Method for recognizing building damage and generating maintenance work order based on space image

Abstract

The invention discloses a method for identifying building damage and generating a maintenance work order based on a space image, which comprises the following steps of: acquiring an orthographic projection panoramic color image of each surface of a building, correcting the orthographic projection panoramic color image into a surface panoramic color image, and converting the surface panoramic color image into a surface panoramic black-and-white image; cutting and segmenting each panoramic color image and each black-and-white image into a regional grid color image and a black-and-white image; grid segmentation lines are drawn on the regional grid black-and-white graph to form a regional number index graph, and the regional number index graph is cut and segmented to form a block black-and-white graph as a block labeling base graph; cutting and segmenting the regional grid color image into block color images, carrying out automatic detection and labeling on the damage condition through an instance segmentation algorithm, and extracting a labeled image layer as a block damage mask image; overlapping each block damage mask pattern on the block marking base pattern to form a block marking damage pattern; and extracting damage information of the block labeling damage graph to generate a damage form, and generating a maintenance work order by using the region number index graph and the block labeling damage graph and the damage form in the region number index graph. The method has the advantages of high damage identification efficiency and high accuracy.

Description

Technical Field

[0001] The present invention relates to the field of building maintenance and construction, and in particular to a method for identifying building damage based on spatial images. Background Art

[0002] When investigating existing buildings, it is necessary to rely on historical drawings to verify and mark the current status. However, many old buildings lack historical data, and draftsmen are still required to carefully measure the on-site dimensions and then draw standard technical drawings. Then, the damaged parts obtained from the on-site investigation are marked on two-dimensional technical drawings to guide on-site operations. The drawing process takes a long time and is prone to omissions. The positioning of the final damage annotation drawings will have a large error with the actual position due to the two-step operation, resulting in inconsistent information during construction briefing. In addition, the facade damage is complex and the number is large. It is easy to miss it through manual identification alone, resulting in inaccurate statistics of the damaged area. Summary of the invention

[0003] The purpose of the present invention is to provide a method for generating a maintenance work order based on spatial image recognition of building damage, so as to solve the problems of low efficiency and low accuracy in drawing engineering drawings with damage information during manual investigation of historical building damage.

[0004] In order to solve the above technical problems, the present invention provides a method for generating a maintenance work order based on spatial image recognition of building damage, comprising: Step S1, obtaining an orthographic panoramic color image of each face of the building; Step S2, measuring the outer contour size of the building and drawing the building frame diagram, placing the orthographic projection composite color map of each face into the frame of the corresponding face of the building frame diagram, and performing distortion correction and edge clipping to form a panoramic color map; Step S3, converting the color panoramic images of each surface into black and white panoramic images; Step S4, cutting and dividing each of the surface panoramic color images and the surface panoramic black-and-white images according to the same grid size and numbering them to form a plurality of regional grid color images and a plurality of regional grid black-and-white images with the same number; Step S5, drawing grid dividing lines on the black-and-white regional grid images of each number according to the same grid size to form a regional number index map, cutting and dividing the black-and-white regional grid images of each number according to the drawn grid dividing lines and numbering them to form a plurality of black-and-white block images with numbers, and using the black-and-white block images as the block annotation base map; Step S6, cutting and segmenting the numbered regional grid color map according to the specifications of the block annotation base map and numbering them to form a number of block color maps with the same number and one-to-one corresponding numbers as the block annotation base map, automatically detecting and marking the damage conditions of the numbered block color maps through the instance segmentation algorithm, and extracting the annotation layer from the numbered block color maps after the damage conditions are marked as the block damage mask map; Step S7, superimposing each numbered block damage mask image onto the corresponding numbered block labeled base image to form the numbered block labeled damage image; Step S8, extract the damage information of each numbered block-marked damage map to generate a damage form, and generate a maintenance work order according to a predetermined format for export or printing by using each numbered area number index map and its included block-marked damage map and damage form.

[0005] Furthermore, in the method for generating maintenance work orders based on spatial image recognition of building damage provided by the present invention, in step S8, each numbered area number index map and each numbered block labeled damage map are scaled to the same scale.

[0006] Furthermore, in the method for generating a maintenance work order based on spatial image recognition of building damage provided by the present invention, in step S6, the instance segmentation algorithm is the YOLACT algorithm.

