A historical building disease intelligent diagnosis system and early warning method

Abstract

The present application belongs to the field of building disease diagnosis, and relates to a historical building disease intelligent diagnosis system and a warning method, which, on one hand, collects images of target wood in all directions, analyzes the detection results of the mold spot, erosion and softening detection units, establishes a decision matrix, and obtains the corrosion disease index by using the calculation method of approaching ideal solution, and on the other hand, collects the position and shape data of insect boring holes, analyzes and calculates the insect boring disease index, and then calculates the overall disease index of the target wood by comprehensively calculating the corrosion disease index and the insect boring disease index, and judges whether the disease grade reaches the warning grade according to the preset historical building wood structure disease grade division standard, and if it reaches, triggers the warning mechanism, comprehensively considers various disease phenomena, realizes the multi-dimensional analysis of a single disease, provides a scientific basis for subsequent diagnosis and warning, and effectively reduces the maintenance cost.

Description

Technical Field

[0001] This invention belongs to the field of building disease diagnosis, and relates to an intelligent diagnostic system and early warning method for historical building diseases. Background Technology

[0002] Historic buildings hold a significant place in the architectural field, with wooden and wood-stone structures being particularly common. However, over time and with environmental changes, wooden materials are susceptible to mold, erosion, softening, and insect infestation. These problems seriously threaten the safety and stability of historic buildings. Therefore, there is an urgent need to develop an intelligent diagnostic system and early warning methods for these defects to ensure timely and effective protection.

[0003] In the existing technology, there are also some related solutions for intelligent diagnosis of defects in wooden historical buildings. For example, Chinese Patent Publication No. CN101986150A describes a method for detecting internal defects in wooden components of ancient buildings by using a stress wave sensor and a resistance meter. This method uses a stress wave sensor and a resistance meter to detect the shape and volume of internal defects in wooden components. The diagnostic method provided improves the efficiency and accuracy of diagnosis without affecting the mechanical strength of the wooden components.

[0004] Another Chinese patent, CN115775314A, discloses an intelligent identification system for cracks in ancient wooden structures. This system automatically identifies and marks cracks in images of ancient buildings by establishing a crack detection system. The diagnostic method it provides saves manpower and improves the accuracy of identification.

[0005] While the above solutions offer some methods for intelligent diagnosis of defects in historical buildings, existing technologies still have the following limitations:

[0006] (1) It focuses on the detection and identification of single diseases, lacks the ability to analyze the comprehensive impact of multiple diseases, and cannot fully assess the health status of historical buildings.

[0007] (2) Only a single manifestation of a disease was analyzed, without comprehensively considering multiple manifestations. For example, when conducting internal defect detection, only the size of the internal defect was considered, while the influence of the defect location on the severity of the disease was ignored. When conducting corrosion detection, the combined effects of other external factors such as mold, softening, and erosion were not fully considered, especially the softening factor, which made it difficult to accurately present the actual severity of the disease. Summary of the Invention

[0008] To address the problems mentioned in the background section, an intelligent diagnostic system and early warning method for historical building defects are proposed.

[0009] The technical solution adopted by the present invention to solve its technical problem is as follows: Firstly, the present invention provides an intelligent diagnostic system for historical building diseases, including: a timber sampling module, a corrosion analysis module, an insect infestation analysis module, a disease diagnosis module, and a disease early warning module.

[0010] The wood sampling module is connected to the corrosion analysis module and the insect infestation analysis module, respectively. The corrosion analysis module is connected to the disease diagnosis module, the insect infestation analysis module is connected to the disease diagnosis module, and the disease diagnosis module is connected to the disease early warning module.

[0011] The timber sampling module is used to select a specific wooden structural component from a historical building as the target timber.

[0012] The corrosion analysis module is used to acquire images of the target wood from all angles, import the corresponding detection results from the mold detection unit, erosion detection unit, and softening detection unit, and obtain the corrosion disease index by establishing a decision matrix and using a calculation method that approximates the ideal solution.

[0013] The insect infestation analysis module is used to collect the location and shape of insect holes using a microwave detector, analyze the distance and volume of the borer holes inside the target wood, and obtain the insect infestation disease index by combining the corresponding preset influencing factors.

[0014] The disease diagnosis module is used to perform linear weighted fusion of corrosion disease index and insect infestation disease index to obtain the overall disease index of the target wood.

[0015] The disease early warning module is used to determine whether the disease level of the target timber reaches the warning level based on the preset classification standard for disease levels of wooden structures in historical buildings and the overall disease index of the target timber. If it does, the early warning mechanism is triggered.

