Accurate diagnosis, renovation and design methods for existing old buildings

Through image recognition technology, the automated diagnosis of old buildings is solved, and the problem of low efficiency of traditional diagnostic methods is achieved, efficient diagnosis and renovation design of old buildings is achieved, and the accuracy and efficiency of design solutions are improved.

CN117994235BActive Publication Date: 2024-11-15GUANGZHOU MUNICIPAL GRP DESIGN INST CO LTD +3
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202410179527.9
Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
Filing Date
2024-02-18
Publication Date
2024-11-15
Estimated Expiration
2044-02-18

AI Technical Summary

Technical Problem

Traditional old buildings diagnosis methods are inefficient, especially when multiple old buildings need to be diagnosed, it takes a long time to effectively promote the renovation design.

Method used

Image recognition technology is used to identify images to be diagnosed in old buildings, and obtain building defects, including automated processes for shooting, image transmission, recognition, and design solutions. Using image recognition devices such as computers and smartphones, shooting and image segmentation through drones , filtering and other technologies improve diagnostic efficiency.

Benefits of technology

It effectively improves the efficiency of old building diagnosis, can effectively promote the progress of diagnosis and transformation, reduces the time cost of manual diagnosis, and improves the accuracy and efficiency of design solutions.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN117994235B_ABST
    Figure CN117994235B_ABST
Patent Text Reader

Abstract

The present invention belongs to the field of image recognition, and discloses a method for accurate diagnosis, reconstruction and design of existing old buildings, including: S1, photographing the old building to obtain an image to be diagnosed; S2, transmitting the image to be diagnosed to an image recognition device; S3, recognizing the image to be diagnosed by the image recognition device, and obtaining the building defects in the image to be diagnosed. Compared with the prior art, the present invention obtains the building defects in the appearance of the old building by image recognition, effectively improves the efficiency of diagnosing the appearance of the old building, and can efficiently promote the progress of diagnosis and reconstruction of the old building.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present invention relates to the field of image recognition, and in particular to a method for accurate diagnosis, reconstruction and design of existing old buildings. Background Art

[0002] Before renovating old buildings, it is usually necessary to diagnose their appearance first, so that the existing problems in the appearance of the old buildings can be transformed in a targeted manner. The traditional diagnosis method is for personnel with diagnosis experience to conduct on-site inspections and then judge the existing problems in the old buildings based on their own experience. However, this diagnosis method is relatively inefficient. When the number of old buildings that need to be diagnosed is large, it takes a long time to get the diagnosis results. Summary of the invention

[0003] The purpose of the present invention is to disclose a method for accurate diagnosis, renovation and design of existing old buildings, so as to solve the problem of how to improve the efficiency of diagnosis when diagnosing old buildings.

[0004] In order to achieve the above object, the present invention provides the following technical solutions:

[0005] The present invention provides a method for accurate diagnosis, reconstruction and design of existing old buildings, including:

[0006] S1, photograph the old building to obtain the image to be diagnosed;

[0007] S2, transmitting the image to be diagnosed to an image recognition device;

[0008] S3, identifying the image to be diagnosed by an image recognition device to obtain the building defects existing in the image to be diagnosed.

[0009] Optionally, S1 includes:

[0010] Photographing the exterior wall of the old building to obtain an image to be diagnosed of the exterior wall of the old building;

[0011] The interior space of the old building is photographed to obtain an image to be diagnosed of the interior space of the old building.

[0012] Optionally, the image recognition device includes a computer and a smart phone.

[0013] Optionally, after S3, also include:

[0014] S4, sending all the building defects and the image to be diagnosed corresponding to each building defect to the office equipment used by the designer;

[0015] S5, receiving the design plan sent by the office equipment, which is obtained by the designer to design for the building defects.

[0016] Optionally, photographing the exterior wall of the old building to obtain the image to be diagnosed of the exterior wall of the old building includes:

[0017] The exterior wall of the old building is photographed by a drone to obtain the image to be diagnosed of the exterior wall of the old building.