[0007] Furthermore, in the method for generating a maintenance work order based on spatial image recognition of building damage provided by the present invention, in step S5, the regional grid black-and-white image is automatically cut and segmented into a block-annotated base map through an instance segmentation algorithm.

[0008] Furthermore, the method for generating maintenance work orders based on spatial image recognition of building damage provided by the present invention, in step S4, automatically crops and segments the surface panoramic color image and the surface panoramic black-and-white image through an instance segmentation algorithm and numbers them to form a number of regional grid color images and a number of regional grid black-and-white images with the same number and the same number.

[0009] Furthermore, in the method for generating a maintenance work order based on spatial image recognition of building damage provided by the present invention, the order of step S5 and step S6 is interchanged.

[0010] Furthermore, in the method for generating a maintenance work order based on spatial image recognition of building damage provided by the present invention, in step S3, the panoramic color image is decolorized and hue-balanced to generate a panoramic black-and-white image.

[0011] Furthermore, in the method for generating a maintenance work order based on spatial image recognition of building damage provided by the present invention, in step S2, the outer contour dimensions of the building are measured by a laser rangefinder or a total station.

[0012] Furthermore, in the method for generating a maintenance work order based on spatial image recognition of building damage provided by the present invention, in step S1, an orthographic panoramic color image of each surface of the building is collected by a drone with a camera.

[0013] Furthermore, the method for generating maintenance work orders based on spatial image recognition of building damage provided by the present invention uses a drone with a camera to take photos of each surface of the building, and all photos of each surface of the building are spliced ​​together through a splicing and synthesis algorithm to generate an orthographic panoramic color image of each surface.

[0014] Compared with the prior art, the present invention has the following beneficial effects: The method for generating a repair work order based on spatial image recognition of building damage provided by the present invention performs distortion correction and edge clipping on the acquired panoramic color images of each face of the building according to the actual proportions of each face of the building to form a corrected face panoramic color image, so that the corrected face panoramic color image is consistent with the proportions of each face of the building; converts each face panoramic color image into a face panoramic black-and-white image, clips and divides each face panoramic color image and face panoramic black-and-white image according to the same specification grid size and numbers them to form a regional grid color image and a regional grid black-and-white image with the same number, so that the specification size and quantity of the regional grid color image and the regional grid black-and-white image are consistent; draws a grid segmentation line on each numbered regional grid black-and-white image to form a regional number index image, and clips each numbered regional grid black-and-white image according to the same grid size. The numbered regional grid black-and-white map and the numbered regional grid color maps are respectively cut into a number of block-annotated base maps and a number of block color maps with the same number and corresponding numbers; the damage conditions of the numbered block color maps are automatically detected and annotated through the instance segmentation algorithm, and the annotation layer is extracted as the block damage mask map; the numbered block damage mask map is superimposed on the corresponding numbered block-annotated base map to form a block-annotated damage map with the same number, thereby realizing automatic detection of the damage conditions of various surfaces of the historical building and generating easy-to-identify block-annotated damage maps, avoiding omissions in manual detection, and there is no need to draw engineering drawings. There will be no problem of confusion in the inspection position marking on the engineering drawings, thereby improving the efficiency and accuracy of the inspection of the damage conditions of historical buildings.

[0015] The method for generating maintenance work orders based on spatial image recognition of building damage provided by the present invention extracts damage information from the block-annotated damage map of each number and generates a damage form, generates a maintenance work order according to a predetermined format based on the area number index map of each number and the block-annotated damage map of each number and the damage form contained therein, and exports or prints it, thereby providing a reference basis for the maintenance, restoration and repair of historical buildings through the maintenance work order.

[0016] The method for generating a maintenance work order based on spatial image recognition of building damage provided by the present invention divides the surface panoramic color map and the surface panoramic black-and-white map with damage information on each surface of the building into a plurality of regional grid color maps and a plurality of regional grid black-and-white maps of the same specification and size, and then divides each regional grid color map and regional grid black-and-white map into a plurality of block color maps and a plurality of block annotation base maps of smaller specifications and sizes, and respectively performs damage annotation processing and graphic overlay processing, thereby avoiding the problem of graphics being too large, difficult to calculate and jamming, and improving the calculation speed, thereby improving the inspection efficiency.

