A method and system for detecting quality of wood boards produced by smart home

Abstract

The present invention relates to the field of image processing technology, and in particular to a method and system for detecting the quality of wooden boards produced in smart homes. The method comprises: extracting all texture areas in a wooden board surface image, calculating texture continuity of the texture areas, performing straight line fitting on all pixel points in the texture areas, constructing a histogram of distances from all pixel points to the fitting straight line, taking the pixel points corresponding to the peak values ​​as inflection points of the texture areas, evaluating the curvature of the texture areas based on the curvature at the inflection points, calculating the smoothness of the texture areas in combination with the curvature, the number of inflection points and texture continuity, determining the similarity between any texture area and other texture areas according to the difference in the smoothness between any texture area and other texture areas, and the ratio in width, screening out abnormal areas based on the size of the similarity, and evaluating the overall quality of the wooden board according to the number of abnormal areas. The present invention improves the accuracy of detecting crack defects on the wooden board surface.

Description

Technical Field

[0001] The present invention relates to the field of image processing technology, and in particular to a method and system for detecting the quality of wooden boards produced by smart home furnishings. Background Art

[0002] With the advancement of science and technology, the smart home market is growing rapidly. Smart home products not only improve the convenience of life, but also enhance the safety and comfort of the living environment. In this context, wood boards are an important raw material for furniture manufacturing in smart homes. The quality of wood boards directly affects the performance and service life of smart home products. The raw material of wood boards is wood, which is very sensitive to the temperature and humidity of the environment. It is easily affected by the external environment during the production process, resulting in deformation and cracking on the surface of the wood boards. These problems not only affect the appearance of the product, but also reduce its structural stability, thereby affecting the quality of the final product. Therefore, it is necessary to conduct quality inspections on the wood boards during the production process to determine whether there are crack defects on their surface.

[0003] For the detection of crack defects, the existing Chinese patent document with the authorization announcement number CN108257138B discloses a method for detecting crack features of porcelain insulators based on edge detection, including: decomposing the collected color insulator image into three channels of R, G, and B, and then using median filtering to remove the noise generated during the image acquisition and transmission process; using an improved edge detection operator to perform edge detection on the three pre-processed single-channel images respectively; using the maximum inter-class variance method to segment the image and extract the crack area of ​​the insulator. For another example, the Chinese patent application document with the application publication number CN107230202A discloses a method for automatically identifying pavement disease images, which method includes: pre-processing the photographed pavement image, including grayscale correction, Gaussian filter enhancement and local adaptive binarization; performing edge detection on the binarized image; performing connected domain contour detection on the image after edge detection to obtain the number of connected domains and the circumscribed rectangle of the connected domain contour; judging the area where the pavement cracks are located according to the shape of the circumscribed rectangle of the connected domain contour; extracting the crack area image from the image after edge detection according to the position information of the area where the pavement cracks are located.

[0004] However, as a natural material, the natural texture of the wood surface is complex and changeable. It is difficult to accurately distinguish cracks from normal textures by relying solely on threshold segmentation and edge detection methods. It is easy to misjudge normal textures as crack areas, thus affecting the accuracy of the detection results. Summary of the invention

[0005] In order to solve the problem that traditional threshold segmentation and edge detection methods are difficult to accurately distinguish cracks from normal textures, and are prone to misjudge normal textures as cracks, thereby affecting the accuracy of wood board quality detection results, the present invention proposes a smart home production wood board quality detection method and system, which specifically adopts the following scheme:

[0006] In one aspect, the present invention provides a method for detecting the quality of a smart home production wood board, comprising:

[0007] Extract all texture areas in the wood board surface image, and calculate the texture continuity of the texture area using the number of pixels and the grayscale fluctuation degree of the pixels in any texture area, wherein the texture continuity is positively correlated with the number of pixels and negatively correlated with the grayscale fluctuation degree;

[0008] Performing straight line fitting on all pixel points in the texture area, constructing a histogram of distances from all pixel points to the fitting straight line, taking the pixel point corresponding to the peak value in the histogram as the inflection point of the texture area, and evaluating the curvature of the texture area based on the curvature at the inflection point;

