonwoven fabric inspection platform is used to acquire images, detect defects in the images, and alarm and mark defective areas.

[0009] The nonwoven fabric testing platform includes at least:

[0010] Coordinate system establishment module, which is used to establish a coordinate system in the image. ;

[0011] The defect detection module is used to locate defective areas in an image and determine the feature points of the defective areas in the coordinate system. coordinate set in ;in, Represented as the first The coordinate system of the image, Represented as the first The coordinate system of the first image A set of coordinates of feature points of a defective area. Represented as the first The first set of feature point coordinates for the defective part The coordinates of the feature points;

[0012] The defect determination module verifies the identified defective parts using a preset verification method.

[0013] The processing module is used to mark the verified defective parts and send alarm information to the user;

[0014] The verification methods include pattern verification and similarity verification. The pattern verification is performed on different coordinate systems. Coordinate sets appear in To verify the patterns, and for coordinate sets without patterns... This is then identified as a defect in the entire coordinate system; for coordinate sets that appear according to a pattern... Then, a similarity verification is performed on the coordinate sets that do not meet the similarity criteria. This indicates that the defect is located in the coordinate system in question.

[0015] The similarity conditions include shape feature similarity conditions and color feature similarity conditions.

[0016] As a further improvement to this technical solution, an image library is also included, which is used to store the acquired images;

[0017] And, the coordinate system established in the image and coordinate set Store it.

[0018] As a further improvement to this technical solution, a post-processing identification module is also included. The post-processing recording module is used to identify the defective parts after marking and alarm processing, and obtain identification information.

[0019] As a further improvement to this technical solution, the image library also stores identification information, which includes positive identification information and negative identification information. Positive identification information affirms the defective part, while negative identification information negates the defective part.

[0020] The defect determination module memorizes defect locations that have negative identification information.

[0021] As a further improvement to this technical solution, the intelligent camera adopts a high-speed line scan color camera to acquire color images of the non-woven fabric in order to complete the detection of color similarity conditions.

[0022] As a further improvement to this technical solution, at least three smart cameras are provided on one side of the nonwoven fabric to meet the requirements of nonwoven fabrics with a width of 1200-1400mm and a moving speed of 58-62m / min.

[0023] As a further improvement to this technical solution, the smart camera is installed on both the front and back sides of the nonwoven fabric.

[0024] As a further improvement to this technical solution, the processing module also includes a detection report generation process, which is used to extract the defect location search status, defect location determination status, and defect location identification status from the image library.

[0025] As a further improvement to this technical solution, the types of defects include holes, oil stains, blemishes, and loose threads.

[0026] As a further improvement to this technical solution, an AI recognition module is also included. The AI ​​recognition module is used to identify the types of defective parts, specifically to identify the color and shape features of the verified defective parts.

[0027] Compared with the prior art, the beneficial effects of the present invention are as follows:

[0028] This intelligent machine vision inspection equipment for fabrics uses an intelligent camera to capture images of the nonwoven fabric surface, then locates the defective parts in the images, establishes multiple coordinate systems to determine the pattern occurrence rules in the nonwoven fabric, and uses the comparison of shape and color features to determine the authenticity of the defective parts, thereby achieving intelligent inspection while ensuring the accuracy of the inspection results. Attached Figure Description

[0029] Figure 1 This is a schematic diagram of the overall structure of the present invention;

[0030] Figure 2 This is a schematic diagram of the internal structure of the nonwoven fabric testing platform of the present invention. Detailed Implementation

[0031] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0032] Furthermore, the terms "first" and "second" are used for descriptive purposes only and should not be construed as indicating or implying relative importance or implicitly specifying the number of technical features indicated. Thus, a feature defined as "first" or "second" may explicitly or implicitly include one or more of that feature. In the description of this invention, "a plurality of" means two or more, unless otherwise explicitly specified.

[0033] Please see Figure 1 and Figure 2 As shown, the purpose of this embodiment is to provide a machine vision intelligent inspection equipment for fabrics, which includes:

[0034] Smart cameras are used to capture images of the surface of non-woven fabrics.

[0035] The non-woven fabric inspection platform is used to acquire images, detect defects in the images, and alarm and mark defective areas.

[0036] A nonwoven fabric testing platform should include at least:

[0037] The coordinate system establishment module is used to establish a coordinate system in an image. ;

[0038] The defect detection module is used to find defective areas in an image and determine the feature points of these defective areas in the coordinate system. coordinate set in ;in, Represented as the first The coordinate system of the image, Represented as the first The coordinate system of the first image A set of coordinates of feature points of a defective area. Represented as the first The first set of feature point coordinates for the defective part The coordinates of the feature points;

[0039] The defect identification module uses a preset verification method to verify the identified defective parts.