[0016] Secondly, the present invention provides an intelligent diagnosis and early warning method for defects in historical buildings, comprising: S1. Selecting a wooden structural component in a historical building as the target timber.

[0017] S2. Perform omnidirectional image acquisition on the target wood, import the corresponding detection results from the mold detection unit, erosion detection unit and softening detection unit respectively, and obtain the corrosion disease index by establishing a decision matrix and using the approximation ideal solution calculation method.

[0018] S3. Collect the location and shape of insect-infested holes using a microwave detector, analyze the distance and volume of the borer holes inside the target wood, and obtain the insect-infestation disease index by combining the corresponding preset influencing factors.

[0019] S4. Perform linear weighted fusion of the corrosion disease index and the insect infestation disease index to obtain the overall disease index of the target wood.

[0020] S5. Based on the preset classification standard for the disease level of wooden structures in historical buildings, and combined with the overall disease index of the target timber, determine whether its disease level reaches the warning level. If it does, trigger the warning mechanism.

[0021] Compared with the prior art, the beneficial effects of the present invention are as follows:

[0022] (1) This invention can capture mold spots, erosion, softening on the surface of wood and insect infestation inside the wood by all-round image acquisition and microwave detection, providing more comprehensive disease information.

[0023] (2) This invention relies on the corrosion disease module, introduces mold, erosion and softening detection units, conducts comprehensive analysis of the detection data of each unit, and then obtains the corrosion disease index, accurately captures potential corrosion risks, comprehensively considers a variety of disease phenomena, realizes multi-dimensional analysis of a single disease, provides a scientific basis for subsequent diagnosis and early warning, and effectively reduces maintenance costs.

[0024] (3) This invention uses the insect-boring disease module to accurately obtain the location and shape data of insect-boring holes, and deeply calculates and analyzes the distance and volume data of the borer core, thereby obtaining the insect-boring disease index, which effectively improves the accuracy of quantifying diseases and improves the efficiency and quality of historical building protection work.

[0025] (4) This invention obtains the comprehensive disease index of historical buildings by combining corrosion and insect infestation disease indices, and establishes an early warning mechanism based on the preset classification standard of disease levels of wooden structures of historical buildings. It realizes the comprehensive assessment of the health status of historical buildings by combining multiple disease factors, effectively improves the intelligent level of disease assessment and early warning, and buys valuable time for the protection and restoration of historical buildings. Attached Figure Description

[0026] To more clearly illustrate the technical solutions of the embodiments of the present invention, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0027] Figure 1 This is a schematic diagram of the module connection of the present invention.

[0028] Figure 2 This is a flowchart illustrating the implementation steps of the method of the present invention.

[0029] Figure 3 This is a flowchart illustrating the logic of obtaining the insect infestation index of the target wood in the insect infestation analysis module of the present invention. Detailed Implementation

[0030] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0031] Please see Figure 1 As shown, the first aspect of the present invention provides an intelligent diagnostic system for historical building diseases, including: a timber sampling module, a corrosion analysis module, an insect infestation analysis module, a disease diagnosis module, and a disease early warning module.

[0032] The wood sampling module is connected to the corrosion analysis module and the insect infestation analysis module, respectively. The corrosion analysis module is connected to the disease diagnosis module, the insect infestation analysis module is connected to the disease diagnosis module, and the disease diagnosis module is connected to the disease early warning module.

[0033] The timber sampling module is used to select a specific wooden structural component from a historical building as the target timber.

[0034] The corrosion analysis module is used to acquire images of the target wood from all angles, import the corresponding detection results from the mold detection unit, erosion detection unit and softening detection unit, and obtain the corrosion disease index by establishing a decision matrix and using a method to approximate the ideal solution.

[0035] As a preferred embodiment, the mold detection unit is used to identify several mold areas in the scanned images of the target wood from various angles, and to obtain the mold detection results of the target wood from various angles based on the basic features of each mold area and the aggregation features between mold areas.

[0036] It should be noted that the process of identifying several moldy areas in the directional scan images of the target wood includes: converting the directional scan image of the target wood into the HSL space, and collecting the color feature set of each pixel in the directional scan image. The color feature set includes brightness, saturation and hue.