[0009] The smoothness of the texture region is calculated by combining the curvature of the texture region, the number of inflection points and the texture continuity, wherein the smoothness is positively correlated with the texture continuity and negatively correlated with the curvature and the number of inflection points;

[0010] Calculate the similarity between the texture region and the other texture regions according to the difference in the smoothness and the ratio in width between the texture region and the other texture regions, wherein the similarity is negatively correlated with the difference and positively correlated with the ratio;

[0011] Based on the degree of similarity, texture areas that are not similar to other texture areas are screened out as abnormal areas, and the overall quality of the wood board is evaluated according to the number of the abnormal areas.

[0012] The above technical solution avoids interference from background areas through effective texture area extraction, improves the pertinence of subsequent analysis, reflects the integrity and consistency of the texture area by evaluating the length continuity of the texture area, distinguishes normal texture areas from areas with possible defects (such as breaks and missing), the greater the length continuity, the more complete and continuous the texture is, the curvature of the texture area can be quantified through straight line fitting and distance histograms, the identification of inflection points can help locate mutation points in the texture, and further analysis of the angle change values ​​at these points can evaluate the curvature of the texture area, and can effectively identify and quantify irregular changes in the texture area. By comprehensively considering the curvature, the number of inflection points and the length continuity, a comprehensive smoothness index can be obtained, which reflects the overall smoothness of the texture area. By comparing the difference in smoothness and width ratio of different texture areas, the similarity between them is evaluated. The similarity reflects the similarity between texture areas, and can effectively identify abnormal areas that are inconsistent with most texture areas, thereby accurately locating potential quality problems, improving the accuracy of the crack defect detection results of wooden boards, and can help timely discover and eliminate wooden boards that do not meet quality standards during the production process, thereby ensuring product quality.

[0013] Furthermore, the calculation formula of the similarity is:

[0014]

[0015] in, It is The texture area and The similarity of the texture regions For the The smoothness of the texture area, For the The smoothness of the texture area, For the The width of the texture region, For the The width of the texture region, To obtain the minimum function, To obtain the maximum value function, is the absolute value symbol, is a natural exponential function.

[0016] The above technical scheme can effectively capture the similarity of texture regions in shape and size, so that the degree of similarity not only depends on the closeness of the degree of smoothness, but also takes into account the proportional relationship of the width, thereby more comprehensively reflecting the similarity between texture regions. Through the smoothing effect of the exponential function, the formula can assign a lower similarity score to a larger difference, while maintaining a higher similarity score for a smaller difference, thereby ensuring that the calculation of the similarity is more reasonable and accurate.

[0017] Furthermore, the method for extracting all texture areas in the wooden board surface image is:

[0018] The Otsu threshold algorithm is used to determine the binarization threshold of the surface image, the binarization threshold is used to distinguish the texture area and the background area of ​​the surface image, the surface image is binarized based on the binarization threshold, the pixel points greater than the binarization threshold are recorded as 0, the pixel points less than or equal to the binarization threshold are recorded as 1, the connected domain analysis is performed on all the pixel points recorded as 1, and the areas corresponding to the various connected domains in the original wooden board surface image are used as various texture areas.

[0019] The above technical solution performs connected domain analysis after binarization of the image, which can accurately identify each independent texture area and avoid interference with detection due to noise or complex background.

[0020] Furthermore, the calculation formula for the smoothness of the texture area is:

[0021]

[0022] In the formula, It is The smoothness of the texture area, It is Texture continuity of texture regions, It is The curvature of the texture area, It is The number of inflection points in a texture region, are preset weights, and maximum, It is a normalization operation.

[0023] This technical solution can quantify the smoothness of each texture area, so that the texture area with higher texture continuity, lower curvature and fewer inflection points has a higher smoothness.

