Positioning method of reserved openings in indoor concrete structures

Abstract

The present invention relates to a method for locating reserved openings in indoor concrete structures, comprising: obtaining an indoor construction site inspection video and the camera pose information corresponding to each frame image in the video; performing target detection on the openings in each frame image and marking the openings with a frame; performing an affine transformation based on the camera pose information and the frame mark of the corresponding opening to determine the orientation and depression angle of the center position of the frame mark relative to the camera; setting the openings with a depression angle greater than 17 degrees as vertical openings and screening images with vertical openings; determining the three-dimensional coordinates of the vertical openings in the images with vertical openings in the three-dimensional world based on the camera pose information; and determining the three-dimensional position information corresponding to the vertical openings based on the three-dimensional coordinates. The present invention solves the technical problem of safety hazards posed by vertical openings in concrete structures in the prior art.

Description

Technical Field

[0001] The present invention relates to the field of building construction, and in particular to a method for positioning a reserved opening indoor in a concrete structure. Background Art

[0002] During construction of the main structure at a construction site, after the cast-in-place concrete structure is completed and the formwork support system is removed, a large number of vertical openings are left in the finished concrete structure. If these openings are not protected, there is a risk of falls from height. Construction site safety management generally involves on-site managers conducting special inspections and arranging for workers to install safety protection facilities to ensure that there are no such safety hazards. However, the complex construction site environment makes it difficult for on-site managers to conduct comprehensive inspections, and existing protective facilities are subject to damage, which increases the risk of project safety management. Summary of the Invention

[0003] In view of the above-mentioned shortcomings in the prior art, the present invention provides a method for positioning a reserved opening in a concrete structure, which solves the technical problem that the vertical openings in the concrete structure in the prior art have potential safety hazards.

[0004] The present invention discloses a method for positioning reserved openings in indoor concrete structures, comprising: obtaining an inspection video of an indoor construction site and camera posture information corresponding to each frame image in the video; performing target detection on the openings in each frame image and marking the openings with a frame; performing affine transformation according to the camera posture information and the frame marks of the corresponding openings, and determining the orientation and depression angle of the center position of the frame marks relative to the camera; setting the openings with a depression angle greater than 17 degrees as vertical openings, and screening images with vertical openings; determining the three-dimensional coordinates of the vertical openings in the images with the vertical openings in the three-dimensional world according to the camera posture information; and determining the three-dimensional position information corresponding to the vertical openings according to the three-dimensional coordinates.

[0005] The method for locating a reserved opening in a concrete structure is further improved in that the step of determining the three-dimensional coordinates of the vertical opening in the three-dimensional world includes:

[0006] a) Determine the equation of a line from the center of the vertical opening to the camera in three-dimensional space using an image having a vertical opening, the position of the vertical opening in the image, the camera pose information corresponding to the image, and the orientation information of the vertical opening relative to the camera:

[0007]

[0008] Where (x0, y0, z0) are the three-dimensional coordinates of the camera, (a, b, c) are the direction vector of the hole center relative to the camera, n is any value greater than 0, indicating that its direction is from the camera to the hole, and (x, y, z) are the three-dimensional coordinates of the vertical hole center.

[0009] b) According to equation (1) and the surface equation of the inspection floor in three-dimensional space: Ax+By+Cz+D=0 (2), the three-dimensional coordinates (x, y, z) of the center of the opening in three-dimensional space are determined as follows:

[0010]

[0011] Among them, A, B, C, and D are all known parameters;

[0012] c) Repeat steps a) and b) above until the three-dimensional coordinates of all vertical openings in three-dimensional space are obtained.

[0013] The method for locating a reserved opening in a concrete structure is further improved in that the step of determining the three-dimensional position information corresponding to the vertical opening according to the three-dimensional coordinates includes:

[0014] d) performing density-based clustering on the three-dimensional coordinates of the same vertical opening in all images containing vertical openings, and obtaining the final three-dimensional coordinates as the three-dimensional position information of the vertical opening;

[0015] e) Repeat step d) until the three-dimensional position information of all vertical openings is obtained.

[0016] The method for positioning a reserved opening in an indoor concrete structure of the present invention is further improved in that the camera position information includes the camera's position coordinate information and the camera's angle posture information.

