Artificial intelligence-based plywood interface bonding quality inspection system and method

Abstract

This invention relates to the technical field of plywood production and discloses an artificial intelligence-based plywood interface bonding quality inspection system and method. Based on plywood bonding defect judgment information, plywood bonding defect shape image information, and combined with image processing tools, the system scientifically generates shape images of plywood interface bonding structural defect identifiers, achieving intelligent combination of plywood interface bonding structural defect type information and plywood interface bonding structural defect coverage shape images. Based on preprocessed image information of plywood bonding structural features, plywood bonding defect shape and location information, and plywood bonding defect identifier shape image information, combined with image processing tools, the system efficiently locates the shape and location of plywood interface bonding structural defect identifiers. Based on data fusion, the system scientifically identifies the type, shape, and location of structural defects in plywood interface bonding, improving the intelligence and applicability of plywood interface bonding quality inspection.

Description

Technical Field

[0001] This invention relates to the technical field of plywood production, specifically to an artificial intelligence-based plywood interface bonding quality inspection system and method. Background Technology

[0002] Plywood bonding is a technique that uses adhesives to bond multiple layers of thin wood veneers together to form a strong board. Early adhesives primarily used natural glues such as animal glue. With the development of the chemical industry, synthetic resins such as urea-formaldehyde resin, phenolic resin, and melamine resin have become mainstream. These adhesives offer high bond strength, good water resistance, and weather resistance, adapting to various environmental requirements. The processing flow includes veneer rotary cutting, drying, gluing, assembly, pre-pressing, and hot pressing. Hot pressing technology uses high temperature and pressure to rapidly cure the adhesive, improving production efficiency and board performance. Cold pressing is suitable for heat-sensitive materials or small-scale production. The development of automated control systems and environmentally friendly adhesives, such as formaldehyde-free adhesives, aims to reduce formaldehyde release and environmental impact, aligning with the trend of green manufacturing. Key technologies in plywood bonding processing involve adhesive formulation optimization, process parameter control (such as temperature, pressure, and time), and quality inspection to ensure the flatness, strength, and durability of the plywood. The plywood bonding process requires the detection of bonding defects at the plywood interface to ensure that the bonding meets specified requirements. However, existing plywood interface bonding quality inspection processes cannot intelligently locate the type, position, and shape of bonding defects, nor can they autonomously and efficiently present bonding defect information on the plywood surface using printing technology, thus reducing the accuracy and effectiveness of plywood bonding quality inspection.

[0003] Chinese invention patent application CN118798603A, published on October 18, 2024, discloses a method and system for managing the production of furniture panels based on big data. This method involves splitting furniture model data into individual orders; arranging panel parts on the processed panels; determining whether the processing tasks contain fixed operation instructions; and sorting the processing tasks in the target combination. However, this technical solution cannot identify and locate processing defects during the panel processing, increasing the scrap rate of the panels. Summary of the Invention

[0004] To address the issues of existing plywood interface bonding quality inspection processes failing to intelligently locate the type, position, and shape of plywood bonding defects, and lacking the ability to autonomously and efficiently present bonding defect information on the plywood surface using printing technology, thus reducing the accuracy and effectiveness of plywood bonding quality inspection, this paper aims to achieve the following: efficient judgment of plywood bonding defects; intelligent identification of the location and shape of plywood bonding defects; real-scene labeling of plywood bonding defect information; visualized intelligent quality inspection of plywood bonding operations; and improved intelligence and industrial convenience of plywood bonding quality inspection.

[0005] This invention is achieved through the following technical solution: an artificial intelligence-based quality inspection method for plywood interface bonding, the method comprising the following steps: Collect plywood bonding structure feature image information, preprocess the plywood interface bonding structure image to obtain plywood bonding structure feature preprocessed image information; detect and process structural defects in plywood interface bonding to obtain plywood bonding defect judgment information; when no defects are found, directly end this plywood bonding quality inspection operation. When defects exist, the shape image of the structural defect in the plywood interface bonding is extracted and processed to obtain the shape image information of the plywood bonding defect; the position of the structural defect in the plywood interface bonding is obtained and processed to obtain the position information of the plywood bonding defect; the shape image of the identifier of the plywood interface bonding structural defect is generated and processed to obtain the shape image information of the plywood bonding defect identifier; the position of the identifier of the plywood interface bonding structural defect is located and processed to obtain the position information of the plywood bonding defect identifier. Create plywood bonding defect identification information and perform plywood bonding defect identification operation.

[0006] Preferably, the following steps are taken: First, acquire image information of the plywood bonding structure features. Second, preprocess the image of the plywood interface bonding structure to obtain preprocessed image information of the plywood bonding structure features. Third, detect and process structural defects in the plywood interface bonding to obtain plywood bonding defect judgment information. Fourth, if no defects are found, directly end the current plywood bonding quality inspection operation. Industrial X-ray inspection equipment is used to collect and process the interface bonding structure image of the entire fiber cross section of the finished plywood product in a vertical direction, and generate plywood bonding structure feature image information. The plywood bonding structure feature image information represents the structural inspection image of the plywood product fiber cross section bonding generated according to the fiber cross section specifications. The plywood bonding structure feature image information is preprocessed by using a bilateral filter to reduce noise in the plywood interface bonding structure image, and the preprocessed plywood bonding structure feature image information is constructed. Based on the preprocessed image information of the plywood bonding structure features and the standard bonding structure feature image information matrix of plywood bonding defects, structural defect detection processing of plywood interface bonding is performed to obtain plywood bonding defect judgment information; the plywood bonding defect judgment information includes no defects and defects; when no defects are found, the current plywood bonding quality inspection operation is directly terminated.