[0037] Extract preset color feature sets of various mold spots on wood stored in the cloud database, including the standard brightness, saturation, and hue data range of mold spots. Compare the color feature sets of each pixel in the azimuth scan image with the preset color feature sets of various mold spots on wood. Calculate the deviation values ​​of the color feature sets of each pixel in the azimuth scan image and the preset color feature sets of various mold spots on wood in terms of brightness, saturation, and hue using the Euclidean distance formula. Sum the deviation values ​​of brightness, saturation, and hue and substitute them into a standard decay exponential function to obtain the comprehensive similarity between the color feature sets of each pixel in the azimuth scan image and the preset color feature sets of various mold spots on wood. Select the pixel with the highest comprehensive similarity for the corresponding mold spot type. If it exceeds the preset comprehensive similarity threshold, the pixel is selected as a pre-selected mold spot pixel. Obtain the connected components constructed by each pre-selected mold spot pixel in the azimuth scan image and record the connected components as pre-selected mold spot regions, thereby obtaining each pre-selected mold spot region in the azimuth scan image.

[0038] Based on the mold spot shape feature set stored in the cloud database, the maximum matching similarity between each pre-selected mold spot region and the mold spot shape feature set in the azimuth scan image is obtained by the shape matching tool in the image processing software. If the maximum matching similarity exceeds the preset matching similarity threshold, the pre-selected mold spot region is taken as the mold spot region, thereby filtering several mold spot regions in the azimuth scan images of the target wood.

[0039] As a preferred option, the specific process for obtaining the mold detection results of the target wood in each direction includes: obtaining the area ratio of each mold region in the target wood directional scanning image.

[0040] It should be noted that the process of obtaining the area ratio of each moldy area includes: counting the number of pixels in each moldy area, and calculating the ratio with the total number of pixels in the azimuth image to obtain the area ratio of each moldy area.

[0041] The scanned images of the target wood from all directions are converted to the HSL color space. The brightness, saturation, and hue of the normal area of ​​the target wood are used as standard reference colors. The brightness, saturation, and hue of each moldy area are measured and compared with the standard reference colors to obtain the color parameter deviation values ​​of each moldy area.

[0042] The mold patch region is extracted by image segmentation, and the edge contour of the mold patch region is identified. The edge density is obtained by calculating the ratio of the number of pixels of the edge contour of the mold patch region to its total number of pixels. The reciprocal of the edge density is used as the continuous value of the mold patch region, and the continuous value of each mold patch region is obtained in this way.

[0043] By integrating area proportion, color parameter deviation value, and continuous value, basic feature values ​​of each mold patch area in the directional area are generated, and the maximum value is extracted as the comprehensive basic feature value of the mold patch area in the directional area.

[0044] It should be noted that the process of obtaining the color parameter deviation values ​​of each moldy area includes: measuring the brightness, saturation, and hue data of the normal area and each moldy area using the color picker tool in the image software; calculating the absolute difference between the brightness of each moldy area and the brightness of the normal area and performing a ratio calculation with the brightness of the normal area to obtain the brightness deviation ratio of each moldy area; similarly, obtaining the saturation deviation ratio and hue deviation ratio of each moldy area; and summing the brightness, saturation, and hue deviation ratios of each moldy area to obtain the color parameter deviation value of each moldy area.

[0045] It should be noted that the process of obtaining the edge contour of the moldy area includes: converting the orientation scan image of the target wood into a grayscale image using image software, and performing Gaussian filtering on the grayscale image to obtain a grayscale image with noise removed.

[0046] By performing a convolution operation on the image using the Laplacian operator, a Laplacian response map is obtained. Each pixel in the Laplacian response map is analyzed, and points with different signs of positive and negative values ​​from their neighboring pixels are identified as zero-crossing points. Erosion processing is then applied to the detected zero-crossing points to obtain the edge contour of the mold patch area.

[0047] It should be noted that the process of integrating the area proportion, color parameter deviation value, and continuous value to generate the basic feature value of each mold patch area in the orientation is mainly obtained through linear weighting. The process includes multiplying the area proportion, color parameter deviation value, and continuous value of each mold patch area by a preset weight and accumulating them to obtain the basic feature value of each mold patch area.

[0048] It should be noted that the preset basic weights for area ratio, color parameter deviation value, and continuous value can be exemplarily 0.5, 0.3, and 0.2. This is based on the fact that the area ratio directly reflects the degree of mold spread and is a core indicator for assessing the severity of wood damage, thus having a higher weight. Color difference can help determine the degree of mold corrosion, but it needs to be combined with the area ratio for comprehensive analysis, so its weight is lower than that of the area ratio. Continuous value reflects the morphology of mold growth and distribution, but it has a smaller impact on the overall assessment and has the lowest weight.