[0017] The method for locating reserved openings in indoor concrete structures of the present invention is further improved in that a method for target detection of openings in each frame of image includes the steps of: performing digital image processing on the acquired video, detecting all openings in the video and corresponding them to each frame of image; performing digital image processing on each acquired frame of image, and directly detecting the openings in each frame of image.

[0018] Compared to existing technologies, the present invention offers significant and positive results. By detecting, classifying, and locating openings, the present invention addresses the existing technical issues of vertical openings in concrete structures, which pose safety risks. The present invention is portable and offers high detection and positioning accuracy. By carrying this device, managers can quickly identify hidden dangers during routine inspections, effectively improving on-site safety management and preventing potential safety incidents. BRIEF DESCRIPTION OF THE DRAWINGS

[0019] In order to more clearly illustrate the technical solutions in the embodiments of the present invention, the following briefly introduces the drawings required for use in the description of the embodiments. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work.

[0020] Figure 1 A framework diagram of the method for positioning a reserved opening in a concrete structure interior according to the present invention.

[0021] Figure 2 A schematic diagram of a method for locating a reserved opening in a concrete structure according to the present invention, which uses affine transformation to determine the orientation and depression angle of the opening relative to a camera.

[0022] Figure 3 This is a schematic diagram of a method for locating a reserved opening in a concrete structure indoor according to the present invention, which utilizes a depression angle to determine a vertical opening within a 5-meter working radius.

[0023] Figure 4 The present invention is a schematic diagram of a method for locating a reserved opening in a concrete structure indoors, for determining the spatial coordinates of the center of the opening in a certain image or a certain frame.

[0024] Figure 5 This is a schematic diagram of the vertical opening center coordinates before density-based clustering of the positioning method for reserved openings in indoor concrete structures of the present invention.

[0025] Figure 6 This is a schematic diagram of the vertical opening center coordinates after density-based clustering in the method for locating indoor reserved openings in concrete structures of the present invention.

[0026] Figure 7 This is a hardware architecture diagram of the method for locating a reserved opening in a concrete structure indoors according to the present invention.

[0027] Figure 8 This is a schematic diagram of information transmission for the method for positioning a reserved opening in a concrete structure indoors according to the present invention. DETAILED DESCRIPTION

[0028] The following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the accompanying drawings. 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 making creative efforts are within the scope of protection of the present invention.

[0029] like Figure 1As shown, the present invention provides a method for locating reserved openings in indoor concrete structures, including: obtaining an indoor construction site inspection video and camera pose information corresponding to each frame image in the video; performing target detection on the openings in each frame image and marking the openings with a frame; performing affine transformation based on the camera pose information and the frame mark of the corresponding opening to determine the orientation and depression angle of the center position of the frame mark relative to the camera, such as Figure 2 As shown in the figure, reference numeral 1 is a camera with a known posture, reference numeral 2 is an image or video frame, reference numeral 3 is the center of the hole in the image or video frame, reference numeral 4 is the depression angle of the camera, and reference numeral 5 is the azimuth ray of the hole center determined by anti-radiation transformation; the hole with a depression angle greater than 17 degrees is set as a vertical hole, and images with vertical holes are screened, such as Figure 3 As shown in the figure, reference numeral 6 is a depression angle of 17 degrees, reference numeral 7 is a vertical hole, reference numeral 8 is a camera, reference numeral R is a vertical hole detection radius of 5 meters, and reference numeral H is a camera working height of approximately 1.5 meters; according to the camera posture information, the three-dimensional coordinates of the vertical hole in the image with the vertical hole in the three-dimensional world are determined respectively; according to the three-dimensional coordinates, the three-dimensional position information corresponding to the vertical hole is determined, and the information is output.