[0007] Preferably, structural defect detection processing of plywood interface bonding is performed based on the preprocessed image information of the plywood bonding structure features and the standard bonding structure feature image information matrix of plywood bonding defects to obtain plywood bonding defect judgment information; the plywood bonding defect judgment information includes no defects and defects present; when no defects are found, the operation steps for directly ending this plywood bonding quality inspection operation are as follows: Establish a standard adhesive structure feature image information matrix for plywood bonding defects ,in Indicates the first The standard bonding structure feature image information of plywood bonding defects corresponding to various types of plywood bonding defects includes delamination, voids, bubbles, and local debonding; the standard bonding structure feature image information of plywood bonding defects represents the structural feature image of standard plywood boundary bonding set for different types of plywood bonding defects. The preprocessed image information of the plywood bonding structure features is compared with the standard bonding structure feature image information matrix of the plywood bonding defects. The image information of standard adhesive structure features of plywood bonding defects described in the text The specific steps for generating plywood bonding defect judgment information are as follows: Image feature matching of the plywood bonding structure is performed, and plywood bonding defect judgment information is generated based on the image feature matching. Step 1321, Parameter Initialization: Identifying Adhesion Defects and Bat Population Size Maximum number of iterations T, objective function Identifying individual bats by adhesive defects The standard adhesive structure feature image information matrix of the plywood adhesive defects Location in the search space , ; and speed sound wave frequency Sound wave loudness and frequency ; Step 1322: In the standard adhesive structure feature image information matrix of plywood adhesive defects The optimal location of the individual bat in the adhesive defect identification bat population was found within the search space. That is, identifying bat populations based on the standard adhesive structure feature image matrix of adhesive defects in plywood. Search the search space to find the standard adhesive structure feature image information of the plywood adhesive defects that best matches the preprocessed image information of the plywood adhesive structure features. The position is determined, and the velocity and position are updated, with the velocity and position update formulas as follows: , ,in Indicating adhesive defects to identify individual bats exist After the next iteration, the standard adhesive structure feature image information matrix of plywood bonding defects was obtained. Speed ​​in the search space Indicating adhesive defects to identify individual bats exist After the next iteration, the standard adhesive structure feature image information matrix of plywood bonding defects was obtained. Speed ​​within the search space; Indicating adhesive defects to identify individual bats exist After the next iteration, the standard adhesive structure feature image information matrix of plywood bonding defects was obtained. The location in the search space, Indicating adhesive defects to identify individual bats exist After the next iteration, the standard adhesive structure feature image information matrix of plywood bonding defects was obtained. The position in the search space; Step 1323: Generate a random number rand1, where rand1 is a random number in the interval [0, 1]. When rand1 > From the best adhesive defect identification bats, select the optimal adhesive defect identification bat, that is, the standard adhesive structure feature image information matrix of the plywood adhesive defect. Search the search space to find the standard adhesive structure feature image information of the plywood adhesive defects that best matches the preprocessed image information of the plywood adhesive structure features. The location, near the selected optimal adhesive defect identification bat individual, is determined by the formula. , , Generate a local solution, namely, the standard adhesive structure feature image information matrix of the plywood adhesive defects. Search the search space to find the standard adhesive structure feature image information of the plywood adhesive defects that best matches the preprocessed image information of the plywood adhesive structure features. ,in Indicating adhesive defects to identify individual bats exist After the next iteration, the standard adhesive structure feature image information matrix of plywood bonding defects was obtained. The loudness of sound waves in the search space. Indicating adhesive defects to identify individual bats exist After the next iteration, the standard adhesive structure feature image information matrix of plywood bonding defects was obtained. Frequency in the search space Indicating adhesive defects to identify individual bats The standard adhesive structure feature image information matrix of the plywood adhesive defects The initial frequency in the search space; otherwise, according to the formula , Update the bat position for identifying adhesive defects, that is, in the standard adhesive structure feature image information matrix of the plywood adhesive defects. Update the search space to find standard bond structure feature images of plywood bonding defects that match the preprocessed image information of the plywood bonding structure features. The location, among which Indicating adhesive defects to identify individual bats exist After the next iteration, the standard adhesive structure feature image information matrix of plywood bonding defects was obtained. The loudness of sound waves in the search space. Indicating adhesive defects to identify individual bats exist After the next iteration, the standard adhesive structure feature image information matrix of plywood bonding defects was obtained. Frequency in the search space Indicates the value random function, and It is a constant and 0 < <1, >0; Step 1324: Generate another random number rand2, which is a random number in [0, 1]; when rand2 < And at this time the objective function The fitness of the solution is better than the new solution in step 1323, that is, in the standard adhesive structure feature image information matrix of the plywood adhesive defects. Search the search space to find standard bond structure feature images of plywood bonding defects that match the preprocessed image information of the plywood bonding structure features. The fitness is better than the standard adhesive structure feature image information of the plywood adhesive defects matched in step 1323. Then accept the standard adhesive structure feature image information of the plywood bonding defects. The standard adhesive structure feature image information matrix of the plywood adhesive defects The search space is located, and the location is updated using the following formula: and according to the formula , Synchronous adjustment and ;in Indicating adhesive defects to identify individual bats The standard adhesive structure feature image information matrix of the plywood adhesive defects New location in the search space Indicating adhesive defects to identify individual bats The standard adhesive structure feature image information matrix of the plywood adhesive defects Old locations in the search space, It represents any number within the interval [-1, 1]. Step 1325: Sort the fitness values ​​of individuals in the bat population identified by the adhesion defect and find the current best. That is, the standard adhesive structure feature image information matrix of the plywood adhesive defects Search the search space to find the standard adhesive structure feature image information of the plywood adhesive defects that best matches the preprocessed image information of the plywood adhesive structure features. Location; Step 1326: Repeat steps 1322 to 1325. When the maximum number of iterations T is met, output the standard bonding structure feature image information of the plywood bonding defects that matches the preprocessed image information of the plywood bonding structure features. ; Step 1327: Based on the preprocessed image information of the plywood bonding structure features output in step 1326 and the standard bonding structure feature image information of the plywood bonding defects, The image matching results are used to generate information on plywood bonding defects. When the standard adhesive structure feature image information of the plywood adhesive defects is matched with the preprocessed image information of the plywood adhesive structure features, If the condition is not present, it means that the plywood finished product has no bonding defects. In this case, the plywood bonding defect judgment information is output as "not present", and the current plywood bonding quality inspection operation is directly ended. When the standard adhesive structure feature image information of the plywood adhesive defects is matched with the preprocessed image information of the plywood adhesive structure features, If the defect exists, it indicates that there is a bonding defect in the finished plywood product. The plywood bonding defect judgment information is then output as "existence," and simultaneously, a standard bonding structure feature image information of the plywood bonding defect that matches the preprocessed image information of the plywood bonding structure feature is output. Text information on the corresponding plywood bonding structure defect type.

[0008] Preferably, when defects exist, the following steps are taken: extracting the shape image of the structural defect in the plywood interface bonding to obtain plywood bonding defect shape image information; acquiring the location of the structural defect shape in the plywood interface bonding to obtain plywood bonding defect shape location information; generating the mark shape image of the plywood interface bonding structural defect to obtain plywood bonding defect mark shape image information; and locating the mark shape location of the plywood interface bonding structural defect to obtain plywood bonding defect mark shape location information: When the plywood bonding defect judgment information indicates the presence of a defect, the structural defect shape image of the plywood interface bonding is extracted based on the preprocessed image information of the plywood bonding structure features and the standard bonding structure feature image information matrix of the plywood bonding defects, to obtain the plywood bonding defect shape image information; the structural defect shape location of the plywood interface bonding is obtained based on the preprocessed image information of the plywood bonding structure features and the plywood bonding defect shape image information, to obtain the plywood bonding defect shape location information. Based on the plywood bonding defect judgment information and the plywood bonding defect shape image information, the plywood interface bonding structure defect identification shape image generation process is performed to obtain the plywood bonding defect identification shape image information. Based on the preprocessed image information of the plywood bonding structure features, the shape and location information of the plywood bonding defects, and the shape image information of the plywood bonding defect markers, the marker shape and location information of the plywood interface bonding structure defects is obtained through the location processing of the marker shape of the plywood bonding defects.