[0049] The preset weights can be set based on industry experience or obtained through a limited number of experimental data. For example, first collect historical data on the area proportion, color parameter deviation, and continuous values ​​of the moldy area, then calculate the influence correlation coefficients of the area proportion, color parameter deviation, and continuous values ​​of different moldy areas, and use regression analysis or logistic regression analysis to determine their contribution to the basic characteristic values ​​of each moldy area. Finally, after normalization, the contribution is converted into preset weights and their sum is 1.

[0050] As a preferred option, the specific acquisition process of the mold spot detection results of the target wood in each direction further includes: establishing a pixel coordinate system of the azimuth scan image, determining the horizontal and vertical coordinates of the center of each mold spot area, calculating the distance from the center of any mold spot area to the center of all other mold spot areas, selecting the minimum value as the nearest neighbor distance, and calculating the average of the nearest neighbor distances of all mold spots as the clustering feature value of the azimuth mold spot area.

[0051] It should be noted that the specific process of determining the horizontal and vertical coordinates of the center of each moldy area includes: summing the horizontal and vertical coordinates of each moldy area and calculating the ratio with the total number of pixels in each moldy area to obtain the horizontal and vertical coordinates of the center of each moldy area.

[0052] Based on the comprehensive basic feature value and clustered feature value of the directional mold patch area, the directional mold patch area detection value is calculated by the geometric mean formula, and the mold patch area detection value of each direction is obtained as the mold patch detection result of each direction of the target wood.

[0053] The erosion detection unit is used to analyze the scanned images of the target wood from all directions and the design images of the target wood stored in the cloud, to obtain the texture feature data of the scanned images and design images of the target wood from all directions, and to compare and analyze the two to obtain the erosion detection results of the target wood from all directions.

[0054] As a preferred option, the specific process for obtaining the erosion detection results of the target wood in each direction includes: constructing a gray-scale co-occurrence matrix by analyzing the directional scanning images of the target wood.

[0055] The texture feature dataset, including contrast, entropy, and energy, is obtained from the gray-level co-occurrence matrix.

[0056] Based on the target timber design image recorded in the historical architectural design drawings stored in the cloud database, a reference texture feature dataset of the target timber design image is obtained similarly. The absolute deviation between the target timber design image and the target timber texture feature dataset is calculated, and then the ratio is calculated with the corresponding parameters of the reference texture feature dataset to obtain the percentage of absolute deviation of each parameter in the target timber texture feature dataset. These percentages are then accumulated to obtain the erosion area detection value in each direction. This is used to obtain the erosion area detection value in each direction as the erosion detection result of the target timber in each direction.

[0057] It should be noted that the specific process of constructing the gray-level co-occurrence matrix includes: converting the orientation scan image of the target wood into a gray-level image using image software, obtaining the maximum and minimum gray-level values ​​in the image, and using the difference between the two as the gray-level range of the orientation scan image.

[0058] The grayscale range of the azimuth scan image is exemplarily divided into 8 equal intervals, and each interval is assigned a grayscale level according to its size.

[0059] The offset is determined by the direction and pixel distance. The direction can be set to 0°, 45°, 90°, or 135° for example, and the pixel distance can be set to 1 for example.

[0060] Initialize a zero matrix of gray levels × gray levels. Determine the neighboring pixels of each pixel based on the offset. Record the gray value pairs of the current pixel and its neighboring pixels. Count the number of times each pixel and its neighboring pixels have the same gray value. Then, normalize the count and use the probability matrix obtained as the gray co-occurrence matrix.

[0061] It should be noted that the texture feature dataset obtained from the gray-level co-occurrence matrix includes contrast, entropy, and energy. The specific acquisition process includes: contrast reflects the roughness of the texture. The more severe the erosion, the higher the contrast. Contrast can be obtained by multiplying each element in the gray-level co-occurrence matrix by the square of the difference between the corresponding row and column coordinates and then summing them up.

[0062] Entropy is an indicator of texture complexity. The more severe the erosion, the higher the entropy. The entropy can be obtained by multiplying each element in the gray-level co-occurrence matrix by its corresponding logarithm, summing them up, and taking the negative value.

[0063] Energy is an indicator of texture uniformity. The more severe the erosion, the lower the energy. Energy can be obtained by calculating the sum of squares of each element in the gray-level co-occurrence matrix.

[0064] The softening detection unit is used to construct a three-dimensional model based on the scanned images of the target wood from various directions, select several detection node positions in each direction, collect hardness data of several detection nodes using a hardness tester and perform benchmark comparison analysis to obtain the softening detection results of the target wood in each direction.

[0065] It should be noted that the above selection of several detection node positions in various directions includes the mortise and tenon connection positions of the target wood, and the determination of each detection node position by pre-setting the spacing distance, starting from one end of the target wood and proceeding from left to right and from top to bottom at equal intervals.