[0018] Optionally, a drone is used to photograph the exterior wall of the old building to obtain an image to be diagnosed of the exterior wall of the old building, including:

[0019] For the same shooting position, the drone is controlled to take multiple photos of the old building to obtain multiple images to be diagnosed.

[0020] Optionally, the process of identifying the image to be diagnosed by an image recognition device and obtaining the building defects existing in the image to be diagnosed includes:

[0021] Obtain the image to be identified at each shooting position respectively;

[0022] The image to be identified is identified to obtain the building defects existing in the image to be identified.

[0023] Optionally, the image to be identified at each shooting position is obtained separately, including:

[0024] respectively calculating the image comparison value of each image to be diagnosed obtained at the same shooting position;

[0025] The image to be diagnosed with the highest image comparison value is used as the image to be identified.

[0026] Optionally, the image to be identified is identified to obtain building defects in the image to be identified, including:

[0027] Perform image segmentation on the image to be identified to obtain the foreground area in the image to be identified;

[0028] The foreground area is identified and the building defects existing in the foreground area are obtained.

[0029] Optionally, performing image segmentation on the image to be identified to obtain a foreground area in the image to be identified includes:

[0030] Filter the image to be recognized to obtain a filtered image;

[0031] Perform image segmentation on the filtered image to obtain the foreground area in the image to be identified.

[0032] Beneficial effects:

[0033] Compared with the prior art, the present invention obtains the architectural defects existing in the appearance of old buildings by means of image recognition, effectively improves the efficiency of diagnosing the appearance of old buildings, and can efficiently promote the progress of diagnosis and renovation of old buildings. BRIEF DESCRIPTION OF THE DRAWINGS

[0034] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the accompanying drawings required for describing the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other accompanying drawings can be obtained based on these accompanying drawings without paying creative work.

[0035] Figure 1 A schematic diagram of the accurate diagnosis, renovation and design method for existing old buildings of the present invention.

[0036] Figure 2 Another schematic diagram of the method for accurate diagnosis, renovation and design of existing old buildings of the present invention. DETAILED DESCRIPTION

[0037] 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 only 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 ordinary technicians in this field without creative work are within the scope parameters of the protection of the present invention.

[0038] like Figure 1 In one embodiment shown, the present invention provides a method for accurate diagnosis, reconstruction and design of existing old buildings, including:

[0039] S1, photograph the old building to obtain the image to be diagnosed;

[0040] S2, transmitting the image to be diagnosed to an image recognition device;

[0041] S3, identifying the image to be diagnosed by an image recognition device to obtain the building defects existing in the image to be diagnosed.

[0042] The invention obtains the architectural defects existing in the appearance of old buildings by means of image recognition, effectively improves the efficiency of diagnosing the appearance of old buildings, and can efficiently promote the progress of diagnosing and renovating old buildings.

[0043] Optionally, S1 includes:

[0044] Photographing the exterior wall of the old building to obtain an image to be diagnosed of the exterior wall of the old building;

[0045] The interior space of the old building is photographed to obtain an image to be diagnosed of the interior space of the old building.

[0046] Specifically, the public areas inside the old building may be photographed to obtain images to be diagnosed of the public areas at various locations inside the old building.

[0047] Optionally, the image recognition device includes a computer and a smart phone.

[0048] Optional, such as Figure 2 As shown, after S3, it also includes:

[0049] S4, sending all the building defects and the image to be diagnosed corresponding to each building defect to the office equipment used by the designer;

[0050] S5, receiving the design plan sent by the office equipment, which is obtained by the designer to design for the building defects.

[0051] Specifically, the office equipment used by designers includes desktop computers and workstations.

[0052] Specifically, for the exterior walls of old buildings, defects include hollowing and cracking of the finishing layer, peeling of the walls, aging of the walls, damaged downpipes, etc.

[0053] For the interior spaces of old buildings, defects include indoor water seepage, exposed pipes, insufficient light, etc.

