Image processing method and system for wood-plastic plate extrusion shaping

Abstract

The invention discloses an image processing method and system for extrusion shaping of a wood-plastic plate, and the method comprises the steps: carrying out the multi-scale analysis of an image of the wood-plastic plate, and recognizing the local features in the image; on the basis of multi-scale image feature extraction, a deformation mode in an image sequence is recognized through a time sequence image data association and dynamic correction method, and image processing parameters are adjusted in real time; according to the dynamically corrected image, processing the image by using morphological operation, and extracting a surface defect area of the wood-plastic plate; after morphological operation, identifying a defect area based on an image gradient optimization method; and on the basis of the sharpened image, analyzing the local flatness in the image by using a region growing algorithm. According to the method, the feature points are extracted from the image sequence and matched, the deformation mode between the images is recognized, and the local surface flatness is accurately analyzed in combination with the region growing algorithm, so that accurate detection of the surface defects and irregular regions of the wood-plastic plate is realized.

Description

Technical Field

[0001] The present invention relates to the technical field of image processing, and particularly to an image processing method and system for the extrusion and shaping of wood-plastic boards. Background Art

[0002] As a new type of environmentally friendly material, wood-plastic boards are widely used in fields such as construction and furniture. Especially in the production process of wood-plastic composites, the surface flatness of wood-plastic boards directly affects the appearance quality of the finished products and the subsequent processing performance. Traditional methods for detecting the surface flatness of wood-plastic boards usually rely on manual visual inspection or manual measurement. This method not only has low efficiency and poor accuracy, but is also easily affected by human factors. In addition, existing automated detection systems generally have difficulty in detecting minute defects or irregularities on the surface of wood-plastic boards in real time and accurately during the dynamic production process. Especially when the light changes and the deformation is large in the production environment, the limitations of the existing technology become particularly obvious. Therefore, how to develop an efficient, accurate and adaptable method for detecting the surface flatness of wood-plastic boards has become an urgent technical problem to be solved. Summary of the Invention

[0003] In view of the above existing problems, the present invention is proposed.

[0004] Therefore, the present invention provides an image processing method for the extrusion and shaping of wood-plastic boards, aiming to solve the problems of low efficiency, low accuracy and low adaptability existing in the existing methods for detecting the surface flatness of wood-plastic boards. Through a detection method based on image processing and region growing algorithm, the surface flatness of wood-plastic boards is analyzed in real time and accurately. Specifically, the present invention extracts feature points in the image sequence and performs matching to identify the deformation mode between images, and combines the region growing algorithm to accurately analyze the local flatness of the surface, thereby realizing the accurate detection of the defective and irregular regions on the surface of wood-plastic boards. The technical solution of the present invention can effectively cope with the problems of light change and deformation, and improve the accuracy and efficiency of detection.

[0005] To solve the above technical problems, the present invention provides the following technical solution, an image processing method for the extrusion and shaping of wood-plastic boards, including: Performing multi-scale analysis on the image of the wood-plastic board to extract image features at different scales and identify local features in the image; Based on the extraction of multi-scale image features, identifying the deformation mode in the image sequence through the method of temporal image data association and dynamic correction, and adjusting the image processing parameters in real time to adapt to the shape change of the wood-plastic board during the extrusion process; Processing the image using morphological operations according to the dynamically corrected image to extract the defective regions on the surface of the wood-plastic board; After the morphological operations, enhancing the clarity of the edge features in the image based on the image gradient optimization method to identify the defective regions; Based on the sharpened image, the regional growth algorithm is used to analyze the local flatness in the image to evaluate the overall flatness of the surface of the wood-plastic board.

[0006] As a preferred solution of the image processing method for the extrusion shaping of wood-plastic boards according to the present invention, wherein: the local features include edge features, corner features, texture features, key point features, color features, and shape features.