[0024] Furthermore, the method for determining the abnormality degree of the texture region based on the similarity degree is:

[0025] Compute the relationship between each texture region and other The similarity of the texture regions is The mean value of similarity is taken as the value of the texture area and other The overall similarity of the texture regions is calculated, and the reciprocal of the overall similarity is normalized as the abnormality of the texture region. is the total number of texture areas;

[0026] If the abnormality of a certain texture area is greater than the preset abnormality threshold, the texture area is not similar to other texture areas and is marked as an abnormal area; if the abnormality of a certain texture area is not greater than the preset abnormality threshold, the texture area is similar to other texture areas and is marked as a normal area.

[0027] Furthermore, the calculation formula of the texture continuity of the texture area is as follows:

[0028]

[0029] In the formula, It is Texture continuity of texture regions, It is The total number of pixels in the texture area, For the The overall grayscale fluctuation degree of a texture area is represented by the variance of the normalized grayscale values ​​of all pixels in the texture area. For the The local grayscale fluctuation degree of the pixel point in the texture area is represented by the variance of the grayscale mean of the eight neighborhoods of all the pixels in the texture area. , , are preset weights, and maximum.

[0030] This formula quantifies the texture continuity of the texture area by calculating the total number of pixels, the overall grayscale fluctuation degree and the local grayscale fluctuation degree and assigning different weights. This makes the calculation of texture continuity more comprehensive and can comprehensively evaluate the continuity and consistency of texture changes in the texture area, which helps to identify those locally inconsistent or broken areas.

[0031] Further, evaluating the curvature of the texture region based on the curvature at the inflection points is to take the average of the curvatures at all inflection points in the texture region as the curvature of the texture region.

[0032] The above technical solution can effectively capture small bending changes in the texture area by adopting the average of the change angles at all inflection points as the curvature of the texture area, thereby improving the detection sensitivity of small defects.

[0033] Furthermore, the method for determining the abnormal area in the wooden board surface image is: taking the texture area whose abnormality degree is greater than the abnormality degree threshold as the abnormal area.

[0034] Furthermore, the method for evaluating the overall quality of the wooden board according to the number of abnormal areas is:

[0035] The evaluation indexes for the overall quality of the constructed wooden board are:

[0036]

[0037] In the formula, It is an evaluation index of the overall quality of the wood board. is the total number of abnormal regions, is the total number of texture areas, is the natural exponential function;

[0038] like , the quality of the board is qualified, if , the quality of the wood board is unqualified, is the preset evaluation index threshold.

[0039] Furthermore, the method for obtaining the curvature at the inflection point is: calculate the sum of the angles between the inflection point and the nearest inflection points on both sides of the inflection point, and take the ratio of the sum of the angles to 180 degrees as the curvature. The curvature at the inflection point is equal to the sum of the angles between the inflection point and the nearest inflection points on both sides of the inflection point.

[0040] On the other hand, the present invention also provides a smart home production wood board quality inspection system, the quality inspection system includes a memory and a processor, the memory stores a computer program, and the processor executes the computer program to implement the steps of any quality inspection method as described in any one of the items.

[0041] The present invention has the following effects:

[0042] The present invention can accurately distinguish normal texture from cracks through effective texture area extraction and evaluation, thereby improving the accuracy and comprehensiveness of wood board quality detection, timely eliminating unqualified products, and ensuring the quality of the final product. BRIEF DESCRIPTION OF THE DRAWINGS

[0043] By reading the following detailed description with reference to the accompanying drawings, the above and other objects, features and advantages of the exemplary embodiments of the present invention will become readily understood. In the accompanying drawings, several embodiments of the present invention are shown in an exemplary and non-limiting manner, and the same or corresponding reference numerals represent the same or corresponding parts, wherein:

[0044] Figure 1 It is a schematic flow chart of the method of the present invention. DETAILED DESCRIPTION

[0045] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative work are within the scope of protection of the present invention.

[0046] The specific embodiments of the present invention are described in detail below with reference to the accompanying drawings.

[0047] Reference Figure 1 , a method for detecting the quality of wood boards produced by smart home, comprising steps S1 to S5, which are specifically as follows:

[0048] S1: Extract all texture areas in the wooden board surface image.