[0054] The design plan is a design solution given by the designer to address the defects in the building. For example, the corresponding design solution for the defects of aging walls is repainting.

[0055] To address the problem of insufficient light, the corresponding design solutions are to replace lighting equipment with brighter ones or to enlarge the windows used for lighting.

[0056] Optionally, photographing the exterior wall of the old building to obtain the image to be diagnosed of the exterior wall of the old building includes:

[0057] The exterior wall of the old building is photographed by a drone to obtain the image to be diagnosed of the exterior wall of the old building.

[0058] Specifically, drones can take photos of exterior walls at high altitudes, which has greater shooting advantages compared to manual photography.

[0059] Optionally, photographing the interior space of the old building to obtain the image to be diagnosed of the interior space of the old building includes:

[0060] The staff carries photographing equipment to photograph various places of the interior space of the old building, and obtains a plurality of images to be diagnosed of the interior space of the old building.

[0061] Optionally, a drone is used to photograph the exterior wall of the old building to obtain an image to be diagnosed of the exterior wall of the old building, including:

[0062] For the same shooting position, the drone is controlled to take multiple photos of the old building to obtain multiple images to be diagnosed.

[0063] Specifically, the outer wall of the old building can be divided into a plurality of shooting areas of the same size, and the center of each shooting area is used as the shooting position.

[0064] Specifically, when multiple shots are taken, the number of shots is greater than 30.

[0065] Optionally, the process of identifying the image to be diagnosed by an image recognition device and obtaining the building defects existing in the image to be diagnosed includes:

[0066] Obtain the image to be identified at each shooting position respectively;

[0067] The image to be identified is identified to obtain the building defects existing in the image to be identified.

[0068] Specifically, screening the images to be identified is helpful to select images to be diagnosed with the highest possible quality, thereby improving the accuracy of subsequent identification.

[0069] Optionally, the image to be identified at each shooting position is obtained separately, including:

[0070] respectively calculating the image comparison value of each image to be diagnosed obtained at the same shooting position;

[0071] The image to be diagnosed with the highest image comparison value is used as the image to be identified.

[0072] Specifically, the larger the image comparison value is, the better the image quality is, thereby achieving effective screening of all images to be diagnosed obtained from the same shooting position.

[0073] Optionally, the image comparison value of each image to be diagnosed obtained at the same shooting position is calculated respectively, including:

[0074] All the images to be diagnosed are numbered consecutively in the order of shooting time from earliest to latest. The numbers are integers and are greater than or equal to 1. The earlier the shooting time, the smaller the number.

[0075] First calculation:

[0076] Calculating the image to be diagnosed numbered 1 to obtain an image comparison value of the image to be diagnosed numbered 1;

[0077] Second calculation:

[0078] According to the preset calculation distance, determine the number num2 of the image to be diagnosed to be calculated for the second time;

[0079] Calculate the image to be diagnosed numbered num2 to obtain the image comparison value of the image to be diagnosed numbered num2;

[0080] The kth calculation, k is greater than 2:

[0081] Obtaining an adaptive calculation distance according to image comparison values ​​obtained by the k-1th calculation and the k-2th calculation;

[0082] Based on the adaptive calculation distance, determine the number num of the image to be diagnosed that needs to calculate the image comparison value during the k-th calculation k ;

[0083] Determine num k Is it greater than the total number of images to be diagnosed obtained from the same shooting position? If not, then k The image to be diagnosed is calculated and the number is num k The image comparison value of the image to be diagnosed is calculated; if so, the image comparison value of the image to be diagnosed for which the image comparison value has not been calculated is set to 0, and the calculation is terminated.

[0084] Different from the prior art, the present invention does not calculate the image comparison value for each image to be diagnosed separately, because too many images need to calculate the image comparison value, which affects the efficiency of screening. Moreover, the present invention does not need to obtain the highest quality image to be diagnosed, but only needs to obtain the image to be diagnosed with the highest quality possible. Therefore, in addition to the first calculation and the second calculation, the present invention uses the image comparison values ​​obtained from the first two calculations and determines the adaptive calculation distance; it realizes the non-equidistant selection of the image to be diagnosed for which the image comparison value is not calculated, and can improve the efficiency of the acquisition process while obtaining the highest quality image to be diagnosed as much as possible.