[0066] It should be noted that the specific process for obtaining the hardness values ​​of each detection node includes: establishing a three-dimensional model of the target wood based on scanned images of the target wood from various angles.

[0067] By maneuvering a drone to the detection points in various directions of the target wood, multiple hardness measurements are taken using the onboard hardness tester, and the average of the multiple measurement results is taken as the hardness value of each detection point.

[0068] It should be noted that if the difference between the hardness value at a certain detection node and the reference value exceeds a preset judgment threshold, the node is judged as a softened node. The specific acquisition process includes: calculating the absolute difference between the hardness value at the node and the reference value, and performing a ratio calculation with the reference value. If the result is greater than the preset judgment threshold, the node is judged as a softened node. The preset judgment threshold can be set to 0.15 for example.

[0069] As a preferred option, the specific process for obtaining the softening test results of the target wood in various aspects includes: using the preset standard hardness of the material corresponding to the target wood as a reference value, and comparing and analyzing the hardness value of each test node with the reference value. If the difference between the hardness value of a certain test node and the reference value exceeds a preset judgment threshold, then the node is judged as a softened node.

[0070] Through formula The softening value of the softening node is calculated, where This is the softening value. This represents the hardness value of the softened node. As the baseline value, This represents the average hardness value of the softened nodes in the area where the softened node is located.

[0071] The maximum softening value in each direction is selected as the softened area detection value of the target wood in each direction, and it is used as the softening detection result of the target wood in each direction.

[0072] It should be noted that the target wood is based on a preset standard hardness as a benchmark value. The specific acquisition process includes: determining the wood material corresponding to the target wood based on historical building design drawings stored in the cloud database, and consulting the industry standard hardness of the wood material as a benchmark value.

[0073] As a preferred option, the corrosion disease index is specifically obtained by: constructing a decision matrix based on the collected detection results of mold spots, erosion areas, and softening areas from various directions, wherein the rows of the matrix correspond to the various directions of the target wood, and the columns correspond to the aforementioned detection indicators.

[0074] The decision matrix is ​​standardized, and the maximum and minimum values ​​of each detection index after standardization are collected.

[0075] The elements in the standardized decision matrix are multiplied by the preset weights of the corresponding detection indicators to obtain the weighted decision matrix.

[0076] The vector formed by the minimum values ​​of each detection index is taken as the positive ideal solution, and the vector formed by the maximum values ​​of each detection index is taken as the negative ideal solution.

[0077] The distances to the positive and negative ideal solutions from each direction are calculated using the Euclidean distance formula.

[0078] Calculate the average distance from each direction to the positive ideal solution, and then compare it with the sum of the average distances from each direction to the positive and negative ideal solutions to obtain the corrosion disease index of the target wood.

[0079] As a preferred embodiment, the method of calculating the distances to the positive and negative ideal solutions from each direction using the Euclidean distance formula specifically includes: [details of the formula would be inserted here]. , The distances to the positive and negative ideal solutions in each direction were calculated, where... , Let be the distance from each direction to the positive ideal solution and the negative ideal solution. , The j-th index value of the positive ideal solution and the negative ideal solution. Let be the j-th index value of the i-th orientation in the weighted decision matrix, where Numbering the various locations of the target timber. , Number each test indicator. .

[0080] This invention relies on a corrosion disease module and introduces mold, erosion and softening detection units. The detection data of each unit are comprehensively analyzed to obtain a corrosion disease index, accurately capture potential corrosion risks, comprehensively consider a variety of disease phenomena, and realize multi-dimensional analysis of a single disease. This provides a scientific basis for subsequent diagnosis and early warning, and effectively reduces maintenance costs.

[0081] The insect infestation analysis module is used to collect the location and shape of insect holes using a microwave detector, analyze the distance and volume of the insect holes inside the target wood, and obtain the insect infestation disease index by combining the corresponding preset influencing factors.

[0082] Please see Figure 3 As shown, as a preferred embodiment, the specific process for obtaining the insect infestation index includes: scanning the target wood with a microwave detector, identifying the location and shape of the insect holes from the microwave reflection signal image, and marking the boundary outline of each insect hole and the geometric center of the target wood stake.

[0083] Calculate the distance from each wormhole to the geometric center of the target stump and define it as the distance from the wormhole to the center of the wormhole. Then, perform a reciprocal operation on the distance from the center of the wormhole to obtain the inverse distance from the center of the wormhole.

[0084] The approximate volume of the insect-eaten hole is calculated by selecting the corresponding volume formula based on the shape of the hole.