[0009] Preferably, when the plywood bonding defect judgment information indicates the presence of a defect, the following steps are taken to extract the shape image of the structural defect in the plywood interface bonding based on the preprocessed image information of the plywood bonding structure features and the standard bonding structure feature image information matrix of the plywood bonding defects, thereby obtaining the shape image information of the plywood bonding defect; the following steps are also taken to obtain the location information of the shape of the structural defect in the plywood interface bonding based on the preprocessed image information of the plywood bonding structure features and the shape image information of the plywood bonding defects, thereby obtaining the location information of the shape of the plywood bonding defect: The preprocessed image information of the plywood bonding structure features is compared with the standard bonding structure feature image information matrix of the plywood bonding defects using image processing tools. The image information of standard adhesive structure features of plywood bonding defects described in the text Perform image feature matching of plywood bonding structures, and extract the standard bonding structure feature image information of plywood bonding defects from the image corresponding to the preprocessed image information of the plywood bonding structure features. The matching plywood bonding defect structural feature image is used to generate plywood bonding defect shape image information after data identification; the plywood bonding defect shape image information represents the structural detection image of bonding defects in the finished plywood product; the image processing tool includes any one of OpenCV, Scikit-image, and VLFeat. Using an image processing tool, a planar coordinate system is established on the image plane corresponding to the preprocessed image information of the plywood bonding structure features based on the image features corresponding to the plywood bonding defect shape image information. The coordinate position information of the image features corresponding to the plywood bonding defect shape image information in the planar coordinate system is measured, and the plywood bonding defect shape position information is generated after data identification. The plywood bonding defect shape position information represents the position information of the plywood bonding structure defect shape on the fiber cross-section of the finished plywood.

[0010] Preferably, the steps for generating a shape image of the plywood interface bonding structure defect based on the plywood bonding defect judgment information and the plywood bonding defect shape image information are as follows: Obtain the plywood bonding defect judgment information and the plywood bonding defect shape image information; The plywood bonding defect judgment information and the corresponding text information of the plywood bonding structure defect type are mapped to the center position of the plywood bonding defect shape image information through text editing using image processing tools. This constructs a plywood bonding defect identifier shape image information, which represents a combined image of the plywood bonding structure defect shape and the plywood bonding structure defect type text information.

[0011] Preferably, the steps for locating the shape and position of the plywood interface bonding structure defects based on the preprocessed image information of the plywood bonding structure features, the shape and position information of the plywood bonding defects, and the shape image information of the plywood bonding defect markers, to obtain the shape and position information of the plywood bonding defect markers, are as follows: Using an image processing tool, the image corresponding to the shape and location information of the plywood bonding defect is located within the plywood bonding structure image corresponding to the preprocessed image information of the plywood bonding structure features. Based on the established planar coordinate system, the position coordinates of the image features corresponding to the shape and location information of the plywood bonding defect are remeasured in the planar coordinate system, and the shape and location information of the plywood bonding defect is generated. The shape and location information of the plywood bonding defect represents the position information of the shape and type of the bonding defect in the fiber cross-section of the finished plywood.

[0012] Preferably, the steps for constructing plywood bonding defect identification information and performing plywood bonding defect identification are as follows: The plywood bonding defect identification image information and the plywood bonding defect identification image location information are combined and identified to generate plywood bonding defect identification information. The plywood bonding defect identification information represents the location information of the combined image of the bonding structure defect shape and bonding defect type text characters on the fiber cross-section of the finished plywood. The visual positioning UV flatbed printer performs plywood bonding defect identification by printing images of the shape and text characters of the plywood bonding structure defect corresponding to the plywood bonding defect identification information, as well as the image position, on the cross-section of the plywood fiber.

[0013] An AI-based plywood interface bonding quality inspection system is used to implement the AI-based plywood interface bonding quality inspection method. The system includes a plywood interface bonding defect object identification module, a plywood interface bonding defect feature identification module, and a plywood interface bonding defect identification module. The plywood interface bonding defect object identification module includes a plywood bonding structure feature image acquisition unit, a plywood bonding structure feature image processing unit, a plywood bonding defect standard bonding structure feature image storage unit, and a plywood bonding defect judgment unit. The plywood bonding structure feature image acquisition unit acquires plywood bonding structure feature image information using industrial X-ray inspection equipment; the plywood bonding structure feature image processing unit performs noise reduction preprocessing on the plywood interface bonding structure image based on the plywood bonding structure feature image information, and constructs preprocessed plywood bonding structure feature image information; the plywood bonding defect standard bonding structure feature image storage unit stores plywood bonding defect standard bonding structure feature image information; the plywood bonding defect judgment unit performs structural defect detection processing on the plywood interface bonding based on the preprocessed plywood bonding structure feature image information and the standard bonding structure feature image information, and obtains plywood bonding defect judgment information. The plywood interface bonding defect feature recognition module includes a plywood bonding defect shape image extraction unit, a plywood bonding defect shape location acquisition unit, a plywood bonding defect identifier shape image generation unit, and a plywood bonding defect identifier shape location positioning unit. The plywood bonding defect shape image extraction unit extracts plywood bonding defect shape image information by preprocessing the plywood bonding structure feature image information, standard bonding structure feature image information of the plywood bonding defects, and combining image processing tools. The plywood bonding defect shape location acquisition unit obtains plywood bonding defect shape location information by preprocessing the plywood bonding structure feature image information, combining plywood bonding defect shape image information, and combining image processing tools. The marking shape image generation unit generates marking shape images of plywood bonding defects based on the plywood bonding defect judgment information, the plywood bonding defect shape image information, and in conjunction with image processing tools, to obtain plywood bonding defect marking shape image information; the plywood bonding defect marking shape position positioning unit locates marking shape images of plywood bonding defects based on the plywood bonding structure feature preprocessed image information, the plywood bonding defect shape position information, and the plywood bonding defect marking shape image information, and in conjunction with image processing tools, to obtain plywood bonding defect marking shape position information. The plywood interface bonding defect identification module includes a plywood bonding defect identification information construction unit and a plywood bonding defect identification operation execution unit. The plywood bonding defect identification information construction unit constructs plywood bonding defect identification information based on the shape image information and position information of the plywood bonding defect identification and in conjunction with data processing; the plywood bonding defect identification operation execution unit performs plywood bonding defect identification operation based on the plywood bonding defect identification information and in conjunction with a visual positioning UV flatbed printer.