[0007] As a preferred solution of the image processing method for the extrusion shaping of wood-plastic boards according to the present invention, wherein: Through multi-scale decomposition, image levels from high resolution to low resolution are obtained, and the original image of the wood-plastic board is decomposed into multiple scale levels to form an image pyramid. By the image pyramid method, images with different resolutions are obtained, expressed as: ; wherein, is the original image, represents the th layer image, and the resolution of each layer of image gradually decreases, , n represents the level of the image pyramid; At each scale, the feature information of the image is extracted, and the feature information of the image includes edge, texture, and color features; It is defined that the feature set extracted at the th layer scale is : ; wherein, is the set of all features extracted from the th layer image, is the th feature extracted from the th layer image, is the number of features extracted at this scale; A weighting coefficient is assigned to the features at each scale, adjusted according to the importance of each scale, and the weighted features are fused through the following formula: ; wherein, is the integrated feature representation after fusion, is the weighting coefficient of the th layer image; The calculation formula of the weighting factor is as follows: ; wherein, represents the The standard deviation of the layer image measures the distribution degree of the feature information of the scale image, which is a constant to avoid division-by-zero errors when the standard deviation is zero.

[0008] As a preferred solution of the image processing method for wood-plastic board extrusion shaping according to the present invention, wherein: the multi-scale analysis further includes performing Fourier transform on the image of each scale , calculating the L2 norm and L1 norm of the image in the frequency domain to obtain the frequency domain weighting factor : ; wherein, represents the Fourier transform result of the -th layer image, extracting the frequency information of the image, represents the L2 norm of the Fourier transform result, reflecting the energy of the high-frequency information in the image, represents the L1 norm of the Fourier transform result, reflecting the energy of the low-frequency information in the image; The combined feature representation after scale weighting and frequency domain weighting is calculated by the following formula: ; wherein, is the final combined feature representation.

[0009] As a preferred solution of the image processing method for wood-plastic board extrusion shaping according to the present invention, wherein: identifying the local features in the image includes, At each scale level, detecting feature points through the scale space, and the feature points are located by calculating the Hessian matrix of the image. The formula is as follows: ; wherein, is the feature point response value of the -th layer image at the position , represents the Hessian matrix of the image at the position ; represents the determinant; For each feature point , extracting descriptors from the surrounding local area, that is, within the window range of size . The calculation of the descriptors is based on the gradient information within this window. The formula is as follows: ; wherein, is the local descriptor of the feature point , represents the image Gradient information at the position , and is a weighting function used to weight the contribution of each pixel in this local area; An adaptive weighting factor is introduced for each descriptor, and this factor dynamically adjusts the contribution of the descriptor based on the standard deviation of the local texture. The formula is as follows: ; where represents the standard deviation of the local window , measuring the texture complexity of the local area; represents the standard deviation; is a constant to avoid division-by-zero errors when the standard deviation is zero; By weighting the descriptors, the texture information in the image is extracted, and the texture feature is obtained through weighted summation. The formula is as follows: ; where is the extracted texture information, is the weighting factor of the feature point , and is the local descriptor of the feature point .

[0010] As a preferred solution of the image processing method for wood-plastic board extrusion shaping described in the present invention, wherein: the time-series image data association and dynamic correction method includes, for consecutive image frames and , feature points in the images are extracted and matched. Suppose the feature points extracted in the image are , and the corresponding feature points in the next image frame are . The matching error between consecutive image frames is calculated through a feature point matching algorithm. Then the error formula for feature point matching is as follows: ; where is the matching error between the point in the -th frame image and the corresponding point in the -th frame image, and respectively represent the feature point coordinates in the images and , and is the tolerance threshold. When the error is less than , the two feature points are considered to be matched; After the feature point matching is completed, analyze the deformation patterns between the feature points. Assume that the feature points in the image frame correspond to the deformation points in the image frame . Then, describe the image deformation through the following transformation model: ; where is the transformation matrix that describes the image deformation, and b is the translation vector that represents the translation part of the image; According to the identified deformation pattern, adjust the parameters in the image processing process in real time; Assume that the image processing parameters at the current moment are . Through the recognition of the deformation pattern, update the image processing parameters. The update formula is as follows: ; where are the image processing parameters at the moment of the image frame , are the updated image processing parameters at the moment of the image frame , and is the adjustment amount of the image processing parameters caused by the deformation pattern.