[0017] The method for generating a maintenance work order based on spatial image recognition of building damage provided by the present invention extracts the block damage mask map of each numbered block color map after the damage situation is marked and superimposes it on the corresponding numbered block marked base map to form the numbered block marked damage map, so that the damage situation can be clearly, quickly and intuitively observed on the block marked base map, avoiding the problem that the color of the annotation information of the block color map marking the damage situation is consistent with the color of the block color map and cannot be identified and distinguished.

[0018] The method for generating maintenance work orders based on spatial image recognition of building damage provided by the present invention can quickly identify the damage conditions of four facades and one top surface of a historical building for rapid inspection, and can output a regional number index map and a block-annotated damage map. The location of each block-annotated damage map can be found in the regional number index map by the number of the block-annotated damage map, which provides convenience for the maintenance of the building. BRIEF DESCRIPTION OF THE DRAWINGS

[0019] Figure 1 is a flow chart of a method for generating maintenance work orders based on spatial image recognition of building damage; Figure 2 It is an area number index map of a certain area of ​​a certain side map of a building; Figure 3 It is a numbered block-marked damage map of a certain area on a certain side of a building. DETAILED DESCRIPTION

[0020] The present invention is described in detail below in conjunction with the accompanying drawings: The advantages and features of the present invention will become more apparent from the following description. It should be noted that the accompanying drawings are in very simplified form and in non-precise proportions, and are only used to conveniently and clearly assist in explaining the purpose of the embodiments of the present invention.

[0021] Please refer to Figure 1 The embodiment of the present invention provides a method for generating a maintenance work order based on spatial image recognition of building damage, comprising: Step S1, obtaining an orthographic panoramic color image of each face of the building. In order to quickly obtain an orthographic panoramic color image of each face of the building, an orthographic panoramic color image of each face of the building can be collected by a drone with a camera. Specifically, a drone with a camera takes photos of each face of the building, and all photos of each face of the building are spliced ​​together by a splicing synthesis algorithm to generate an orthographic panoramic color image of each face.

[0022] Step S2, measuring the outer contour size of the building to draw the building frame diagram, placing the orthographic projection composite color map of each face into the frame of the corresponding face of the building frame diagram, and performing distortion correction and edge clipping to form a face panoramic color map. In order to draw a building frame diagram with precise size, the outer contour size of the building can be measured by a laser rangefinder or a total station, and the building frame diagram can be drawn according to the measured size information.

[0023] Step S3, converting the surface panoramic color images of each surface into surface panoramic black-and-white images. In order to obtain the surface panoramic black-and-white image, the surface panoramic color image may be subjected to decolorization and hue balancing processing to generate the surface panoramic black-and-white image.

[0024] Step S4, cropping and segmenting each panoramic color map and panoramic black-and-white map according to the same grid size and numbering them to form a number of regional grid color maps and a number of regional grid black-and-white maps with the same number. In order to quickly generate regional grid color maps and regional grid black-and-white maps in batches, an instance segmentation algorithm can be used to automatically crop and segment according to the size specifications of the drawn grid segmentation lines to form a number of regional grid color maps and a number of regional grid black-and-white maps with one-to-one corresponding numbers and the same number.

[0025] Step S5, draw grid dividing lines on the black and white grid images of each numbered area according to the same grid size to form an area number index map, such as Figure 2 As shown, the example segmentation algorithm can be used to automatically cut and segment the black-and-white grid images of each numbered region according to the drawn grid segmentation lines and number them to form a number of black-and-white block images with numbers, and the black-and-white block images are used as the block annotation base map. Figure 2 The example is an A3 drawing with a scale of about 1:100 and a regional number index map with a map length ratio of 4:3. The size specifications of the grid dividing lines include but are not limited to 3.6m×2.4m. Figure 2 The numbers in the table can be automatically numbered by the software. After setting the starting point, the numbers are numbered from left to right and from top to bottom.

[0026] Step S6, each numbered regional grid color map is cut and segmented according to the specifications of the block annotation base map and numbered to form a number of block color maps with the same number and one-to-one corresponding numbers as the block annotation base map, and the damage of each numbered block color map is automatically detected and annotated by the instance segmentation algorithm, and the annotation layer is extracted from each numbered block color map after the damage is annotated as a block damage mask map. The order of step S5 and step S6 can be interchanged. The instance segmentation algorithm includes but is not limited to the YOLACT algorithm.