[0085] Optionally, the preset calculation distance is 3.

[0086] Optionally, calculating the image to be diagnosed numbered 1 to obtain the image comparison value of the image to be diagnosed numbered 1 includes:

[0087] In the first step, the total number of images to be diagnosed obtained from the same shooting position is represented as K, and the image to be diagnosed numbered 1 and the image to be diagnosed numbered K are represented as p1 and p2, respectively. K ;

[0088] The second step is based on p K Calculate the judgment value of each pixel in p1 respectively;

[0089] The third step is to obtain a set Up1 of pixel points in p1 whose judgment values ​​are greater than the set judgment value threshold;

[0090] The fourth step is to perform the following calculations on each pixel in Up1:

[0091] For a pixel point pixUp1 in Up1, if there is no pixel point in Up1 within the range of radius R with pixUp1 as the center, then pixUp1 is deleted from Up1;

[0092] Step 5: Get the maximum value x of the horizontal coordinate of the pixel point in Up1 a , the minimum value x of the horizontal axis i , the maximum value of the vertical coordinate y a and the minimum value y of the ordinate i ;

[0093] Step 6: Store the coordinates of the pixels whose horizontal and vertical coordinates meet the following conditions into the set sUp1:

[0094]

[0095] The seventh step is to calculate the image comparison value of the pixel point corresponding to the coordinate in sUp1, so as to obtain the image comparison value of the image to be diagnosed numbered 1.

[0096] When calculating the image comparison value of the image to be diagnosed numbered 1, the present invention first obtains the set sUp1 and then calculates the image comparison value based on sUp1. Thus, subsequent calculations can calculate the image comparison value based on the set sUp1, avoiding calculating the image comparison value based on all pixels, and further improving the calculation efficiency of the image comparison value.

[0097] In the process of obtaining sUp1, the present invention calculates the judgment value of each pixel in p1 respectively, and takes the pixel points whose judgment value is greater than the set judgment value threshold as the pixel points obtained in the first round of screening. K There are pixel points in different areas, because the drone may encounter lighting changes when shooting, and the lighting changes have different effects on different areas of the exterior wall. Therefore, the present invention can calculate the image comparison value based on the area that is more affected by the lighting changes, so that the image can be screened more accurately to obtain the highest quality image to be diagnosed. If the image comparison value is calculated for the global image, then when the area that is more affected by the lighting changes is small, the calculated image comparison value is difficult to effectively represent the true quality of the image to be diagnosed.

[0098] Optional, based on p K Calculate the judgment value of each pixel in p1 separately, including:

[0099] The p Kand p1 are converted to Lab color space to obtain p K The image Lp corresponding to the brightness component of p1 in the Lab color space K and Lp1;

[0100] For the pixel point p1(x,y) with coordinates (x,y) in p1, the judgment value of p1(x,y) is calculated using the following formula:

[0101] judp1(x,y)=|Lp1(x,y)-Lp K (x,y)|

[0102] Lp1(x,y) and Lp K (x,y) are Lp K In Lp1 and Lp2, the coordinates are the brightness values ​​of the pixels at the coordinates (x, y).

[0103] Specifically, by converting the image to be diagnosed into the Lab color space, it is possible to reduce the amount of calculation while retaining most of the details of the image to be diagnosed, because only the brightness component needs to be calculated. K The brightness difference between Lp1 and Lp2 can be used to obtain the judgment value.

[0104] Optionally, the judgment value threshold is set to 5.

[0105] Optional, the value of R is 3.