[0011] As a preferred solution of the image processing method for wood-plastic board extrusion and shaping according to the present invention, wherein: the dynamic correction includes, based on the image sequence, by calculating the differences between images, adjusting the image brightness and contrast parameters in real time to compensate for the illumination changes and deformations caused during the extrusion process; The morphological operations include applying erosion, dilation, opening operation, and closing operation to extract the edges of the surface defects of the wood-plastic board, highlighting the defect areas in the image, and at the same time, identifying cracks, bubbles, scratches, or surface unevenness defects in the image through the features obtained by the morphological operations, and marking the defect areas.

[0012] As a preferred solution of the image processing system for wood-plastic board extrusion and shaping according to the present invention, wherein: it includes a data acquisition module, an image recognition module, and a dynamic correction module; the data acquisition module is used to acquire data information; the image recognition module is used to process images; the dynamic correction module is used to compensate for the illumination changes and deformations that may be caused during the extrusion process based on the image sequence.

[0013] A computer device includes a memory and a processor. The memory stores a computer program, and when the processor executes the computer program, it implements the steps of the image processing method for wood-plastic board extrusion and shaping.

[0014] A computer-readable storage medium stores a computer program thereon. When the computer program is executed by a processor, it implements the steps of an image processing method for the extrusion and shaping of wood-plastic boards.

[0015] Advantages of the present invention: The wood-plastic board surface flatness detection method provided by the present invention adopts feature point matching and region growing algorithms based on image processing, and can analyze the flatness of the wood-plastic board surface in real time and accurately. By dynamically adjusting image processing parameters, this method adapts to the light changes and deformations in the production process, ensuring high-precision and stable detection. At the same time, the region growing algorithm can effectively analyze the geometric characteristics of local regions, identify and locate surface defects, greatly improving the detection efficiency and accuracy. In addition, this method has strong adaptability under different lighting conditions, solving the limitations of the prior art in complex production environments and having significant practical value. BRIEF DESCRIPTION OF THE DRAWINGS

[0016] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the following will briefly introduce the drawings required for the description of the embodiments. Obviously, the following drawings are only some embodiments of the present invention. For those of ordinary skill in the art, without creative efforts, other drawings can be obtained based on these drawings.

[0017] Figure 1 It is a schematic flowchart of an image processing method for the extrusion and shaping of wood-plastic boards provided by an embodiment of the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0018] To make the above objects, features, and advantages of the present invention more obvious and understandable, the following will give a detailed description of the specific embodiments of the present invention in conjunction with the drawings of the specification. Obviously, the described embodiments are some embodiments of the present invention, rather than all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative efforts shall fall within the scope of protection of the present invention.

[0019] Many specific details are set forth in the following description in order to fully understand the present invention. However, the present invention can also be implemented in other ways different from those described herein. Those skilled in the art can make similar generalizations without departing from the connotation of the present invention. Therefore, the present invention is not limited by the specific embodiments disclosed below.

[0020] Secondly, the so-called "one embodiment" or "embodiment" herein refers to a specific feature, structure, or characteristic that can be included in at least one implementation manner of the present invention. The phrase "in one embodiment" appearing in different places in this specification does not necessarily refer to the same embodiment, nor is it a separate or alternative embodiment that excludes other embodiments.

[0021] The present invention will be described in detail with reference to the schematic diagrams. When describing the embodiments of the present invention, for the convenience of explanation, the cross-sectional views showing the device structure will be enlarged locally in a non-general proportion, and the schematic diagrams are only examples and should not limit the scope of protection of the present invention herein. In addition, in actual production, three-dimensional spatial dimensions including length, width, and depth should be included.

[0022] Meanwhile, in the description of the present invention, it should be noted that the orientation or positional relationship indicated by terms such as "upper, lower, inner, and outer" is based on the orientation or positional relationship shown in the drawings, and is only for the convenience of describing the present invention and simplifying the description, rather than indicating or implying that the device or element referred to must have a specific orientation, be constructed and operated in a specific orientation, and therefore should not be construed as a limitation to the present invention. In addition, the terms "first, second, or third" are only used for descriptive purposes and cannot be construed as indicating or implying relative importance.