[0027] Step S7, superimpose each numbered block damage mask image onto the corresponding numbered block annotation base image to form the numbered block annotation damage image, such as Figure 3 As shown. The damage information of the block-marked damage map includes but is limited to the drawing number, legend, damage location, damage type, damage outline and damage area. The damage types include but are not limited to cracking, overgrowth, cracking, defect, alkali efflorescence, etc. Figure 3 Examples of defective situations are shown.

[0028] Step S8, extract the damage information of the block-marked damage map of each number to generate a damage form, and generate a maintenance work order according to a predetermined format for export or printing by using the area number index map of each number and the block-marked damage map and damage form included in the map. In order to facilitate reading of the maintenance work order, the area number index map of each number and the block-marked damage map of each number are scaled to the same scale. Figure 2 When exporting the area number index map, you can scale the map to 1:100 and print it on A3 or A2 sheets according to the actual size of a certain side of the building. Figure 3 The export or print ratio example is 1:10. After testing, the complete damage morphology can be presented on the A3 map. The size of each damage can be viewed through the corresponding form, which is convenient for quickly locating the damage position. It can also be used to count the number of overall damages and compare the severity of damage in different parts, assisting in the rapid formulation of on-site construction plans.

[0029] The method for generating a repair work order based on spatial image recognition of building damage provided by an embodiment of the present invention performs distortion correction and edge cropping on the acquired panoramic color images of each face of the building according to the actual proportions of each face of the building to form a corrected panoramic color image, so that the corrected panoramic color image is consistent with the proportions of each face of the building; converts each panoramic color image into a panoramic black-and-white image, and crops and divides each panoramic color image and panoramic black-and-white image according to the same grid size and numbers them to form a regional grid color image and a regional grid black-and-white image with the same number, so that the specifications and sizes of the regional grid color image and the regional grid black-and-white image are consistent; draws grid segmentation lines on the regional grid black-and-white images with each number to form a regional number index image, and divides each numbered regional grid black-and-white image according to the same grid size. The numbered regional grid black-and-white map and the numbered regional grid color maps are respectively cut into a number of block-annotated base maps and a number of block color maps with the same number and corresponding numbers; the damage conditions of the numbered block color maps are automatically detected and annotated through the instance segmentation algorithm, and the annotation layer is extracted as the block damage mask map; the numbered block damage mask map is superimposed on the corresponding numbered block-annotated base map to form a block-annotated damage map with the same number, thereby realizing automatic detection of the damage conditions of various surfaces of the historical building and generating easy-to-identify block-annotated damage maps, avoiding omissions in manual detection, and there is no need to draw engineering drawings. There will be no problem of confusion in the inspection position marking on the engineering drawings, thereby improving the efficiency and accuracy of the inspection of the damage conditions of historical buildings.

[0030] The method for generating maintenance work orders based on spatial image recognition of building damage provided in an embodiment of the present invention extracts damage information from block-annotated damage maps of each number and generates a damage form, generates a maintenance work order according to a predetermined format for each numbered area number index map and all numbered block-annotated damage maps and damage forms contained therein, and exports or prints them, thereby providing a reference basis for the maintenance and restoration of historical buildings through maintenance work orders.

[0031] The method for generating maintenance work orders based on spatial image recognition of building damage provided by an embodiment of the present invention divides the surface panoramic color map and the surface panoramic black-and-white map with damage information on each surface of the building into a number of regional grid color maps and a number of regional grid black-and-white maps of the same specifications and sizes, and then divides each regional grid color map and regional grid black-and-white map into a number of block color maps and a number of block annotation base maps of smaller specifications and sizes, and respectively performs damage annotation processing and graphic overlay processing, thereby avoiding the problem of graphics being too large, difficult to calculate, and jamming, and improving the calculation speed, thereby improving the inspection efficiency.

[0032] The method for generating a maintenance work order based on spatial image recognition of building damage provided by an embodiment of the present invention extracts the block damage mask map of each numbered block color map after the damage situation is marked and superimposes it on the corresponding numbered block marked base map to form the numbered block marked damage map, so that the damage situation can be clearly, quickly and intuitively observed on the block marked base map, avoiding the problem that the color of the annotation information of the block color map marking the damage situation is consistent with the color of the block color map and cannot be identified and distinguished.

[0033] The method for generating maintenance work orders based on spatial image recognition of building damage provided by an embodiment of the present invention can quickly identify the damage conditions of four facades and one top surface of a historical building for rapid inspection, and can output an area number index map and a block labeled damage map. The location of each block labeled damage map can be found in the area number index map by the number of the block labeled damage map, which provides convenience for the maintenance of the building.