[0049] Place the board on a stable surface, ensure that the lighting environment is sufficient and uniform, avoid reflections, select an appropriate angle and distance according to the actual situation, use a high-resolution camera to capture the surface image of the board, and then pre-process the surface image of the board, including:

[0050] Crop the image: remove unnecessary background parts to ensure that the image only contains the surface of the board. De-noise: use the median filter algorithm to denoise the surface image and improve the clarity of the image. Enhance contrast: apply the histogram equalization algorithm to enhance the contrast in the image and make the image details more obvious. Finally, grayscale processing is performed. Through these pre-processing steps, the image quality can be improved and the foundation for subsequent analysis can be laid.

[0051] Next, the texture area in the surface image is extracted, including:

[0052] The Otsu threshold algorithm is used to determine the binarization threshold of the surface image, the binarization threshold is used to distinguish the texture area and the background area of ​​the surface image, and the surface image is binarized based on the binarization threshold. Since the natural texture of the wooden board surface is darker than the background, the natural texture is close to black and has a small grayscale value, the pixel points greater than the binarization threshold are recorded as 0, and the pixel points less than or equal to the binarization threshold are recorded as 1 to obtain the corresponding binary image, and a connected domain analysis is performed on all the pixels recorded as 1, and the areas corresponding to these connected domains in the binary image in the surface image are used as various texture areas in the surface image.

[0053] S2: Analyze the texture continuity of the texture region.

[0054] This step mainly evaluates the texture continuity of the texture region by analyzing the grayscale distribution characteristics of the pixel points in each texture region. The texture continuity of the texture region is mainly reflected in the continuity of the length of the texture region and the continuity of the grayscale distribution of the texture region.

[0055] Among them, the continuity of length mainly reflects the overall coherence of the texture area, which is indirectly reflected by counting the total number of pixels in the texture area. The more pixels in the texture area, the more continuous the texture area is as a whole. If the texture is interrupted, the corresponding number of pixels will also decrease.

[0056] Among them, the continuity of grayscale distribution mainly includes the overall grayscale fluctuation degree and local grayscale fluctuation degree of the texture area. No matter whether the overall grayscale fluctuation degree is greater or the local grayscale fluctuation degree is greater, the continuity of grayscale distribution will be poor. It can be seen that the magnitude of grayscale fluctuation degree and the continuity of grayscale distribution present a negative correlation. Specifically, the overall grayscale fluctuation degree of the texture area is characterized by the variance of the grayscale values ​​of all pixels in the texture area. The larger the variance, the greater the grayscale fluctuation, and the worse the overall grayscale continuity of the texture area. However, there are limitations, because some texture areas appear to be continuous as a whole, but the continuity between pixels in some places is weak, or even disconnected. The grayscale values ​​of these pixels with weak continuity will have large fluctuations. This means that if only the grayscale changes of the texture area as a whole are analyzed, it is easy to ignore some discontinuous grayscale changes inside the texture area, affecting the accuracy of the grayscale distribution continuity analysis results. Therefore, the grayscale mean of the eight-neighborhood pixels of each pixel is calculated to reflect the local grayscale characteristics of the pixel. Then, the variance of the grayscale mean of the eight-neighborhood pixels of all pixels is calculated to reflect the degree of local grayscale fluctuation of all pixels in the texture area. In this way, even if there are parts with weak continuity in the texture area, it can be judged whether they are truly continuous by the changes in the grayscale values ​​in the eight-neighborhood pixels.