[0040] The processing module is used to mark the verified defective areas and send alarm information to the user.

[0041] Verification methods include pattern verification and similarity verification. Pattern verification is performed on different coordinate systems. Coordinate sets appear in To verify the patterns, and for coordinate sets without patterns... This is then identified as a defect in the entire coordinate system; for coordinate sets that appear according to a pattern... Then, a similarity verification is performed on the coordinate sets that do not meet the similarity criteria. This indicates that the defect is located in the coordinate system in question.

[0042] Similarity conditions include shape feature similarity conditions and color feature similarity conditions.

[0043] First, the intelligent camera acquires images of the non-woven fabric (this method can also be applied to the detection of other fabrics). The method for determining the acquisition nodes is as follows:

[0044] S1. Obtain the acquisition length l of the smart camera;

[0045] S2. Obtain the moving speed v of the nonwoven fabric;

[0046] S3. Calculate the time t taken for the nonwoven fabric to move a length l, i.e., t = l / v;

[0047] The smart camera's acquisition nodes collect images periodically with time t, enabling continuous acquisition of images of the non-woven fabric surface without image overlap. A coordinate system is then established on the images according to the acquisition sequence. superscript It is used to represent the order of images. N = L / l, where L is the total length of the nonwoven fabric, with a remainder. If we increment by 1, assuming l is 600mm and L is 2000mm, then L / l = 2000 / 600, which means there is a remainder. So we increment L by 1 to get L = 4, to ensure that the obtained image can cover the entire nonwoven fabric.

[0048] That is to say, there are four coordinate systems, namely , , and Next, the defect detection module obtains a grayscale image based on the boundary detection method of the color image, that is:

[0049] Convert the image to the CIELab color space;

[0050] The target equation is established by correlating the color difference between adjacent pixels of the target grayscale image with the CIELab color distance of the corresponding pixel.

[0051] Solve the objective equation to obtain a grayscale image that retains the maximum contrast.

[0052] Then, the grayscale threshold of the nonwoven fabric is obtained, grayscale blocks exceeding the grayscale threshold are found, and the defective areas are obtained after the boundaries are determined. The defective areas are then identified. Because nonwoven fabrics inevitably contain patterns, simply determining the defective areas based on the grayscale threshold would reduce accuracy. Therefore, this embodiment utilizes the regularity of patterns in nonwoven fabrics, thus determining the length D of the regular occurrence, and then obtaining the corresponding coordinate system based on the length D. If a defective area appears in the same position in the corresponding coordinate system (which needs to be determined by...), then... If the coordinates in the coordinate system are determined, then it can be determined that the defect appears in a regular pattern; however, if there are no defective parts that appear in a regular pattern, for example, if they appear in the same position in the corresponding coordinate system, but also appear in other non-corresponding coordinate systems, then they do not appear in a regular pattern. In this case, the defective parts appearing in all coordinate systems are real.

[0053] Specifically, for regularly occurring defects, further identification is needed, namely, determining whether the defective areas meet the conditions of similar shape and color characteristics. Color characteristics are determined based on the grayscale values ​​of the grayscale blocks, and shape characteristics are determined based on the coordinates of the boundaries (i.e.,...). The coordinates of the boundary are determined.

[0054] Specifically, it also includes an image library, which is used to store the acquired images;

[0055] And, the coordinate system established in the image and coordinate set Store it.

[0056] In addition, it includes a post-processing identification module and a post-processing recording module, which is used to identify the defective parts after marking and alarm processing, and obtain identification information. This identification requires staff to identify the defective parts and provide feedback to the identification module based on the identification results to avoid incorrect identification that could delay the production of nonwoven fabric.

[0057] Specifically, the image library also stores identification information, which includes positive identification information and negative identification information. Positive identification information affirms the defective part, while negative identification information negates the defective part.

[0058] The defect identification module memorizes defective parts with negative identification information, so that subsequent occurrences of the same defective parts will not be identified, thus ensuring the efficiency of nonwoven fabric production.

[0059] Specifically, the smart camera uses a high-speed line scan color camera to acquire color images of the nonwoven fabric, thereby completing the detection of color similarity conditions and making it suitable for materials with colored backgrounds.

[0060] Specifically, at least three smart cameras are installed on one side of the nonwoven fabric to meet the requirements of nonwoven fabrics with a width of 1200-1400mm and a moving speed of 58-62m / min.