[0023] Unless otherwise clearly defined and limited in the present invention, the terms "mounted, connected, and coupled" shall be understood in a broad sense. For example, it may be a fixed connection, a detachable connection, or an integral connection; it may also be a mechanical connection, an electrical connection, or a direct connection, or may be indirectly connected through an intermediate medium, or may be the communication inside two elements. For those of ordinary skill in the art, the specific meanings of the above terms in the present invention can be understood according to specific situations.

[0024] Example 1, referring to Figure 1 , which is the first embodiment of the present invention. This embodiment provides an image processing method for the extrusion and shaping of wood-plastic boards, including: Performing multi-scale analysis on the image of the wood-plastic board to extract image features at different scales and identify local features in the image.

[0025] Based on the extraction of multi-scale image features, through the method of temporal image data association and dynamic correction, identify the deformation patterns in the image sequence, and adjust the image processing parameters in real time to adapt to the shape changes of the wood-plastic board during the extrusion process.

[0026] According to the dynamically corrected image, use morphological operations to process the image and extract the surface defect areas of the wood-plastic board.

[0027] After the morphological operations, based on the image gradient optimization method, enhance the clarity of the edge features in the image and identify the defect areas.

[0028] Based on the sharpened image, use the region growing algorithm to analyze the local flatness in the image to evaluate the overall flatness of the surface of the wood-plastic board.

[0029] The local features include edge features, corner features, texture features, key point features, color features, and shape features.

[0030] The multi-scale analysis includes obtaining image hierarchies from high resolution to low resolution through multi-scale decomposition, and decomposing the original image of the wood-plastic board into multiple scale levels to form an image pyramid. Through the image pyramid method, images with different resolutions are obtained, expressed as: ; wherein, is the original image, represents the -th layer image, and the resolution of each layer image gradually decreases, , and n represents the level of the image pyramid; At each scale, the feature information of the image is extracted. The feature information of the image includes edge, texture, and color features; Define the feature set extracted at the -th layer scale as : ; wherein, is the set of all features extracted from the -th layer image, is the -th feature extracted from the -th layer image, is the number of features extracted at this scale; A weighting coefficient is assigned to the features at each scale and adjusted according to the importance of each scale. The weighted features are fused through the following formula: ; wherein, is the integrated feature representation after fusion, is the weighting coefficient of the -th layer image; The calculation formula of the weighting factor is as follows: ; wherein, represents the standard deviation of the -th layer image, which measures the distribution degree of the feature information of the scale image, is a constant to avoid division-by-zero error when the standard deviation is zero.

[0031] The multi-scale analysis also includes performing Fourier transform on the image of each scale , calculating the L2 norm and L1 norm of the image in the frequency domain, and obtaining the frequency-domain weighting factor : ; Among them, represents the Fourier transform result of the -th layer image, extracting the frequency information of the image, represents the L2 norm of the Fourier transform result, reflecting the energy of the high-frequency information in the image, represents the L1 norm of the Fourier transform result, reflecting the energy of the low-frequency information in the image; The feature representation after combining scale weighting and frequency domain weighting is calculated by the following formula: ; Among them, is the final feature representation after synthesis.

[0032] The local features in the recognition image include, at each scale level, detecting feature points through the scale space, and the feature points are located by calculating the Hessian matrix of the image. The formula is as follows: ; Among them, is the feature point response value of the -th layer image at the position , represents the Hessian matrix (second-order partial derivative) of the image at the position ; represents the determinant; For each feature point , a descriptor is extracted from the surrounding local area, that is, within the window range of size . The calculation of the descriptor is based on the gradient information within this window. The formula is as follows: ; Among them, is the local descriptor of the feature point , represents the gradient information of the image at the position , is a weighting function used to weight the contribution of each pixel within this local area; An adaptive weighting factor is introduced for each descriptor, and this factor dynamically adjusts the contribution of the descriptor based on the standard deviation of the local texture. The formula is as follows: ; Among them, represents the standard deviation of the local window , measuring the texture complexity of the local area; represents the standard deviation; is a constant to avoid division by zero error when the standard deviation is zero; Extract the texture information in the image through weighted descriptors, and obtain the texture features through weighted summation. The formula is as follows: ; Among them, is the extracted texture information, is the feature point 's weighting factor, is the feature point 's local descriptor.