[0014] This invention provides an artificial intelligence-based quality inspection system and method for plywood interface bonding. It has the following beneficial effects: I. Utilizing industrial X-ray inspection equipment, we efficiently acquire characteristic images of plywood bonding structures based on the actual specifications of the plywood bonding cross-section. Simultaneously, we perform image processing to reduce noise in the plywood interface bonding structure images, achieving accurate acquisition of plywood bonding structure information and providing reliable data support for the accurate identification of plywood bonding defect types. Based on the preprocessed image information of plywood bonding structure features, combined with AI algorithms and standard bonding structure feature image information of plywood bonding defects stored in big data, we perform intelligent detection of plywood interface bonding structure defects. This achieves efficient and high-precision detection of plywood bonding defects based on artificial intelligence algorithms, improving the quality of plywood interface bonding quality inspection.

[0015] II. Based on preprocessed image information of plywood bonding structure features and standard bonding structure feature image information of plywood bonding defects, and combined with image processing tools, the shape image of plywood interface bonding structure defects is accurately extracted to achieve accurate acquisition of the coverage shape of plywood interface bonding structure defects based on the specifications of the plywood bonding cross section; based on the preprocessed image information of plywood bonding structure features and the shape image information of plywood bonding defects, and combined with image processing tools, the location of the shape of plywood interface bonding structure defects is intelligently located to achieve accurate location of the coverage position of plywood interface bonding structure defects based on the specifications of the plywood bonding cross section; based on the plywood bonding defect judgment information and the shape of the plywood bonding defects... This method scientifically generates shape images of plywood interface bonding defects by combining image information with image processing tools. It achieves intelligent combination of plywood interface bonding defect type information and plywood interface bonding defect coverage shape image. Based on the preprocessed image information of plywood bonding structure features, plywood bonding defect shape and location information, and plywood bonding defect mark shape image information, it efficiently locates the shape and location of plywood interface bonding defect marks by combining image processing tools. Based on data fusion, it scientifically identifies the structural defect type, structural defect coverage shape, and structural defect coverage location of plywood interface bonding, improving the intelligence and applicability of plywood interface bonding quality inspection.

[0016] Third, based on the shape and location information of plywood bonding defect markers and combined with data processing, plywood bonding defect marker information is autonomously and efficiently constructed, realizing the efficient integration of plywood interface bonding structure defect type, location, and shape information; at the same time, based on the plywood bonding defect marker information and combined with a visual positioning UV flatbed printer, the plywood bonding defect marker operation is reliably executed, realizing intuitive feedback of bonding defect type, location, and shape information on the plywood bonding cross-section, improving the output and quality of plywood interface bonding quality inspection. Attached Figure Description

[0017] Figure 1 A schematic diagram of the modules of the artificial intelligence-based plywood interface bonding quality inspection system provided by the present invention; Figure 2 The flowchart shows the artificial intelligence-based quality inspection method for plywood interface bonding provided by the present invention. Detailed Implementation

[0018] 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, other embodiments obtained by those skilled in the art without creative effort are all within the scope of protection of the present invention.

[0019] An embodiment of the artificial intelligence-based plywood interface bonding quality inspection system and method is as follows: Example 1: Please refer to Figures 1-2 An artificial intelligence-based quality inspection method for plywood interface bonding includes the following steps: The process involves: acquiring structural feature images of plywood bonding; preprocessing the images to obtain preprocessed structural feature images of the plywood bonding interface; detecting structural defects in the plywood bonding interface to obtain defect judgment information; terminating the plywood bonding quality inspection operation when no defects are found; extracting the shape image of the structural defect in the plywood bonding interface to obtain shape image information of the plywood bonding defect; acquiring the location of the structural defect shape to obtain the location information of the plywood bonding defect shape; generating the shape image of the identifier of the plywood bonding structural defect to obtain the shape image information of the plywood bonding defect identifier; locating the location of the identifier shape of the plywood bonding structural defect to obtain the location information of the plywood bonding defect identifier shape; and finally, constructing the plywood bonding defect identifier information and performing the plywood bonding defect identifier operation.

[0020] For further details, please refer to Figures 1-2The process involves acquiring image information of the plywood bonding structure features, preprocessing the images of the plywood interface bonding structure to obtain preprocessed image information of the plywood bonding structure features, detecting and processing structural defects in the plywood interface bonding to obtain plywood bonding defect judgment information, and directly ending the plywood bonding quality inspection operation when no defects are found. Step 11: Using industrial X-ray inspection equipment, the entire fiber cross-section of the finished plywood product after bonding is captured and processed vertically to obtain an overall image of the interface bonding structure, and a plywood bonding structure feature image information is generated. The plywood bonding structure feature image information represents the structural inspection image of the plywood product's fiber cross-section bonding generated according to the fiber cross-section specifications. Step 12: Perform noise reduction preprocessing on the plywood bonding structure feature image information using a bilateral filter, and construct the plywood bonding structure feature preprocessed image information. Step 13: Based on the preprocessed image information of plywood bonding structure features and the standard bonding structure feature image information matrix of plywood bonding defects, perform structural defect detection processing on the plywood interface bonding to obtain plywood bonding defect judgment information; the plywood bonding defect judgment information includes no defects and defects; when no defects are found, the current plywood bonding quality inspection operation is directly terminated.