[0061] Specifically, the smart camera is installed on both the front and back sides of the non-woven fabric.

[0062] In addition, the processing module also includes a detection report generation process, which is used to extract the blemish location search, blemish location determination, and blemish location identification from the image library.

[0063] It should be noted that the types of defects include holes, oil stains, blemishes, and loose threads.

[0064] In addition, it includes an AI recognition module, which is used to identify the types of defective parts (parts that are finally identified as real defects after verification by the defect determination module). Specifically, it identifies the color and shape characteristics of the verified defective parts. Its implementation principle is to use the characteristics of each category of defective parts in terms of color and shape to identify the types of defective parts, so as to facilitate the staff to determine the solution. In other words, the recognition results should also appear in the inspection report.

[0065] The foregoing has shown and described the basic principles, main features, and advantages of the present invention. Those skilled in the art should understand that the present invention is not limited to the above embodiments. The embodiments and descriptions in the specification are merely preferred examples and are not intended to limit the invention. Various changes and modifications can be made to the invention without departing from its spirit and scope, and all such changes and modifications fall within the scope of the present invention as claimed. The scope of protection of the present invention is defined by the appended claims and their equivalents.

Claims

1. A machine vision intelligent inspection equipment for fabrics, characterized in that, It includes: A smart camera, used to capture images of the surface of a nonwoven fabric; A nonwoven fabric inspection platform is used to acquire images, detect defects in the images, and alarm and mark defective areas. The nonwoven fabric testing platform includes at least: Coordinate system establishment module, which is used to establish a coordinate system in the image. ; The defect detection module is used to locate defective areas in an image and determine the feature points of the defective areas in the coordinate system. coordinate set in ;in, Represented as the first The coordinate system of the image, Represented as the first The coordinate system of the first image A set of coordinates of feature points of a defective area. Represented as the first The first set of feature point coordinates for the defective part The coordinates of the feature points; The defect determination module verifies the identified defective parts using a preset verification method. The processing module is used to mark the verified defective parts and send alarm information to the user; The verification methods include pattern verification and similarity verification. The pattern verification is performed on different coordinate systems. Coordinate sets appear in To verify the patterns, and for coordinate sets without patterns... This is then identified as a defect in the entire coordinate system; for coordinate sets that appear according to a pattern... Then, a similarity verification is performed on the coordinate sets that do not meet the similarity criteria. This indicates that the defect is located in the coordinate system in question. The similarity conditions include shape feature similarity conditions and color feature similarity conditions. The color features are determined based on the grayscale values ​​of the grayscale blocks, and the shape features are determined based on the coordinates of the boundaries. The AI ​​recognition module is used to identify the types of defects, specifically by recognizing the color and shape characteristics of the verified defects.

2. The intelligent machine vision inspection equipment for fabrics according to claim 1, characterized in that, It also includes an image library, which is used to store the acquired images; And, the coordinate system established in the image and coordinate set Store it.

3. The intelligent machine vision inspection equipment for fabrics according to claim 2, characterized in that, It also includes a post-processing identification module, which is used to identify the defective parts after marking and alarm processing, and obtain identification information.

4. The intelligent machine vision inspection equipment for fabrics according to claim 3, characterized in that, The image library also stores identification information, which includes positive identification information and negative identification information. Positive identification information affirms the defective part, while negative identification information negates the defective part. The defect determination module memorizes defect locations that have negative identification information.

5. The intelligent machine vision inspection equipment for fabrics according to claim 1, characterized in that, The intelligent camera is a high-speed line scan color camera, which acquires color images of the non-woven fabric to complete the detection of color similarity conditions.

6. The intelligent machine vision inspection equipment for fabrics according to claim 5, characterized in that, The intelligent camera is provided with at least three cameras on one side of the nonwoven fabric to meet the requirements of nonwoven fabric with a width of 1200-1400mm and a moving speed of 58-62m / min.

7. The intelligent machine vision inspection equipment for fabrics according to claim 6, characterized in that, The smart camera is installed on both the front and back sides of the non-woven fabric.

8. The intelligent machine vision inspection equipment for fabrics according to claim 4, characterized in that, The processing module also includes a detection report generation process, which is used to extract the defect location search status, defect location determination status, and defect location identification status from the image library.

9. The intelligent machine vision inspection equipment for fabrics according to any one of claims 1-8, characterized in that, The types of defects include holes, oil stains, blemishes, and loose threads.

Images