[0033] The time-series image data association and dynamic correction method includes, for consecutive image frames and , extract the feature points in the image and perform matching. Suppose the feature points extracted in the image are , and the corresponding feature points in the next image frame are . Calculate the matching error between consecutive image frames through the feature point matching algorithm. Then the error formula for feature point matching is as follows: ; Among them, is the matching error between the point in the -th frame image and the corresponding point in the -th frame image, and respectively represent the feature point coordinates in the images and , is the tolerance threshold. When the error is less than , it is considered that the two feature points match; After the feature point matching is completed, analyze the deformation mode between the feature points. Suppose the feature point in the image frame has the corresponding deformed point in the image frame . Then the deformation of the image is described by the following transformation model: ; Among them, represents the feature point coordinates in the image frame , represents the feature point coordinates in the image frame , is the transformation matrix, which describes the deformation of the image, and b is the translation vector, representing the translation part of the image; According to the identified deformation mode (through the transformation matrix and the translation vector ), and adjust the parameters in the image processing process in real time; Let the image processing parameters at the current moment be , and through the recognition of the deformation mode, update the image processing parameters. The update formula is as follows: ; where ; where, is the image processing parameter at the time of the image frame , is the updated image processing parameter at the time of the image frame , is the adjustment amount of the image processing parameter caused by the deformation mode (through M).

[0034] The dynamic correction includes, based on the image sequence, by calculating the differences between images, adjusting the image brightness and contrast parameters in real time to compensate for the illumination changes and deformations caused during the extrusion process; The morphological operations include applying erosion, dilation, opening operation, and closing operation to extract the edges of the surface defects of the wood-plastic board, highlighting the defect areas in the image, and at the same time, through the features obtained by the morphological operations, identifying cracks, bubbles, scratches, or surface unevenness defects in the image and marking the defect areas.

[0035] The image gradient optimization method includes: using an image gradient operator, namely the Sobel operator, to calculate the gradient information of each pixel in the image, and enhancing the edge features through image sharpening technology; the sharpening technology is to sharpen the image by applying a high-pass filter or a Laplace operator to enhance the edge features of the image and improve the visibility of the defect areas in the image.

[0036] The region growing algorithm is used to analyze the local flatness of the wood-plastic board surface, including: by selecting initial seed points and calculating the similarity of the pixels adjacent to these seed points, gradually expanding the region. After the region growing is completed, by analyzing the geometric characteristics of the region, evaluating whether there are flatness problems on the wood-plastic board surface and providing a basis for correction.

[0037] Furthermore, the region growing algorithm first selects one or more initial seed points in the image. These seed points are determined by the edge or texture features of the image. Usually, points located in the obvious feature regions of the surface are selected. By calculating the similarity of the adjacent pixels to these seed points, and by comparing the gray values, color values or texture features of the pixels, it is judged whether they meet the similarity criteria. Only the adjacent pixels that meet the preset conditions can be added to the current region. The region growing starts from the initial seed points and gradually expands to the pixel regions similar to them. During the expansion process, the growth of the region is controlled by setting a similarity threshold to prevent irrelevant regions from being wrongly included. After the region growing is completed, the geometric properties of each growing region, such as the area, shape, and local convexity and concavity of the region, are calculated to analyze whether there are flatness problems on the surface of the wood-plastic board. Finally, through the analysis of the geometric properties, it is determined whether there are local flatness problems and a basis is provided for subsequent correction measures.

[0038] Embodiment 2 is another embodiment of the present invention, which provides an image processing system for the extrusion and shaping of wood-plastic boards, including a data acquisition module, an image recognition module, and a dynamic correction module; the data acquisition module is used to acquire data information; the image recognition module is used to process the image; the dynamic correction module is used to compensate for the possible illumination changes and deformations during the extrusion process based on the image sequence.

[0039] Embodiment 3, the third embodiment of the present invention, is different from the previous two embodiments in that: If the said function is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on such an understanding, the technical solution of the present invention, in essence, or the part that contributes to the prior art, or a part of this technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions for causing a computer device (which may be a personal computer, a server, or a network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of the present invention. And the aforementioned storage medium includes: USB flash drives, mobile hard disks, read-only memories (ROM, Read-Only Memory), random access memories (RAM, Random Access Memory), magnetic disks or optical discs and other various media that can store program codes.