[0021] Based on the preprocessed image information of plywood bonding structure features and the standard bonding structure feature image information matrix of plywood bonding defects, structural defect detection processing of plywood interface bonding is performed to obtain plywood bonding defect judgment information. The plywood bonding defect judgment information includes whether there is no defect or whether there is a defect. When there is no defect, the operation steps to directly end this plywood bonding quality inspection are as follows: Step 131: Establish a matrix of standard adhesive structure feature images of plywood bonding defects. ,in Indicates the first The standard bonding structure feature image information of plywood bonding defects corresponding to various types of plywood bonding defects includes delamination, voids, bubbles, and local debonding; the standard bonding structure feature image information of plywood bonding defects represents the structural feature image of standard plywood boundary bonding set for different types of plywood bonding defects. Step 132: Combine the preprocessed image information of plywood bonding structure features with the standard bonding structure feature image information matrix of plywood bonding defects. Standard adhesive structure feature image information for adhesive defects in plywood The following steps are taken to generate plywood bonding defect judgment information based on the image feature matching of the plywood bonding structure: Step 1321, Parameter Initialization: Identifying Adhesion Defects and Bat Population Size Maximum number of iterations T, objective function Identifying individual bats by adhesive defects Image information matrix of standard adhesive structure features of plywood bonding defects Location in the search space , ; and speed sound wave frequency Sound wave loudness and frequency ; Step 1322: In the standard adhesive structure feature image information matrix of plywood bonding defects The optimal location of the individual bat in the adhesive defect identification bat population was found within the search space. That is, identifying adhesion defects in bat populations within the standard adhesive structure feature image matrix of plywood. Search the search space to find the standard adhesive structure feature image information of plywood adhesive defects that best matches the preprocessed image information of the plywood adhesive structure features. The position is determined, and the velocity and position are updated, with the velocity and position update formulas as follows: , ,in Indicating adhesive defects to identify individual bats exist After the second iteration, the standard adhesive structure feature image information matrix of plywood adhesive defects was obtained. Speed ​​in the search space Indicating adhesive defects to identify individual bats exist After the second iteration, the standard adhesive structure feature image information matrix of plywood adhesive defects was obtained. Speed ​​within the search space; Indicating adhesive defects to identify individual bats exist After the second iteration, the standard adhesive structure feature image information matrix of plywood adhesive defects was obtained. The location in the search space, Indicating adhesive defects to identify individual bats exist After the second iteration, the standard adhesive structure feature image information matrix of plywood adhesive defects was obtained. The position in the search space; Step 1323: Generate a random number rand1, where rand1 is a random number in the interval [0, 1]. When rand1 > The optimal adhesive defect identification bat is selected from the best adhesive defect identification bats, that is, the standard adhesive structure feature image information matrix of plywood adhesive defects. Search the search space to find the standard adhesive structure feature image information of plywood adhesive defects that best matches the preprocessed image information of the plywood adhesive structure features. The location, near the selected optimal adhesive defect identification bat individual, is determined by the formula. , , Generate a local solution, that is, in the standard adhesive structure feature image information matrix of plywood bonding defects. Search the search space to find the standard adhesive structure feature image information of plywood adhesive defects that best matches the preprocessed image information of the plywood adhesive structure features. ,in Indicating adhesive defects to identify individual bats exist After the second iteration, the standard adhesive structure feature image information matrix of plywood adhesive defects was obtained. The loudness of sound waves in the search space. Indicating adhesive defects to identify individual bats exist After the second iteration, the standard adhesive structure feature image information matrix of plywood adhesive defects was obtained. Frequency in the search space Indicating adhesive defects to identify individual bats Image information matrix of standard adhesive structure features of plywood bonding defects The initial frequency in the search space; otherwise, according to the formula , Update the bat location for identifying adhesive defects, that is, in the standard adhesive structure feature image information matrix of plywood adhesive defects. Update the search space to find standard bond structure feature images of plywood bonding defects that match the preprocessed image information of the plywood bonding structure features. The location, among which Indicating adhesive defects to identify individual bats exist After the second iteration, the standard adhesive structure feature image information matrix of plywood adhesive defects was obtained. The loudness of sound waves in the search space. Indicating adhesive defects to identify individual bats exist After the second iteration, the standard adhesive structure feature image information matrix of plywood adhesive defects was obtained. Frequency in the search space Indicates the value random function, and It is a constant and 0 < <1, >0; Step 1324: Generate another random number rand2, which is a random number in [0, 1]; when rand2 < And at this time the objective function The fitness of the solution is better than the new solution in step 1323, that is, in the standard adhesive structure feature image information matrix of plywood adhesive defects. Search the search space for standard bond structure feature images of plywood bonding defects that match the preprocessed image information of the plywood bonding structure features. The fitness is better than the standard adhesive structure feature image information of the plywood adhesive defects matched in step 1323. Then accept the standard adhesive structure feature image information of plywood bonding defects. Image information matrix of standard adhesive structure features of plywood bonding defects The search space is located, and the location is updated using the following formula: and according to the formula , Synchronous adjustment and ;in Indicating adhesive defects to identify individual bats Image information matrix of standard adhesive structure features of plywood bonding defects New location in the search space Indicating adhesive defects to identify individual bats Image information matrix of standard adhesive structure features of plywood bonding defects Old locations in the search space, It represents any number within the interval [-1, 1]. Step 1325: Sort the fitness values ​​of individuals in the bat population identified by the adhesion defect and find the current best. That is, in the standard adhesive structure feature image information matrix of plywood bonding defects Search the search space to find the standard adhesive structure feature image information of plywood adhesive defects that best matches the preprocessed image information of the plywood adhesive structure features. Location; Step 1326: Repeat steps 1322 to 1325. When the maximum number of iterations T is met, output the standard bonding structure feature image information of plywood bonding defects that matches the preprocessed image information of the plywood bonding structure features. ; Step 1327: Based on the preprocessed image information of plywood bonding structure features output in Step 1326 and the standard bonding structure feature image information of plywood bonding defects... The image matching results are used to generate information on plywood bonding defects. Standard adhesive structure feature image information of plywood adhesive defects that is matched with preprocessed image information of plywood adhesive structure features. If the condition is not present, it means that the plywood finished product has no bonding defects. In this case, the plywood bonding defect judgment information is output as "not present", and the current plywood bonding quality inspection operation is directly ended. Standard adhesive structure feature image information of plywood adhesive defects that is matched with preprocessed image information of plywood adhesive structure features. If the defect exists, it indicates that there is a bonding defect in the finished plywood product. The output will then show the plywood bonding defect as present, and simultaneously output the standard bonding structure feature image information of the plywood bonding defect that matches the preprocessed image information of the plywood bonding structure feature. Text information on the corresponding plywood bonding structure defect type.

[0022] Industrial X-ray inspection equipment efficiently acquires plywood bonding structure feature images based on the actual specifications of the plywood bonding cross-section. Simultaneously, image processing is used for noise reduction preprocessing of the plywood interface bonding structure images, enabling accurate acquisition of plywood bonding structure information and providing reliable data support for accurate identification of plywood bonding defect types. Based on the preprocessed plywood bonding structure feature image information, combined with AI algorithms and standard bonding structure feature image information of plywood bonding defects stored in big data, intelligent detection of plywood interface bonding structure defects is achieved. This enables efficient and high-precision detection of plywood bonding defects based on artificial intelligence algorithms, improving the quality of plywood interface bonding quality inspection.

[0023] For further details, please refer to Figures 1-2 When defects exist, the following steps are taken: First, the shape image of the structural defect in the plywood interface bonding is extracted to obtain the plywood bonding defect shape image information. Second, the location of the structural defect shape in the plywood interface bonding is acquired to obtain the plywood bonding defect shape location information. Third, the shape image of the marker of the plywood interface bonding structural defect is generated to obtain the plywood bonding defect marker shape image information. Finally, the location of the marker shape of the plywood interface bonding structural defect is determined to obtain the plywood bonding defect marker shape location information. Step 21: When the plywood bonding defect judgment information indicates the presence of a defect, the shape image of the structural defect in the plywood interface bonding is extracted based on the preprocessed image information of the plywood bonding structure features and the standard bonding structure feature image information matrix of the plywood bonding defects, thus obtaining the shape image information of the plywood bonding defect; the location of the shape of the structural defect in the plywood interface bonding is obtained based on the preprocessed image information of the plywood bonding structure features and the shape image information of the plywood bonding defect, thus obtaining the location information of the shape of the plywood bonding defect. Step 22: Based on the plywood bonding defect judgment information and the plywood bonding defect shape image information, perform the identification shape image generation process for the plywood interface bonding structure defect to obtain the plywood bonding defect identification shape image information. Step 23: Based on the preprocessed image information of the plywood bonding structure features, the shape and location information of the plywood bonding defects, and the shape image information of the plywood bonding defect markers, perform the mark shape and location processing of the plywood interface bonding structure defects to obtain the shape and location information of the plywood bonding defect markers.