[0057] In order to more accurately evaluate the texture continuity of the texture area, it is necessary to consider the above two factors at the same time and obtain the calculation formula of the texture continuity of the texture area:

[0058]

[0059] In the formula, It is Texture continuity of texture regions, It is The total number of pixels in the texture area, For the The overall grayscale fluctuation degree of a texture area is expressed by the variance of the normalized grayscale values ​​of all pixels in the texture area. The larger the variance of the normalized grayscale values ​​of all pixels, the greater the overall grayscale fluctuation degree of all pixels in the texture area, and the worse the overall grayscale continuity. The smaller, For the The degree of local grayscale fluctuation in a texture area is expressed by the variance of the grayscale means of the eight neighborhoods of all pixels in the texture area. The larger the variance of the grayscale means of the eight neighborhoods of all pixels, the greater the grayscale fluctuation of all pixels in the texture area is locally, that is, the worse the grayscale continuity is locally.

[0060] In this formula, 、 、 are all preset weights. ， ， , since the texture continuity of the texture area is affected by and The joint impact of and All are analyzed from a holistic perspective. The analysis is more detailed, so Set a larger weight to highlight How much to affect the texture continuity of this texture region.

[0061] The smaller the total number of all pixels in the texture area, the worse the overall grayscale continuity of all pixels, and the worse the local grayscale continuity of all pixels, the lower the texture continuity of the texture area, and vice versa.

[0062] S3: Obtain the inflection point of the texture region and calculate the curvature of the texture region.

[0063] First, a rectangular coordinate system is constructed with the upper left corner of the surface image as the coordinate origin, the horizontal direction as the horizontal axis, and the vertical direction as the vertical axis to obtain the coordinate position of each pixel in each texture area.

[0064] Then, a straight line is fitted to all the pixels in each texture area to obtain a fitted line, and the distance from each pixel in the texture area to the fitted line is calculated. Then, a grayscale histogram is constructed with the sequence number of each pixel as the horizontal axis and the distance from each pixel to the fitted line as the vertical axis. All peaks in the grayscale histogram are found, and the pixels corresponding to all peaks are taken as the inflection points of the texture area.

[0065] Next, the curvature of the texture region is evaluated based on the curvature at the inflection point. Specifically, the average curvature at all inflection points in the texture region is used as the curvature of the texture region. The curvature at the inflection point is obtained by calculating the sum of the angles between the inflection point and the inflection points closest to both sides, and taking the ratio of the sum of the angles to 180 degrees as the curvature. The specific acquisition method is as follows:

[0066] In this texture area Take the inflection point as an example, the inflection points on both sides of it are The turning point and The turning point, The distance from the inflection point to the fitting straight line is , No. The distance from the inflection point to the fitting straight line is , No. The distance from the inflection point to the fitting straight line is , No. The turning point and The Euclidean distance between inflection points is , No. The turning point and The Euclidean distance between inflection points is .

[0067] Calculate the The curvature at the inflection point is:

[0068]

[0069] In the formula, For the The curvature at the inflection point, Indicates The turning point and The angle between the inflection points, Indicates The turning point and The sum of these two angles represents the angle between the inflection points. The angle change value at the inflection point (maximum 180 degrees), It indicates the ratio of the angle change value to the maximum angle change value. The larger the ratio is, the The greater the angle change at the inflection point, the more severe the bending at the inflection point. The greater the curvature of the inflection point, the greater the curvature, and vice versa.

[0070] S4: Calculate the smoothness of the texture region by combining the number of inflection points of the texture region, the curvature of the texture region, and the texture continuity of the texture region.

[0071] Specifically, if the texture continuity of a texture region is better, the number of inflection points is smaller, and the curvature is smaller, it indicates that the texture region has the characteristics of texture continuity, uniformity and smoothness, that is, the texture region has a higher degree of smoothness.

[0072] Among them, the calculation formula for the smoothness of the texture area is:

[0073]

[0074] In the formula, It is The smoothness of the texture area, It is Texture continuity of texture regions, It is The curvature of the texture area, It is The number of inflection points in a texture region, are preset weights, and maximum, To normalize the operation, you can use Function is normalized, maximum and minimum values ​​are normalized, etc. For example, this step is set Considering that curvature can better highlight whether the texture area is smooth and uniform, a larger weight is set for it.

[0075] S5: The similarity between texture regions is evaluated based on their smoothness and width, abnormal regions are screened out based on the similarity, and the overall quality of the board is evaluated based on the number of abnormal regions.