[0040] This application is described with reference to the flowcharts and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of the present application. It should be understood that each flow and / or block in the flowchart and / or block diagram can be implemented by computer program instructions, and the combination of flows and / or blocks in the flowchart and / or block diagram can also be implemented. These computer program instructions can be provided to the processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing devices to generate a machine, so that the instructions executed by the processor of the computer or other programmable data processing devices generate a device for implementing the functions specified in one or more flows Figure 1 one or more flows and / or blocks Figure 1 or one or more blocks.

[0041] These computer program instructions can also be stored in a computer-readable memory that can direct a computer or other programmable data processing device to work in a specific manner, so that the instructions stored in the computer-readable memory generate a manufactured article including an instruction device, and the instruction device implements the functions specified in one or more flows Figure 1 one or more flows and / or blocks Figure 1 or one or more blocks.

[0042] These computer program instructions can also be loaded onto a computer or other programmable data processing device, so that a series of operation steps are executed on the computer or other programmable device to generate a computer-implemented process, and thus the instructions executed on the computer or other programmable device provide steps for implementing the functions specified in one or more flows Figure 1 one or more flows and / or blocks Figure 1 or one or more blocks.

[0043] Although the preferred embodiments of the present application have been described, those skilled in the art can make additional changes and modifications once they know the basic creative concepts. Therefore, the appended claims are intended to be construed as including the preferred embodiments and all changes and modifications falling within the scope of the present application.

[0044] Obviously, those skilled in the art can make various changes and modifications to the present application without departing from the spirit and scope of the present application. Thus, if these modifications and variations of the present application fall within the scope of the claims of the present application and their equivalent technologies, the present application is also intended to include these changes and variations.

Claims

1. An image processing method for extrusion shaping of wood plastic panels, characterized in that: include, Perform multi-scale analysis on the images of wood-plastic panels to extract image features at different scales and identify local features in the images; Based on multi-scale image feature extraction, the deformation patterns in the image sequence are identified through time-series image data association and dynamic correction methods, and the image processing parameters are adjusted in real time to adapt to the shape changes of the wood-plastic board during the extrusion process; According to the dynamically corrected image, the image is processed using morphological operations to extract the surface defect area of ​​the wood-plastic board; After morphological operations, the image gradient optimization method is used to enhance the clarity of edge features in the image and identify defective areas. After image sharpening, the region growing algorithm is used to analyze the local flatness in the image to evaluate the overall flatness of the WPC surface.

2. The image processing method for extrusion shaping of wood-plastic panels according to claim 1, characterized in that: The local features include edge features, corner features, texture features, key point features, color features and shape features.

3. The image processing method for extrusion shaping of wood-plastic panels according to claim 2, characterized in that: The multi-scale analysis includes obtaining an image hierarchy from high resolution to low resolution by multi-scale decomposition, converting the original image of the wood-plastic board into Decompose into multiple scale levels to form an image pyramid. Through the image pyramid method, images of different resolutions are obtained, which are expressed as: ； in, is the original image, Indicates Layers of images, and the resolution of each layer of images gradually decreases. , n represents the level of the image pyramid; At each scale, extract the feature information of the image, which includes edge, texture and color features; Defined in The feature set extracted at the layer scale is : ; in, For the The collection of all features extracted from the layer image, For the The first layer image is extracted Features, is the number of features extracted at this scale; Assign a weighting coefficient to each feature at each scale , according to the importance adjustment of each scale, the weighted features are fused through the following formula: ； in, is the comprehensive feature representation after fusion, For the Weighting coefficients of layer images; The weighting factor is calculated as follows: ； in, Indicates The standard deviation of the layer image measures the distribution of the feature information of the scale image. is a constant that avoids division by zero when the standard deviation is zero.