[0024] When the plywood bonding defect assessment indicates the presence of a defect, the following steps are taken to extract the shape image of the structural defect in the plywood interface bonding based on the preprocessed image information of the plywood bonding structure features and the standard bonding structure feature image information matrix of the plywood bonding defects. The operation steps for obtaining the location information of the structural defect shape in the plywood interface bonding based on the preprocessed image information of the plywood bonding structure features and the plywood bonding defect shape image information are as follows: Step 211: Using image processing tools, preprocess the image information of the plywood bonding structure features and combine it with the standard bonding structure feature image information matrix of plywood bonding defects. Standard adhesive structure feature image information for adhesive defects in plywood Perform image feature matching of plywood bonding structures, and extract standard bonding structure feature image information of plywood bonding defects from the images corresponding to the preprocessed image information of the plywood bonding structure features. Matching structural feature images of plywood bonding defects, and generating plywood bonding defect shape image information through data identification; plywood bonding defect shape image information represents structural detection images of bonding defects in finished plywood products, and the image processing tools include any one of OpenCV, Scikit-image, and VLFeat; Step 212: Using image processing tools, establish a planar coordinate system on the image plane corresponding to the image features of the plywood bonding structure feature preprocessing image information based on the image features corresponding to the plywood bonding defect shape image information. Measure the coordinate position information of the image features corresponding to the plywood bonding defect shape image information in the planar coordinate system. After data identification, generate the plywood bonding defect shape position information, which represents the position information of the plywood bonding structure defect shape on the fiber cross-section of the finished plywood.

[0025] The steps for generating shape images of plywood interface bonding defects based on plywood bonding defect judgment information and plywood bonding defect shape image information are as follows: Step 221: Obtain information on the determination of plywood bonding defects and image information on the shape of plywood bonding defects; Step 222: Using image processing tools, the text information of the plywood bonding structure defect type corresponding to the plywood bonding defect judgment information is mapped to the center position of the plywood bonding defect shape image corresponding to the plywood bonding defect shape image information through text editing, and a plywood bonding defect identifier shape image information is constructed. The plywood bonding defect identifier shape image information represents a combined image of the plywood bonding structure defect shape and the plywood bonding structure defect type text information displayed synchronously.

[0026] The steps for locating the shape and position of plywood bonding defects based on preprocessed image information of plywood bonding structure features, shape and position information of plywood bonding defects, and shape image information of plywood bonding defect markers are as follows: Step 231: Using an image processing tool, locate the image corresponding to the shape image information of the plywood bonding defect to the plywood bonding structure image corresponding to the preprocessed image information of the plywood bonding structure features based on the image position information corresponding to the shape image information of the plywood bonding defect. Based on the established plane coordinate system, remeasure the position coordinate information of the image features corresponding to the shape image information of the plywood bonding defect in the plane coordinate system, and generate the shape position information of the plywood bonding defect. The shape position information of the plywood bonding defect indicates the position information of the shape of the plywood bonding defect structure and the text character of the bonding structure defect type on the fiber cross-section of the finished plywood.

[0027] Based on preprocessed image information of plywood bonding structure features and standard bonding structure feature image information of plywood bonding defects, and combined with image processing tools, the shape image of plywood interface bonding structure defects is accurately extracted, achieving accurate acquisition of the coverage shape of plywood interface bonding structure defects based on the plywood bonding cross-section specifications. Based on the preprocessed image information of plywood bonding structure features and the shape image information of plywood bonding defects, and combined with image processing tools, the location of the shape of plywood interface bonding structure defects is intelligently located, achieving accurate positioning of the coverage position of plywood interface bonding structure defects based on the plywood bonding cross-section specifications. Based on plywood bonding defect judgment information and the shape of the plywood bonding defects... Image information is combined with image processing tools to scientifically generate shape images of defects in plywood interface bonding structures, enabling intelligent combination of defect type information and defect coverage shape images. Based on preprocessed image information of plywood bonding structure features, defect shape and location information, and defect identification shape image information, combined with image processing tools, the defect identification shape and location of plywood interface bonding structures are efficiently located. Based on data fusion, the structural defect type, defect coverage shape, and defect coverage location of plywood interface bonding are scientifically identified, improving the intelligence and applicability of plywood interface bonding quality inspection.

[0028] For further details, please refer to Figures 1-2 The steps for constructing plywood bonding defect identification information and performing plywood bonding defect identification are as follows: Step 31: Combine the shape image information and location information of the plywood bonding defect mark to generate plywood bonding defect mark information. The plywood bonding defect mark information represents the location information of the combined image of the shape of the bonding structure defect and the text character of the bonding defect type on the fiber cross-section of the finished plywood. Step 32: Using a visual positioning UV flatbed printer, based on the plywood bonding defect identification information, the UV flatbed printer prints the plywood bonding defect shape image and the text image of the bonding defect type on the plywood fiber cross-section to perform the plywood bonding defect identification operation.

[0029] Based on the shape and location information of plywood bonding defect markers, and combined with data processing, plywood bonding defect marker information is autonomously and efficiently constructed, achieving efficient integration of the type, location, and shape information of plywood interface bonding structure defects. Simultaneously, based on the plywood bonding defect marker information and combined with a vision-positioning UV flatbed printer, the plywood bonding defect marking operation is reliably performed, enabling intuitive feedback of the type, location, and shape information of bonding defects on the plywood bonding cross-section, thereby improving the output and quality of plywood interface bonding quality inspection. Example