[0085] The standardized data of the distance and volume of the borer holes are multiplied by the preset influence factors. The sum of squares of the multiplication results is then calculated, and the square root of the sum of squares is taken. The maximum value among all the calculation results is selected as the insect infestation index of the target wood.

[0086] It should be noted that, in order to avoid the case where the reciprocal of the distance to the borer's core is 0, the distance to the borer's core is added to a preset constant when the reciprocal operation is performed. The sum of these additions is then used in the reciprocal operation to obtain the inverse distance to the borer's core of the borer hole. The preset constant can be set to 1 for example.

[0087] It should be noted that the preset influence factors for the borer distance and volume data can be set to 0.7 and 0.3, respectively, based on the fact that borer holes near the center will significantly weaken the load-bearing capacity of the wood. Although large holes cause obvious damage, if they are located at the edge, their impact on the overall structure may be less than that of smaller holes near the center.

[0088] This invention utilizes an insect-borne disease module to accurately acquire data on the location and shape of insect-borne holes, and performs in-depth calculations and analyses on the distance and volume of the borer core, thereby deriving an insect-borne disease index. This effectively improves the accuracy of disease quantification and enhances the efficiency and quality of historical building preservation work.

[0089] The disease diagnosis module is used to perform linear weighted fusion of corrosion disease index and insect infestation disease index to obtain the overall disease index of the target wood.

[0090] It should be noted that the linear weighted fusion of the corrosion disease index and the insect infestation disease index specifically includes the following process: multiplying the corrosion disease index and the insect infestation disease index by preset weights and summing them to obtain the basic feature values ​​of each moldy area.

[0091] It should be noted that the weights of the corrosion disease index and the insect infestation disease index can be 0.6 and 0.4, respectively. This is because corrosion diseases cause more significant damage to the structural integrity of wood and therefore require a higher weight, while insect infestation diseases cause relatively localized damage and therefore have a lower weight.

[0092] The disease early warning module is used to determine whether the disease level of the target timber reaches the warning level based on the preset classification standard for disease levels of wooden structures of historical buildings and the overall disease index of the target timber. If it does, the early warning mechanism is triggered.

[0093] It should be noted that the specific acquisition process of the preset classification standard for the wooden structure disease of historical buildings includes: based on the preset disease levels and corresponding overall disease index ranges of various functional characteristics of each structural component stored in the cloud database; determining the target timber functional characteristic type according to the historical building design drawings; extracting the preset disease levels and corresponding overall disease index ranges of the target timber functional characteristic type stored in the cloud database; and using this as the preset classification standard for the wooden structure disease of historical buildings. The disease levels can be set as Level 1, Level 2, Level 3, and Level 4 in order of increasing disease severity.

[0094] The aforementioned functional characteristics specifically refer to the specific roles of wooden components in historical buildings. For example, the ground column, as the main supporting structure of the building, can be divided into central columns, eaves columns, pedestal columns, and mezzanine columns. Among them, the central column is the core load-bearing component of the building and plays a decisive role in maintaining the overall stability of the building. The eaves column is the main load-bearing component on the exterior of the building, bearing the weight of the eaves. The pedestal columns play a role in auxiliary load-bearing and space division. The mezzanine columns are usually used between beams and rafters to transfer loads and adjust the height of the beam frame.

[0095] It should be noted that the warning level can be set to level three for example.

[0096] This invention provides a comprehensive disease index for historical buildings by combining corrosion and insect infestation indices. Based on a preset classification standard for the disease levels of wooden structures in historical buildings, an early warning mechanism is established. This enables a comprehensive assessment of the health status of historical buildings by combining multiple disease factors, effectively improving the intelligence level of disease assessment and early warning, and saving valuable time for the protection and restoration of historical buildings.

[0097] Reference Figure 2 As shown, the second aspect of the present invention provides an intelligent diagnosis and early warning method for defects in historical buildings, comprising: S1. Selecting a wooden structural component in a historical building as the target timber.

[0098] S2. Perform omnidirectional image acquisition on the target wood, import the corresponding detection results from the mold detection unit, erosion detection unit and softening detection unit respectively, and obtain the corrosion disease index by establishing a decision matrix and using the approximation ideal solution calculation method.

[0099] S3. Collect the location and shape of insect-infested holes using a microwave detector, analyze the distance and volume of the borer holes inside the target wood, and obtain the insect-infestation disease index by combining the corresponding preset influencing factors.

[0100] S4. Perform linear weighted fusion of the corrosion disease index and the insect infestation disease index to obtain the overall disease index of the target wood.