[0030] In indoor construction sites, in addition to vertical openings, there are also horizontal openings. This method can be used to remove horizontal openings and only screen out vertical openings. The present invention aims to solve the risk problem of dangerous openings exposed at construction sites. By carrying this device during daily inspections, dangerous openings that have not been discovered by inspectors can be quickly and accurately located, thereby improving the level of on-site safety management. The method is mainly implemented through three steps: detection of openings, screening and classification of openings, and positioning of openings. For the opening detection step, the openings are non-a priori and are completely acquired through the target detection algorithm; for the opening classification step, the purpose is to classify the openings detected in the previous step into horizontal openings and vertical openings, because unprotected vertical openings constitute dangerous openings; for the opening positioning step, the purpose is to accurately locate the targets identified and classified as vertical openings in the previous step in three-dimensional space, so that management personnel can find unprotected vertical openings. In the present invention, the location of the hole is unknown and needs to be located by the method and device proposed in this solution. The object to be identified is the hole itself, and the object to be monitored is all the holes in the construction project. In addition, the method of the present invention needs to be implemented through daily inspections by management personnel, and the camera is mobile and dynamic. The equipment used in the present invention needs to comprehensively use target detection, SLAM, machine learning and other methods, as well as the classification and positioning algorithms proposed in this solution to achieve the proposed function. This solution uses a mobile camera and a mobile edge computing device, and the internal parameters of the camera are known. The process of solving the problem is target identification, classification, and positioning. The data obtained by the equipment used in the present invention needs to be deeply processed, and then a variety of more complex algorithms such as target detection, SLAM, machine learning, and the classification and positioning algorithms proposed in the solution are comprehensively used to achieve the final function.

[0031] Preferably, the step of determining the three-dimensional coordinates of the vertical opening in the three-dimensional world includes:

[0032] a) Determine the equation of a line from the center of the vertical opening to the camera in three-dimensional space using an image having a vertical opening, the position of the vertical opening in the image, the camera pose information corresponding to the image, and the orientation information of the vertical opening relative to the camera:

[0033]

[0034] Where (x0, y0, z0) are the three-dimensional coordinates of the camera, (a, b, c) are the direction vector of the hole center relative to the camera, n is any value greater than 0, indicating that its direction is from the camera to the vertical hole, and (x, y, z) are the three-dimensional coordinates of the vertical hole center.

[0035] b) Based on equation (1) and the surface equation of the inspection floor in three-dimensional space Ax+By+Cz+D=0(2), perform simultaneous calculations. Taking the calculation of y as an example, the simultaneous calculation process is as follows:

[0036]

[0037] Similarly, the analytical expressions of x and z can be derived to determine the three-dimensional coordinates (x, y, z) of the center of the hole in three-dimensional space:

[0038]

[0039] Among them, A, B, C, and D are all known parameters, and the surface equation is assumed to be known (the elevation of the structural surface or building surface of a certain building or layer is a known value). Figure 4 As shown in the figure, reference numeral 5 is the azimuth ray of the center of the cave entrance determined by anti-radiation transformation, reference numeral 9 is the ground, and reference numeral 10 is the intersection of the azimuth ray of the center of the cave entrance determined by anti-radiation transformation and the ground.

[0040] c) Repeat steps a) and b) above until the three-dimensional coordinates of all vertical openings in three-dimensional space are obtained.

[0041] Preferably, the step of determining the three-dimensional position information of the corresponding vertical opening according to the three-dimensional coordinates includes: d) performing density-based clustering on the three-dimensional coordinates of the same vertical opening in all images containing vertical openings to obtain the final three-dimensional coordinates as the three-dimensional position information of the vertical opening; e) repeating step d) until the three-dimensional position information of all vertical openings is obtained. The above-mentioned density-based clustering method includes all publicly available density-based clustering methods, including but not limited to the DBSCAN algorithm, the OPTICS algorithm, the DENCLUE algorithm, etc. Specifically, Figure 5 As shown in the figure, the center point of the hole determined after steps a) and b) is shown, and the reference numeral 9 is the ground; Figure 6 The center point of the hole determined after the clustering operation is shown, and the reference numeral 9 represents the ground.

[0042] Preferably, the camera pose information includes the camera's position coordinate information and the camera's angular pose information. The above-mentioned video and camera pose information are obtained by the device's built-in camera, and the camera pose information is obtained by the camera's built-in IMU. The device includes a fixed unit, a camera, an edge computing device, a data transmission device, a speaker, and a battery pack. Simply put, an edge computing device is a small, portable computer that can perform simple calculations, such as the Jetson Nano produced by NVIDIA; the data transmission device is used to transmit data to a remote end, and includes a 4G / 5G module, a USB communication unit, and the like.

[0043] Preferably, the method for detecting the hole in each frame of the image comprises the steps of: performing digital image processing on the acquired video, detecting all the holes in the video and corresponding them to each frame of the image; performing digital image processing on each acquired frame of the image, and directly detecting the hole in each frame of the image. The video pre-processing method is selected according to the hole target detection model, and the processing method includes but is not limited to various digital image processing methods such as image compression, noise filtering, edge extraction, and image sharpening, and its goal is to make the original video data meet the requirements of the target detection model; the above-mentioned video pre-processing process is completed by the edge computing device provided by the device. The objects of the hole target detection include videos, and also include any frame images in the video selected by any sampling method; the above-mentioned target detection is realized by the target detection model; the above-mentioned target detection model is realized by training the target data set through various types of deep learning methods.