4. The image processing method for extrusion shaping of wood-plastic panels according to claim 3, characterized in that: The multi-scale analysis also includes performing Fourier transform on the image at each scale. , calculate the L2 norm and L1 norm of the image in the frequency domain, and obtain the frequency domain weighting factor : ； in, Indicates The Fourier transform result of the layer image is used to extract the frequency information of the image. Represents the L2 norm of the Fourier transform result, reflecting the energy of high-frequency information in the image. It represents the L1 norm of the Fourier transform result, reflecting the energy of the low-frequency information in the image; The feature representation after combining scale weighting and frequency domain weighting is calculated by the following formula: ； in, is the final feature representation after integration.

5. The image processing method for extrusion shaping of wood-plastic panels according to claim 4, characterized in that: The local features in the identified image include: At each scale level, feature points are detected through scale space, and feature points are located by calculating the Hessian matrix of the image. The formula is as follows: ； in, For the Layer image at position The response value of the feature point at Representing images In Location The Hessian matrix at ; represents the determinant; For each feature point , from the surrounding local area, i.e., the size is The descriptor is extracted within the window range, and the calculation of the descriptor is based on the gradient information within the window. The formula is as follows: ； in, For feature points The local descriptor of Representing images In Location The gradient information at is a weighting function used to weight the contribution of each pixel in the local area; An adaptive weighting factor is introduced for each descriptor, which dynamically adjusts the contribution of the descriptor based on the standard deviation of the local texture. The formula is as follows: ； in, Represents a local window The standard deviation of , which measures the texture complexity of the local area; represents standard deviation; is a constant to avoid division by zero error when the standard deviation is zero; Through the weighted descriptor, the texture information in the image is extracted, and the texture feature is obtained by weighted summation. The formula is as follows: ； in, To extract texture information, For feature points The weighting factor of For feature points The local descriptor of .

6. The image processing method for extrusion shaping of wood-plastic panels according to claim 5, characterized in that: The time-series image data association and dynamic correction method comprises: for consecutive image frames, and , extract the feature points in the image and match them. The feature points extracted from , and in the next frame image The corresponding feature points are , the matching error between consecutive image frames is calculated by the feature point matching algorithm, and the error formula of feature point matching is as follows: ; in, It is Frame image midpoint and Corresponding points in the frame image The matching error, and Respectively represent images and The coordinates of the feature points in is the tolerance value. When the error is less than When , the two feature points are considered to match; After the feature points are matched, the deformation pattern between the feature points is analyzed, assuming that the image frame The feature points in In the image frame The corresponding deformation point in is , the deformation of the image is described by the following transformation model: ； in, is the transformation matrix, describing the deformation of the image, and b is the translation vector, representing the translation part of the image; According to the identified deformation patterns, the parameters in the image processing process are adjusted in real time; Assume that the image processing parameters at the current moment are , through the recognition of deformation mode, the image processing parameters are updated, and the update formula is as follows: ； in, In the image frame Image processing parameters at time, In the image frame The updated image processing parameters at time, is the amount of adjustment of the image processing parameters due to the deformation mode.

7. The image processing method for extrusion shaping of wood-plastic panels according to claim 6, characterized in that: The dynamic correction includes adjusting the image brightness and contrast parameters in real time by calculating the difference between the images based on the image sequence to compensate for the illumination change and deformation caused by the extrusion process; The morphological operation includes applying corrosion, expansion, opening operation and closing operation to extract the edges of surface defects of the wood-plastic board and highlight the defective area in the image. At the same time, the features obtained by the morphological operation are used to identify cracks, bubbles, scratches or surface uneven defects in the image and mark the defective area.

8. An image processing system for extrusion shaping of wood-plastic panels, applied to the image processing method for extrusion shaping of wood-plastic panels according to any one of claims 1 to 7, characterized in that: Including data acquisition module, image recognition module and dynamic correction module; The data acquisition module is used to collect data information; The image recognition module is used to process the image; The dynamic correction module is used to compensate for the illumination change and deformation that may be caused during the extrusion process based on the image sequence.

9. A computer device comprising a memory and a processor, wherein the memory stores a computer program, wherein: When the processor executes the computer program, the steps of the image processing method for extrusion shaping of wood-plastic panels according to any one of claims 1 to 7 are implemented.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the steps of the image processing method for extrusion shaping of wood-plastic panels according to any one of claims 1 to 7 are implemented.

Images