[0101] S5. Based on the preset classification standard for the disease level of wooden structures in historical buildings, and combined with the overall disease index of the target timber, determine whether its disease level reaches the warning level. If it does, trigger the warning mechanism.

[0102] The above content is merely an example and illustration of the concept of the present invention. Those skilled in the art can make various modifications or additions to the specific embodiments described, or use similar methods to replace them, as long as they do not deviate from the concept of the invention or exceed the scope defined by the present invention, and all such modifications and additions should fall within the protection scope of the present invention.

Claims

1. An intelligent diagnostic system for defects in historical buildings, characterized in that, include: The timber sampling module selects a specific wooden structural component from a historical building as the target timber. The corrosion analysis module performs comprehensive image acquisition on the target wood, imports the corresponding detection results from the mold detection unit, erosion detection unit and softening detection unit, and obtains the corrosion disease index by establishing a decision matrix and using the approximation ideal solution calculation method. The insect infestation analysis module uses a microwave detector to collect the location and shape of insect holes, analyzes the distance and volume of the borer holes inside the target wood, and combines them with the corresponding preset influencing factors to obtain the insect infestation disease index. The disease diagnosis module performs linear weighted fusion of corrosion disease index and insect infestation disease index to obtain the overall disease index of the target wood. The disease early warning module, based on the preset classification standard for disease levels of wooden structures in historical buildings, combined with the overall disease index of the target timber, determines whether its disease level has reached the early warning level. If it has, the early warning mechanism is triggered. The mold detection unit is used to identify several mold areas in the scanned images of the target wood from various directions. Based on the comprehensive basic features of the mold areas from various directions and the analysis of the aggregation features between the mold areas, the mold detection results of the target wood from various directions are obtained. The comprehensive basic features of mold spots in each direction are obtained as follows: The area ratio of each mold spot region within the scanned image of the target wood in each direction is obtained; the scanned images of the target wood in each direction are converted to the HSL color space, and the brightness, saturation, and hue of the normal area of ​​the target wood are used as standard reference colors. The brightness, saturation, and hue of each mold spot region are measured and compared with the standard reference colors to obtain the color parameter deviation value of each mold spot region; the mold spot regions are extracted through image segmentation, and the edge contours of the mold spot regions are identified. The edge density is obtained by calculating the ratio of the number of pixels of the edge contour of the mold spot region to its total number of pixels. The reciprocal of the edge density is used as the continuous value of the mold spot region, and this is used to obtain the continuous value of each mold spot region; the area ratio, color parameter deviation value, and continuous value are integrated to generate the basic feature value of each mold spot region in each direction, and the maximum value is extracted as the comprehensive basic feature value of the mold spot region in each direction. The erosion detection unit is used to analyze the scanned images of the target wood from all directions and the design images of the target wood stored in the cloud, to obtain the texture feature data of the scanned images and design images of the target wood from all directions, and to compare and analyze the two to obtain the erosion detection results of the target wood from all directions.

2. The intelligent diagnostic system for historical building defects according to claim 1, characterized in that, The softening detection unit is used to construct a three-dimensional model based on the scanned images of the target wood from various directions, select several detection node positions in each direction, collect hardness data of several detection nodes using a hardness tester and perform benchmark comparison analysis to obtain the softening detection results of the target wood in each direction.

3. The intelligent diagnostic system for historical building defects according to claim 2, characterized in that, The specific process for obtaining the mold spot detection results from all directions of the target wood also includes: Establish a pixel coordinate system for the azimuth scan image and determine the horizontal and vertical coordinates of the center of each mold patch region. For any mold patch region, calculate the distance from its center to the center of all other mold patch regions, select the minimum value as the nearest neighbor distance, and calculate the average of the nearest neighbor distances of all mold patches as the clustering characteristic value of the azimuth mold patch region. Based on the comprehensive basic feature value and clustered feature value of the directional mold patch area, the directional mold patch area detection value is calculated by the geometric mean formula, and the mold patch area detection value of each direction is obtained as the mold patch detection result of each direction of the target wood.

4. The intelligent diagnostic system for historical building defects according to claim 2, characterized in that, The specific process for obtaining the erosion detection results of the target wood in each direction is as follows: By analyzing the directional scan image of the target wood, a gray-level co-occurrence matrix is ​​constructed. Based on the gray-level co-occurrence matrix, a texture feature dataset is obtained, including contrast, entropy, and energy. Based on the target wood design image recorded in the historical building design drawings stored in the cloud database, a reference texture feature dataset of the target wood design image is obtained according to the method for obtaining the texture feature dataset of the target wood scan image. The absolute deviation between the reference texture feature dataset and the target wood scan image texture feature dataset is calculated, and then a ratio operation is performed with the corresponding parameters of the reference texture feature dataset to obtain the absolute deviation ratio of each parameter in the target wood texture feature dataset. These ratios are then accumulated to obtain the erosion area detection value in each direction. The erosion area detection values ​​in each direction are thus obtained as the erosion detection results of the target wood in each direction.