[0030] Please see Figures 1-2 An AI-based plywood interface bonding quality inspection system is used to implement an AI-based plywood interface bonding quality inspection method. The system includes a plywood interface bonding defect object identification module, a plywood interface bonding defect feature identification module, and a plywood interface bonding defect identification module. The plywood interface bonding defect object recognition module includes a plywood bonding structure feature image acquisition unit, a plywood bonding structure feature image processing unit, a plywood bonding defect standard bonding structure feature image storage unit, and a plywood bonding defect judgment unit. The system includes: a plywood bonding structure feature image acquisition unit, which acquires plywood bonding structure feature image information using industrial X-ray inspection equipment; a plywood bonding structure feature image processing unit, which performs noise reduction preprocessing on the plywood interface bonding structure image based on the plywood bonding structure feature image information and constructs preprocessed plywood bonding structure feature image information; a plywood bonding defect standard bonding structure feature image storage unit, which stores plywood bonding defect standard bonding structure feature image information; and a plywood bonding defect judgment unit, which performs structural defect detection processing on the plywood interface bonding based on the preprocessed plywood bonding structure feature image information and the standard bonding structure feature image information to obtain plywood bonding defect judgment information. The plywood interface bonding defect feature recognition module includes a plywood bonding defect shape image extraction unit, a plywood bonding defect shape location acquisition unit, a plywood bonding defect identifier shape image generation unit, and a plywood bonding defect identifier shape location positioning unit. The plywood bonding defect shape image extraction unit extracts plywood bonding defect shape image information by preprocessing image information of plywood bonding structure features and standard bonding structure feature image information of plywood bonding defects, and combines image processing tools to extract the shape image information of structural defects in plywood interface bonding. The plywood bonding defect shape location acquisition unit obtains plywood bonding defect shape location information by preprocessing image information of plywood bonding structure features and standard bonding structure feature image information of plywood bonding defects, and combines image processing tools to obtain the shape location information of structural defects in plywood interface bonding. The shape image generation unit generates shape image information of plywood bonding defects based on the plywood bonding defect judgment information and the shape image information of the plywood bonding defects, and combines image processing tools to generate shape image information of the plywood bonding defect identification shape. The plywood bonding defect identification shape location unit locates the shape image information of the plywood bonding defects based on the preprocessed image information of the plywood bonding structure features, the shape location information of the plywood bonding defects, and the shape image information of the plywood bonding defects, and combines image processing tools to locate the shape image information of the plywood bonding defects. The plywood interface bonding defect identification module includes a plywood bonding defect identification information construction unit and a plywood bonding defect identification operation execution unit. The plywood bonding defect identification information construction unit constructs plywood bonding defect identification information based on the shape image information and shape location information of the plywood bonding defect identification, and combines data processing; the plywood bonding defect identification operation execution unit performs plywood bonding defect identification operation based on the plywood bonding defect identification information and in conjunction with a visual positioning UV flatbed printer.

[0031] Although embodiments of the invention have been shown and described, it will be understood by those skilled in the art that various changes, modifications, substitutions and alterations can be made to these embodiments without departing from the principles and spirit of the invention, the scope of which is defined by the appended claims and their equivalents.

Claims

1. A quality inspection method for plywood interface bonding based on artificial intelligence, characterized in that, The method includes the following steps: The system collects image information of the plywood bonding structure features, preprocesses the structural images of the plywood interface bonding to obtain preprocessed image information of the plywood bonding structure features, and detects and processes structural defects in the plywood interface bonding to obtain plywood bonding defect judgment information. If no defects are found, the plywood bonding quality inspection work will be terminated directly. When defects exist, the shape image of the structural defect in the plywood interface bonding is extracted and processed to obtain the shape image information of the plywood bonding defect; the position of the structural defect in the plywood interface bonding is obtained and processed to obtain the position information of the plywood bonding defect; the shape image of the identifier of the plywood interface bonding structural defect is generated and processed to obtain the shape image information of the plywood bonding defect identifier; the position of the identifier of the plywood interface bonding structural defect is located and processed to obtain the position information of the plywood bonding defect identifier. Create plywood bonding defect identification information and perform plywood bonding defect identification operation.

2. The artificial intelligence-based quality inspection method for plywood interface bonding according to claim 1, characterized in that: The process involves acquiring structural feature images of plywood bonding, preprocessing the images to obtain preprocessed structural feature images, detecting defects in the plywood bonding interface to obtain defect judgment information, and finally ending the plywood bonding quality inspection if no defects are found. Industrial X-ray inspection equipment is used to collect and process the interface bonding structure image of the entire fiber cross-section of the finished plywood product in a vertical direction, and generate plywood bonding structure feature image information. The plywood bonding structure feature image information is preprocessed by using a bilateral filter to reduce noise in the plywood interface bonding structure image, and the preprocessed plywood bonding structure feature image information is constructed. Based on the preprocessed image information of the plywood bonding structure features and the standard bonding structure feature image information matrix of plywood bonding defects, structural defect detection processing of plywood interface bonding is performed to obtain plywood bonding defect judgment information; the plywood bonding defect judgment information includes no defects and defects; when no defects are found, the current plywood bonding quality inspection operation is directly terminated.

3. The artificial intelligence-based quality inspection method for plywood interface bonding according to claim 2, characterized in that: Based on the preprocessed image information of the plywood bonding structure features and the standard bonding structure feature image information matrix of plywood bonding defects, structural defect detection processing of plywood interface bonding is performed to obtain plywood bonding defect judgment information; the plywood bonding defect judgment information includes no defects and defects present; when no defects are found, the operation steps to directly end this plywood bonding quality inspection operation are as follows: Establish a standard adhesive structure feature image information matrix for plywood bonding defects The include ;in Indicates the first Standard adhesive structure feature image information of plywood adhesive defects corresponding to various types of adhesive structure defects; The preprocessed image information of the plywood bonding structure features is compared with the... The above The specific steps for generating plywood bonding defect judgment information are as follows: Image feature matching of the plywood bonding structure is performed, and plywood bonding defect judgment information is generated based on the image feature matching. Step 1321, Parameter Initialization: Identifying Adhesion Defects and Bat Population Size Maximum number of iterations T, objective function Identifying individual bats by adhesive defects In the Location in the search space , ; and speed sound wave frequency Sound wave loudness and frequency ; Step 1322, in the above The optimal location of the individual bat in the adhesive defect identification bat population was found within the search space. That is, the identification of bat populations by adhesion defects in the above The search space is used to find the image that best matches the preprocessed image information of the plywood bonding structure features. The location, and update the speed and location; Step 1323: Generate a random number rand1, where rand1 is a random number in the interval [0, 1]. When rand1 > Select the optimal adhesive defect identification bat from the best adhesive defect identification bat individuals, that is, in the... The search space is used to find the image that best matches the preprocessed image information of the plywood bonding structure features. The location, near the selected optimal adhesive defect identification bat individual, generates a local solution, that is, in the... The search space is used to find the image that best matches the preprocessed image information of the plywood bonding structure features. Otherwise, update the location of the identified adhesive defect bat, i.e., in the... The search space is updated to find images that match the preprocessed image information of the plywood bonding structure features. Location; Step 1324: Generate another random number rand2, which is a random number in [0, 1]; when rand2 < And at this time the objective function The fitness of the solution is better than that of the new solution in step 1323, that is, in the Search the search space to find the image that matches the preprocessed image information of the plywood bonding structure features. The fitness is better than the match in step 1323. Then accept the above. In the The location of the search space is determined and the location is updated; Step 1325: Sort the fitness values ​​of individuals in the bat population identified by the adhesion defect and find the current best. That is, in the The search space is used to find the image that best matches the preprocessed image information of the plywood bonding structure features. Location; Step 1326: Repeat steps 1322 to 1325. When the maximum number of iterations T is satisfied, output the image that matches the preprocessed image information of the plywood bonding structure features. ; Step 1327: Based on the preprocessed image information of the plywood bonding structure features output in step 1326, and the... The image matching results are used to generate information on plywood bonding defects. When the preprocessed image information matching the plywood bonding structure features is obtained... If the condition is not present, it means that the plywood finished product has no bonding defects. In this case, the plywood bonding defect judgment information is output as "not present", and the current plywood bonding quality inspection operation is directly ended. When the preprocessed image information matching the plywood bonding structure features is obtained... If the defect exists, it indicates that the plywood product has a bonding defect. The plywood bonding defect judgment information is then output as "existence," and the image information matching the preprocessed image information of the plywood bonding structure features is simultaneously output. Text information on the corresponding plywood bonding structure defect type.