[0106] Optionally, calculating the image comparison value of the pixel corresponding to the coordinate in sUp1 includes:

[0107] Get the set of pixel points Ulp1 corresponding to the coordinates in sUp1 in Lp1;

[0108] The image comparison value of the image to be diagnosed numbered 1 is calculated using the following formula:

[0109]

[0110] compval1 represents the image comparison value of the image to be diagnosed numbered 1, Lp1(u) represents the brightness value of the pixel u in Ulp1 in Lp1, nUlp1 represents the number of pixels in Ulp1, and Lp 1,i represents the minimum brightness value of the pixel points in Ulp1 in Lp1, nsobel1 represents the number of pixels whose gradient is greater than the set gradient threshold calculated by the sobel algorithm for the pixel points in Ulp1; Φ1 and Φ2 represent the set brightness distribution weight and effective information weight, respectively.

[0111] The image comparison value is calculated from the difference characteristics of the pixels in the image to be diagnosed on the one hand, and from the effectiveness of the pixels in the image to be diagnosed on the other hand. Since the quality of the image to be diagnosed is calculated from different angles, the quality of the image to be diagnosed can be more accurately represented.

[0112] The greater the brightness difference of the pixels in the image and the more pixels whose values ​​are greater than the set gradient threshold, the larger the image comparison value, indicating that the image to be diagnosed carries more valid data and can more realistically restore the condition of the building's exterior wall.

[0113] Optionally, the brightness distribution weight and the effective information weight are set to 0.2 and 0.8 respectively.

[0114] Optionally, the process of calculating the image comparison value for the to-be-diagnosed image with a number greater than 1 includes:

[0115] For number num v Images to be diagnosed v is greater than or equal to 2;

[0116] Get Luminance component of the image in Lab color space

[0117] exist Get the set of pixel points corresponding to the coordinates in sUp1

[0118] Use the following formula to calculate the number num v Image comparison value of the image to be diagnosed:

[0119]

[0120] Indicates the number is num v The image comparison value of the image to be diagnosed, express The pixel u in The brightness value in express The number of pixels in express The pixels in The minimum brightness value in Indicates the use of the Sobel algorithm The number of pixels whose gradients are greater than the set gradient threshold is calculated for the pixel points in ; Φ1 and Φ2 represent the set brightness distribution weight and effective information weight, respectively.

[0121] The present invention calculates the set sUp1 through the first image and the last image. When subsequently calculating the image comparison value, it can be calculated directly based on sUp1, thereby avoiding the calculation of the image comparison value based on all pixel points, thereby ensuring the validity of the calculation result while improving the calculation efficiency.

[0122] Optionally, according to a preset calculation distance, determining the number num2 of the image to be diagnosed to be calculated for the second time includes:

[0123] num2=1+presetdis

[0124] presetdis indicates the preset calculation distance.

[0125] Optionally, obtaining the adaptive calculation distance according to the image comparison values ​​obtained by the k-1th calculation and the k-2th calculation includes:

[0126] The numbers of the images to be diagnosed corresponding to the image comparison values ​​obtained by the k-1th calculation and the k-2th calculation are represented as num k-1 and num k-2 ;

[0127] The calculation formula for adaptive distance is:

[0128]

[0129] caldis k-1 represents the adaptive calculation distance obtained based on the image comparison values ​​obtained by the k-1th calculation and the k-2th calculation, and Respectively represent the number num k- 1 and num k- 2, and norma means normalizing the variables in the brackets.

[0130] The adaptive calculation distance of the present invention can adaptively change with the change of the size of the difference between the image comparison values ​​obtained by two adjacent calculations. Specifically, when the difference between the image comparison values ​​is larger, the adaptive calculation distance of the present invention is smaller, indicating that the difference of the images to be diagnosed obtained by continuous shooting at this time varies greatly. Therefore, at this time, the present invention reduces the adaptive calculation distance, which can make the image to be diagnosed selected for calculating the image comparison value closer to the high-quality image to be diagnosed. On the contrary, it means that the difference of the images to be diagnosed obtained by continuous shooting does not vary much. The present invention expands the adaptive calculation distance and reduces the number of images to be diagnosed for calculating the image comparison value as much as possible, thereby further improving the computational efficiency of the process of screening out images to be diagnosed with the highest possible image quality.