5. The intelligent diagnostic system for historical building defects according to claim 2, characterized in that, The softening test results of the target wood in various directions are obtained through the following process: The material standard hardness corresponding to the target wood is used as the benchmark value, and the hardness value of each detection node is compared with the benchmark value. If the difference between the hardness value of a certain detection node and the benchmark value exceeds the preset judgment threshold, the node is judged as a softened node. Through formula The softening value of the softening node is calculated, where This is the softening value. This represents the hardness value of the softened node. As the baseline value, This represents the average hardness value of the softened nodes in the location of the softened node; The maximum softening value in each direction is selected as the softened area detection value of the target wood in each direction, and it is used as the softening detection result of the target wood in each direction.

6. The intelligent diagnostic system for historical building defects according to claim 1, characterized in that: The corrosion disease index is obtained through the following process: A decision matrix is ​​constructed based on the collected detection results of mold spots, erosion areas, and softened areas from various directions. The rows of the matrix correspond to the various directions of the target wood, and the columns correspond to the above detection indicators. The decision matrix is ​​standardized, and the maximum and minimum values ​​of each detection index after standardization are collected. Multiply the elements in the standardized decision matrix by the preset weights of the corresponding detection indicators to obtain the weighted decision matrix; The vector formed by the minimum values ​​of each detection index is taken as the positive ideal solution, and the vector formed by the maximum values ​​of each detection index is taken as the negative ideal solution. The distances to the positive and negative ideal solutions from each direction are calculated using the Euclidean distance formula. Calculate the average distance from each direction to the positive ideal solution, and then compare it with the sum of the average distances from each direction to the positive and negative ideal solutions to obtain the corrosion disease index of the target wood.

7. The intelligent diagnostic system for historical building defects according to claim 6, characterized in that, The process of calculating the distances to the positive and negative ideal solutions from each direction using the Euclidean distance formula is as follows: From the formula , The distances to the positive and negative ideal solutions in each direction were calculated, where... , Let be the distance from each direction to the positive ideal solution and the negative ideal solution. , The j-th index value of the positive ideal solution and the negative ideal solution. Let be the j-th index value of the i-th orientation in the weighted decision matrix, where Numbering the various locations of the target timber. , Number each test indicator. .

8. The intelligent diagnostic system for historical building defects according to claim 1, characterized in that, The specific process for obtaining the insect infestation index is as follows: The target wood is scanned by a microwave detector, and the location and shape of the insect holes are identified from the microwave reflection signal image. The boundary outline of each insect hole and the geometric center of the target wood stake are marked. Calculate the distance from each wormhole to the geometric center of the target stump and define it as the distance from the wormhole's core. Then, perform a reciprocal operation on the distance from the wormhole's core to obtain the inverse distance from the wormhole's core. Calculate the approximate volume of the insect-eaten hole by selecting the corresponding volume formula based on the shape of the hole; The standardized data of the distance and volume of the borer holes are multiplied by the preset influence factors. The sum of squares of the multiplication results is then calculated, and the square root of the sum of squares is taken. The maximum value among all the calculation results is selected as the insect infestation index of the target wood.

9. A method for intelligent diagnosis and early warning of defects in historical buildings, comprising the following steps performed by an intelligent diagnosis system for defects in historical buildings as described in any one of claims 1-7, characterized in that, include: S1. Select a wooden structural component from a historical building as the target timber; S2. Perform omnidirectional image acquisition on the target wood, import the corresponding detection results from the mold detection unit, erosion detection unit and softening detection unit respectively, and obtain the corrosion disease index by establishing a decision matrix and using the approximation ideal solution calculation method. S3. Collect the location and shape of insect-infested holes using a microwave detector, analyze the distance and volume of the borer holes inside the target wood, and obtain the insect-infestation disease index by combining the corresponding preset influencing factors. S4. Perform linear weighted fusion of the corrosion disease index and the insect infestation disease index to obtain the overall disease index of the target wood; S5. Based on the preset classification standard for the disease level of wooden structures in historical buildings, and combined with the overall disease index of the target timber, determine whether its disease level reaches the warning level. If it does, trigger the warning mechanism.

Images