[0034] The method for generating a maintenance work order based on spatial image recognition of building damage provided by an embodiment of the present invention is particularly suitable for situations where there are no drawings and BIM models of historical buildings.

[0035] The present invention is not limited to the above-mentioned specific implementation modes. Obviously, the above-mentioned embodiments are only some embodiments of the embodiments of the present invention, but not all embodiments. Based on the described embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field belong to the scope of protection of the present invention. Those skilled in the art can make other levels of modifications and changes to the present invention. In this way, if these modifications and changes of the present invention fall within the scope of the claims of the present invention, the present invention is also intended to include these changes and changes.

Claims

1. A method for generating a maintenance work order based on spatial image recognition of building damage, characterized in that: include: Step S1, obtaining an orthographic panoramic color image of each face of the building; Step S2, measuring the outer contour size of the building and drawing the building frame diagram, placing the orthographic projection composite color map of each face into the frame of the corresponding face of the building frame diagram, and performing distortion correction and edge clipping to form a corrected face panoramic color map; Step S3, converting the color panoramic images of each surface into black and white panoramic images; Step S4, cutting and dividing each of the surface panoramic color images and the surface panoramic black-and-white images according to the same grid size and numbering them to form a plurality of regional grid color images and a plurality of regional grid black-and-white images with the same number; Step S5, drawing grid dividing lines on the black-and-white regional grid images of each number according to the same grid size to form a regional number index map, cutting and dividing the black-and-white regional grid images of each number according to the drawn grid dividing lines and numbering them to form a plurality of black-and-white block images with numbers, and using the black-and-white block images as the block annotation base map; Step S6, cutting and segmenting the numbered regional grid color map according to the specifications of the block annotation base map and numbering them to form a number of block color maps with the same number and one-to-one corresponding numbers as the block annotation base map, automatically detecting and marking the damage conditions of the numbered block color maps through the instance segmentation algorithm, and extracting the annotation layer from the numbered block color maps after the damage conditions are marked as the block damage mask map; Step S7, superimposing each numbered block damage mask image onto the corresponding numbered block labeled base image to form the numbered block labeled damage image; Step S8, extract the damage information of each numbered block-marked damage map to generate a damage form, and generate a maintenance work order according to a predetermined format for export or printing by using each numbered area number index map and its included block-marked damage map and damage form.

2. The method for generating a maintenance work order based on spatial image recognition of building damage according to claim 1 is characterized in that: In step S8, each numbered area number index map and each numbered block labeled damage map included therein are scaled at the same ratio.

3. The method for generating a maintenance work order based on spatial image recognition of building damage according to claim 1 is characterized in that: In step S6, the instance segmentation algorithm is the YOLACT algorithm.

4. The method for generating a maintenance work order based on spatial image recognition of building damage according to claim 1 is characterized in that: In step S5, the regional grid black-and-white image is automatically cropped and segmented into block-labeled base maps using an instance segmentation algorithm.

5. The method for generating a maintenance work order based on spatial image recognition of building damage according to claim 1 is characterized in that: In step S4, the panoramic color image and the panoramic black-and-white image are automatically cropped and segmented by an instance segmentation algorithm and numbered to form a plurality of regional grid color images and a plurality of regional grid black-and-white images with the same number and the same number.

6. The method for generating a maintenance work order based on spatial image recognition of building damage according to claim 1 is characterized in that: The order of step S5 and step S6 is interchanged.

7. The method for generating a maintenance work order based on spatial image recognition of building damage according to claim 1 is characterized in that: In step S3, the panoramic color image is subjected to decolorization and hue balancing processing to generate a panoramic black-and-white image.

8. The method for generating a maintenance work order based on spatial image recognition of building damage according to claim 1, characterized in that: In step S2, the outer contour dimensions of the building are measured by a laser rangefinder or a total station.

9. The method for generating a maintenance work order based on spatial image recognition of building damage according to claim 1, characterized in that: In step S1, an orthographic panoramic color image of each face of a building is collected by using a drone with a camera.

10. The method for generating a maintenance work order based on spatial image recognition of building damage according to claim 9, characterized in that: Photos of each face of the building are taken by a drone with a camera, and all the photos of each face of the building are stitched together using a stitching synthesis algorithm to generate an orthographic panoramic color image of each face.

Images