[0131] Optionally, based on the adaptive calculation distance, determine the number num of the image to be diagnosed for which the image comparison value needs to be calculated during the k-th calculation k ,include:

[0132] Use the following formula to calculate num k :

[0133] num k =num k-1 +caldis k-1 .

[0134] Optionally, the image to be identified is identified to obtain building defects in the image to be identified, including:

[0135] Perform image segmentation on the image to be identified to obtain the foreground area in the image to be identified;

[0136] The foreground area is identified and the building defects existing in the foreground area are obtained.

[0137] Specifically, the foreground area may be input into a trained recognition model for recognition, thereby obtaining the building defects existing in the foreground area.

[0138] Optionally, performing image segmentation on the image to be identified to obtain a foreground area in the image to be identified includes:

[0139] Filter the image to be recognized to obtain a filtered image;

[0140] Perform image segmentation on the filtered image to obtain the foreground area in the image to be identified.

[0141] Filtering can improve the quality of the image, thereby improving the accuracy of the image segmentation results.

[0142] Optionally, performing image segmentation on the filtered image to obtain a foreground area in the image to be identified includes:

[0143] The KSW entropy algorithm is used to segment the filtered image to obtain the foreground area in the image to be identified.

[0144] Specifically, the pixel points that are larger than the segmentation threshold are regarded as the pixel points of the foreground area.

[0145] In the several embodiments provided by the present invention, it should be understood that the disclosed devices and methods can be implemented in other ways. For example, the device embodiments described above are only schematic, for example, the division of units is only a logical function division, and there may be other division methods in actual implementation, such as multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the mutual coupling or direct coupling or communication connection shown or discussed can be through some interfaces, indirect coupling or communication connection of devices or units, which can be electrical, mechanical or other forms.

[0146] The units described as separate components may or may not be physically separated, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed on multiple network units. Some or all of the units may be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0147] In addition, each functional unit in each embodiment of the present invention may be integrated into one processing unit, or each unit may exist physically separately, or two or more units may be integrated into one unit. The above-mentioned integrated unit may be implemented in the form of hardware or in the form of software functional units.

[0148] The above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit the same. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that the technical solutions described in the aforementioned embodiments may still be modified, or some of the technical features may be replaced by equivalents. However, these modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope parameters of the technical solutions of the embodiments of the present invention.