[0044] like Figure 7 and Figure 8 As shown, the three-dimensional coordinate information of all the above-mentioned openings is collected, and the collected information is structured to output structured data; in the above-mentioned processing process, the calculation is completed by the edge computing device built into the device, and the data transmission is completed by the data transmission device built into the device.

[0045] The present invention is applied in the main structure construction phase, which can effectively reduce the risk of falling from heights in the construction site and avoid accidents caused by falling from heights. The present invention solves the technical problem of safety hazards in vertical openings in concrete structures in the prior art by detecting, classifying and locating openings. The present invention has the characteristics of being easy to carry and use, and having high detection and positioning accuracy. By carrying this equipment, management personnel can quickly discover hidden dangers in openings during daily inspections, effectively improving the level of on-site safety management. By avoiding possible safety accidents. The present invention has a high degree of standardization, is easy to assemble, is compact and portable, simple to use, has a low learning cost, and is easy to promote in actual management applications.

[0046] The parts not mentioned in the present invention are the same as the existing technology or can be implemented by using the existing technology. The above description is only the preferred embodiment of the present invention and does not limit the present invention in any form. Although the present invention has been disclosed as the preferred embodiment as above, it is not used to limit the present invention. Any technician familiar with this profession can make some changes or modifications to the equivalent embodiments of equivalent changes by using the technical content disclosed above without departing from the scope of the technical solution of the present invention. However, any simple modification, equivalent change and modification made to the above embodiment based on the technical essence of the present invention without departing from the content of the technical solution of the present invention still falls within the scope of the technical solution of the present invention.

Claims

1. A method for positioning a reserved opening in a concrete structure, characterized in that: include: Obtain inspection videos of indoor construction sites and the camera pose information corresponding to each frame of the video; Perform target detection on the hole in each frame image and mark the hole with a frame; Performing an affine transformation based on the camera pose information and the frame mark of the corresponding hole to determine the orientation and depression angle of the center position of the frame mark relative to the camera; The openings with a depression angle greater than 17 degrees are set as vertical openings, and images with vertical openings are filtered; determining the three-dimensional coordinates of the vertical openings in the image having the vertical openings in the three-dimensional world according to the camera pose information; Determining the three-dimensional position information of the corresponding vertical opening according to the three-dimensional coordinates; The steps to determine the 3D coordinates of a vertical opening in the 3D world include: a) Determine, using an image having a vertical opening, the position of the vertical opening in the image, camera pose information corresponding to the image, and orientation information of the vertical opening relative to the camera, an equation of a line from the center of the vertical opening to the camera in three-dimensional space: Where (x0, y0, z0) are the three-dimensional coordinates of the camera, (a, b, c) are the direction vector of the hole center relative to the camera, n is any value greater than 0, indicating that its direction is from the camera to the vertical hole, and (x, y, z) are the three-dimensional coordinates of the vertical hole center. b) According to equation (1) and the surface equation of the inspection floor in three-dimensional space: Ax+By+Cz+D=0 (2), the three-dimensional coordinates (x, y, z) of the center of the opening in three-dimensional space are determined as follows: Among them, A, B, C, and D are all known parameters; c) repeating the above steps a) and b) until the three-dimensional coordinates of all vertical openings in three-dimensional space are obtained; The step of determining the three-dimensional position information of the corresponding vertical opening according to the three-dimensional coordinates includes: d) performing density-based clustering on the three-dimensional coordinates of the same vertical opening in all images containing the vertical opening, and obtaining final three-dimensional coordinates as the three-dimensional position information of the vertical opening; e) Repeat step d) until the three-dimensional position information of all vertical openings is obtained.

2. The method for positioning a reserved opening in a concrete structure according to claim 1, characterized in that: The camera position information includes the camera's position coordinate information and the camera's angle posture information.

3. The method for positioning a reserved opening in a concrete structure according to claim 1, characterized in that: The method for detecting a target hole in each frame of an image comprises the steps of: performing digital image processing on an acquired video, detecting all holes in the video and corresponding them to each frame of the image; Each frame of the image is digitally processed to directly detect the hole in each frame.

Images