[0076] Specifically, the similarity between each texture region and other texture regions is calculated according to the difference in smoothness between the two texture regions and the width ratio between the two texture regions:

[0077]

[0078] In this formula, It is The texture area and The similarity of the texture regions For the The smoothness of the texture area, For the The smoothness of the texture area, For the The width of the texture region, For the The width of the texture region, To obtain the minimum function, To obtain the maximum value function, is the absolute value symbol, is a natural exponential function.

[0079] Next, based on the similarity between each texture region and all other texture regions, the texture regions that are dissimilar to other texture regions are screened out as abnormal regions. The specific method is as follows:

[0080] Compute the relationship between each texture region and other The similarity of the texture regions is similarity, this The average of the similarities between the texture area and other The overall similarity of the texture regions is calculated, and the inverse of the overall similarity is taken as the abnormality of the texture region. is the total number of texture regions, and the abnormality degree is negatively correlated with the similarity degree. For example, if there are 10 texture regions, the similarity degree between the first texture region and the other 9 texture regions is calculated, and a total of 9 similarities are obtained. The mean of these 9 normalized similarities is calculated, and the inverse of the mean is taken as the abnormality degree of the texture region.

[0081] The preset abnormality threshold is 0.6, and the abnormality degree of each texture area is compared with the threshold. If the abnormality degree of the texture area is greater than 0.6, it means that the texture area is not similar to other texture areas, and the texture area is marked as an abnormal area. If the abnormality degree of the texture area is less than or equal to 0.6, it means that the texture area is similar to other texture areas, and the texture area is marked as a normal area.

[0082] Next, the total number of abnormal areas is counted, and the evaluation index of the overall quality of the wood board is constructed as follows:

[0083]

[0084] In the formula, It is an evaluation index of the overall quality of the board. is the total number of abnormal regions, is the total number of texture areas, is a natural exponential function.

[0085] Finally, according to The size of the board is used to evaluate the quality of the board: , the quality of the board is qualified, if , the quality of the wood board is unqualified, is the preset evaluation index threshold, The empirical value is 0.7.

[0086] The present invention also provides a smart home production wood board quality inspection system, the quality inspection system includes a memory and a processor, the memory stores a computer program, and the processor executes the computer program to implement the steps of the above-mentioned quality inspection method.

[0087] In the description of this specification, "plurality" or "several" means at least two, such as two, three or more, etc., unless otherwise clearly and specifically defined.

[0088] Although this specification has shown and described a number of embodiments of the present invention, it will be apparent to those skilled in the art that such embodiments are provided by way of example only. Those skilled in the art will conceive of many modifications, changes and alternatives without departing from the ideas and spirit of the present invention. It should be understood that in the practice of the present invention, various alternatives to the embodiments of the present invention described herein may be employed.

Claims

1. A method for detecting the quality of wood boards produced by smart home, characterized in that: include: Extract all texture areas in the wood board surface image, and calculate the texture continuity of the texture area using the number of pixels and the grayscale fluctuation degree of the pixels in any texture area, wherein the texture continuity is positively correlated with the number of pixels and negatively correlated with the grayscale fluctuation degree; The overall grayscale fluctuation degree of the texture area is characterized by the variance of the grayscale values ​​of all pixels in the texture area; Performing straight line fitting on all pixel points in the texture area, constructing a histogram of distances from all pixel points to the fitting straight line, taking the pixel point corresponding to the peak value in the histogram as the inflection point of the texture area, and evaluating the curvature of the texture area based on the curvature at the inflection point; The smoothness of the texture region is calculated by combining the curvature of the texture region, the number of inflection points and the texture continuity, wherein the smoothness is positively correlated with the texture continuity and negatively correlated with the curvature and the number of inflection points; Calculate the similarity between the texture region and the other texture regions according to the difference in the smoothness and the ratio in width between the texture region and the other texture regions, wherein the similarity is negatively correlated with the difference and positively correlated with the ratio; Based on the degree of similarity, texture areas that are not similar to other texture areas are screened out as abnormal areas, and the overall quality of the wood board is evaluated according to the number of the abnormal areas.