Claims

1. Accurate diagnosis, renovation and design method for existing old buildings, characterized by: include: S1, photograph the old building to obtain the image to be diagnosed; S2, transmitting the image to be diagnosed to an image recognition device; S3, identifying the image to be diagnosed by an image recognition device to obtain the building defects existing in the image to be diagnosed; S1 includes: Photographing the exterior wall of the old building to obtain an image to be diagnosed of the exterior wall of the old building; Photographing the interior space of the old building to obtain the image to be diagnosed of the interior space of the old building; The exterior wall of the old building is photographed to obtain the image to be diagnosed of the exterior wall of the old building, including: Use a drone to photograph the exterior wall of the old building to obtain an image to be diagnosed of the exterior wall of the old building; Use a drone to photograph the exterior wall of the old building to obtain the image to be diagnosed of the exterior wall of the old building, including: For the same shooting position, the drone is controlled to shoot the old building multiple times to obtain multiple images to be diagnosed; The process of identifying the image to be diagnosed by an image recognition device and obtaining the building defects in the image to be diagnosed includes: Obtain the image to be identified at each shooting position respectively; Identify the image to be identified, and obtain the building defects existing in the image to be identified; Obtain the image to be identified at each shooting position respectively, including: respectively calculating the image comparison value of each image to be diagnosed obtained at the same shooting position; The image to be diagnosed with the highest image comparison value is used as the image to be identified; Calculate the image comparison value of each image to be diagnosed obtained at the same shooting position, including: All the images to be diagnosed are numbered consecutively in the order of shooting time from earliest to latest. The numbers are integers and are greater than or equal to 1. The earlier the shooting time, the smaller the number. First calculation: Calculating the image to be diagnosed numbered 1 to obtain an image comparison value of the image to be diagnosed numbered 1; Second calculation: According to the preset calculation distance, determine the number num2 of the image to be diagnosed to be calculated for the second time; Calculate the image to be diagnosed numbered num2 to obtain the image comparison value of the image to be diagnosed numbered num2; The kth calculation, k is greater than 2: Obtaining an adaptive calculation distance according to image comparison values ​​obtained by the k-1th calculation and the k-2th calculation; Based on the adaptive calculation distance, determine the number num of the image to be diagnosed that needs to calculate the image comparison value during the k-th calculation k ; Determine num k Is it greater than the total number of images to be diagnosed obtained from the same shooting position? If not, then k The image to be diagnosed is calculated and the number is num k The image comparison value of the image to be diagnosed; if so, the image comparison value of the image to be diagnosed for which the image comparison value has not been calculated is set to 0, and the calculation is terminated; Calculating the image to be diagnosed numbered 1 to obtain an image comparison value of the image to be diagnosed numbered 1 includes: In the first step, the total number of images to be diagnosed obtained from the same shooting position is represented as K, and the image to be diagnosed numbered 1 and the image to be diagnosed numbered K are represented as p1 and p2, respectively. K ; The second step is based on p K Calculate the judgment value of each pixel in p1 respectively; The third step is to obtain a set Up1 of pixel points in p1 whose judgment values ​​are greater than the set judgment value threshold; The fourth step is to perform the following calculations on each pixel in Up1: For a pixel point pixUp1 in Up1, if there is no pixel point in Up1 within the range of radius R with pixUp1 as the center, then pixUp1 is deleted from Up1; Step 5: Get the maximum value x of the horizontal coordinate of the pixel point in Up1 a , the minimum value x of the horizontal axis i , the maximum value of the vertical coordinate y a and the minimum value y of the ordinate i ; Step 6: Store the coordinates of the pixels whose horizontal and vertical coordinates meet the following conditions into the set sUp1: Step 7: Calculate the image comparison value of the pixel point corresponding to the coordinate in sUp1, so as to obtain the image comparison value of the image to be diagnosed numbered 1; Based on p K Calculate the judgment value of each pixel in p1 separately, including: The p K and p1 are converted to Lab color space to obtain p K The image Lp corresponding to the brightness component of p1 in the Lab color space K and Lp1; For the pixel point p1(x,y) with coordinates (x,y) in p1, the judgment value of p1(x,y) is calculated using the following formula: judp1(x,y)=|Lp1(x,y)-Lp K (x,y) Lp1(x,y) and Lp K (x,y) are Lp K In Lp1 and Lp2, the coordinates are the brightness values ​​of the pixels at the coordinates (x, y).

2. The method for accurate diagnosis, reconstruction and design of existing old buildings according to claim 1 is characterized in that: Image recognition devices include computers and smartphones.

3. The method for accurate diagnosis, reconstruction and design of existing old buildings according to claim 1 is characterized in that: After S3, it also includes: S4, sending all the building defects and the image to be diagnosed corresponding to each building defect to the office equipment used by the designer; S5, receiving the design plan sent by the office equipment, which is obtained by the designer to design the building defects.

4. The method for accurate diagnosis, reconstruction and design of existing old buildings according to claim 1 is characterized in that: Identify the image to be identified and obtain the building defects in the image to be identified, including: Perform image segmentation on the image to be identified to obtain the foreground area in the image to be identified; The foreground area is identified to obtain the building defects existing in the foreground area.

5. The method for accurate diagnosis, reconstruction and design of existing old buildings according to claim 4 is characterized in that: Perform image segmentation on the image to be identified to obtain the foreground area in the image to be identified, including: Filter the image to be recognized to obtain a filtered image; Perform image segmentation on the filtered image to obtain the foreground area in the image to be identified.