4. The artificial intelligence-based plywood interface bonding quality inspection method according to claim 3, characterized in that: When defects exist, the following steps are taken: First, the shape image of the structural defect in the plywood interface bonding is extracted to obtain the plywood bonding defect shape image information. Second, the location of the structural defect shape in the plywood interface bonding is acquired to obtain the plywood bonding defect shape location information. Third, the shape image of the marker of the plywood interface bonding structural defect is generated to obtain the plywood bonding defect marker shape image information. Finally, the location of the marker shape of the plywood interface bonding structural defect is determined to obtain the plywood bonding defect marker shape location information. When the plywood bonding defect judgment information indicates the presence of a defect, the structural defect shape image of the plywood interface bonding is extracted based on the preprocessed image information of the plywood bonding structure features and the standard bonding structure feature image information matrix of the plywood bonding defects, to obtain the plywood bonding defect shape image information; the structural defect shape location of the plywood interface bonding is obtained based on the preprocessed image information of the plywood bonding structure features and the plywood bonding defect shape image information, to obtain the plywood bonding defect shape location information. Based on the plywood bonding defect judgment information and the plywood bonding defect shape image information, the plywood interface bonding structure defect identification shape image generation process is performed to obtain the plywood bonding defect identification shape image information. Based on the preprocessed image information of the plywood bonding structure features, the shape and location information of the plywood bonding defects, and the shape image information of the plywood bonding defect markers, the marker shape and location information of the plywood interface bonding structure defects is obtained through the location processing of the marker shape of the plywood bonding defects.

5. The artificial intelligence-based quality inspection method for plywood interface bonding according to claim 4, characterized in that: When the plywood bonding defect judgment information indicates the presence of a defect, the following steps are taken to extract the shape image of the structural defect in the plywood interface bonding based on the preprocessed image information of the plywood bonding structure features and the standard bonding structure feature image information matrix of the plywood bonding defects, thereby obtaining the shape image information of the plywood bonding defect; the following steps are also taken to obtain the location information of the shape of the structural defect in the plywood interface bonding based on the preprocessed image information of the plywood bonding structure features and the shape image information of the plywood bonding defects, thereby obtaining the shape location information of the plywood bonding defect: The preprocessed image information of the plywood bonding structure features is compared with the standard bonding structure feature image information matrix of the plywood bonding defects using image processing tools. The image information of standard adhesive structure features of plywood bonding defects described in the text Perform image feature matching of plywood bonding structures, and extract the standard bonding structure feature image information of plywood bonding defects from the image corresponding to the preprocessed image information of the plywood bonding structure features. Matching images of the structural features of plywood bonding defects, and generating shape image information of plywood bonding defects through data identification; Using an image processing tool, a planar coordinate system is established on the image plane corresponding to the preprocessed image information of the plywood bonding structure features based on the image features corresponding to the shape image information of the plywood bonding defect. The coordinate position information of the image features corresponding to the shape image information of the plywood bonding defect in the planar coordinate system is measured, and the plywood bonding defect shape position information is generated after data identification.

6. The artificial intelligence-based quality inspection method for plywood interface bonding according to claim 5, characterized in that: The steps for generating shape images of plywood interface bonding structure defects based on the plywood bonding defect judgment information and the plywood bonding defect shape image information are as follows: Obtain the plywood bonding defect judgment information and the plywood bonding defect shape image information; Using image processing tools, the text information of the plywood bonding defect type corresponding to the plywood bonding defect judgment information is mapped to the center position of the plywood bonding defect shape image corresponding to the plywood bonding defect shape image information through text editing, and a plywood bonding defect identification shape image information is constructed.

7. The artificial intelligence-based quality inspection method for plywood interface bonding according to claim 6, characterized in that: The steps for locating the shape and position of plywood bonding defects based on the preprocessed image information of the plywood bonding structure features, the shape and position information of the plywood bonding defects, and the shape image information of the plywood bonding defect markers are as follows: Using an image processing tool, the image corresponding to the shape image information of the plywood bonding defect is located in the plywood bonding structure image corresponding to the preprocessed image information of the plywood bonding structure features, based on the image position information corresponding to the shape image information of the plywood bonding defect. Then, based on the established planar coordinate system, the position coordinate information of the image feature corresponding to the shape image information of the plywood bonding defect is remeasured in the planar coordinate system, and the shape position information of the plywood bonding defect is generated.

8. The artificial intelligence-based quality inspection method for plywood interface bonding according to claim 7, characterized in that: The steps for generating plywood bonding defect identification information and performing plywood bonding defect identification are as follows: The plywood bonding defect identification image information and the plywood bonding defect identification image location information are combined and identified to generate plywood bonding defect identification information. The plywood bonding defect identification information represents the location information of the combined image of the bonding structure defect shape and bonding defect type text characters on the fiber cross-section of the finished plywood. The visual positioning UV flatbed printer performs plywood bonding defect identification by printing images of the shape and text characters of the plywood bonding structure defect corresponding to the plywood bonding defect identification information, as well as the image position, on the cross-section of the plywood fiber.

9. An artificial intelligence-based plywood interface bonding quality inspection system, used to implement the artificial intelligence-based plywood interface bonding quality inspection method according to any one of claims 1-8, characterized in that: The system includes a plywood interface bonding defect object identification module, a plywood interface bonding defect feature identification module, and a plywood interface bonding defect marking module. The plywood interface bonding defect object recognition module is used to collect plywood bonding structure feature image information, preprocess the plywood interface bonding structure image to obtain plywood bonding structure feature preprocessed image information, and detect and process plywood interface bonding structural defects to obtain plywood bonding defect judgment information. The plywood interface bonding defect feature recognition module is used to extract and process the shape image of the structural defect in the plywood interface bonding to obtain plywood bonding defect shape image information; to acquire and process the position of the structural defect shape in the plywood interface bonding to obtain plywood bonding defect shape position information; to generate and process the mark shape image of the plywood interface bonding structural defect to obtain plywood bonding defect mark shape image information; and to locate the position of the mark shape of the plywood interface bonding structural defect to obtain plywood bonding defect mark shape position information. The plywood interface bonding defect identification module is used to construct plywood bonding defect identification information and perform plywood bonding defect identification operations.

Images