2. The method for detecting the quality of wood boards produced by smart home according to claim 1, characterized in that: The calculation formula of the similarity is: ； in, It is The texture area and The similarity of the texture regions For the The smoothness of the texture area, For the The smoothness of the texture area, For the The width of the texture region, For the The width of the texture region, To obtain the minimum function, To obtain the maximum value function, is the absolute value symbol, is a natural exponential function.

3. The method for detecting the quality of wood boards produced by smart home according to claim 1, characterized in that: The method for extracting all texture areas in the wooden board surface image is: The Otsu threshold algorithm is used to determine the binarization threshold of the surface image, the binarization threshold is used to distinguish the texture area and the background area of ​​the surface image, the surface image is binarized based on the binarization threshold, the pixel points greater than the binarization threshold are recorded as 0, the pixel points less than or equal to the binarization threshold are recorded as 1, the connected domain analysis is performed on all the pixel points recorded as 1, and the areas corresponding to the various connected domains in the original wooden board surface image are used as various texture areas.

4. The method for detecting the quality of wood boards produced by smart home according to claim 2, characterized in that: The calculation formula for the smoothness of the texture area is: ； In the formula, It is The smoothness of the texture area, It is Texture continuity of texture regions, It is The curvature of the texture area, It is The number of inflection points in a texture region, are preset weights, and maximum, It is a normalization operation.

5. The method for detecting the quality of wood boards produced by smart home according to claim 2, characterized in that: The method of screening out texture regions that are dissimilar to other texture regions as abnormal regions is as follows: Compute the relationship between each texture region and other The similarity of the texture regions is The average of the similarities between the texture area and other The overall similarity of the texture regions is calculated, and the reciprocal of the overall similarity is normalized as the abnormality of the texture region. is the total number of texture areas; If the abnormality of a certain texture area is greater than the preset abnormality threshold, the texture area is not similar to other texture areas and is marked as an abnormal area; if the abnormality of a certain texture area is not greater than the preset abnormality threshold, the texture area is similar to other texture areas and is marked as a normal area.

6. The method for detecting the quality of wood boards produced by smart home according to claim 4, characterized in that: The calculation formula of the texture continuity of the texture area is as follows: ； In the formula, It is Texture continuity of texture regions, It is The total number of pixels in the texture area, For the The overall grayscale fluctuation degree of a texture area is represented by the variance of the normalized grayscale values ​​of all pixels in the texture area. For the The degree of local grayscale fluctuation in a texture area is represented by the variance of the grayscale mean of the eight neighborhoods of all pixels in the texture area. , and are preset weights, and maximum.

7. The method for detecting the quality of wood boards produced by smart home according to claim 4, characterized in that: Evaluating the curvature of the texture region based on the curvature at the inflection points is to take the average of the curvatures at all the inflection points in the texture region as the curvature of the texture region.

8. The method for detecting the quality of wood boards produced in smart home according to claim 5, characterized in that: The method for evaluating the overall quality of the wood board based on the number of abnormal areas is: The evaluation indexes for the overall quality of the constructed wooden board are: ； In the formula, It is an evaluation index of the overall quality of the board. is the total number of abnormal regions, is the total number of texture areas, is the natural exponential function; like , the quality of the board is qualified, if , the quality of the wood board is unqualified, is the preset evaluation index threshold.

9. The method for detecting the quality of wood boards produced in smart home according to claim 7, characterized in that: The method for obtaining the curvature at the inflection point is: calculating the sum of the angles between the inflection point and the inflection points closest to both sides thereof, and taking the ratio of the sum of the angles to 180 degrees as the curvature.

10. A smart home production wood board quality detection system, characterized in that: The quality inspection system comprises a memory and a processor, wherein a computer program is stored in the memory, and the processor executes the computer program to implement the steps of the quality inspection method according to any one of claims 1 to 9.

Images