Verticality monitoring system and method for tower crane during lifting process based on image recognition
The tower crane verticality monitoring system based on image recognition, utilizing laser emitters and image recognition technology, solves the safety and accuracy issues of tower crane verticality monitoring, and achieves rapid and accurate monitoring of the tower crane lifting process.
Patent Information
- Application Number
- CN202210428280.0
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- Filing Date
- 2022-04-22
- Publication Date
- 2024-11-22
- Estimated Expiration
- 2042-04-22
AI Technical Summary
Existing methods for monitoring the verticality of tower cranes have problems such as high safety and large measurement errors, making it difficult to achieve safe, fast, and accurate monitoring.
A verticality monitoring system based on image recognition is adopted during the tower crane jacking process. It includes components such as a laser transmitter, a wireless signal transmitter, a camera, a laser receiving target, a personal computer, a WIFI router, and a smartphone. The system processes the movement trajectory of the laser point through image recognition and calculates the offset distance during the tower crane jacking process.
It enables safe, fast, and accurate monitoring of the verticality of the tower crane during the jacking process, reduces measurement errors, and improves construction safety.
Smart Images

Figure CN114772467B_ABST
Abstract
Description
Technical Field
[0001] The invention relates to the technical field of verticality monitoring during the tower crane lifting process, and in particular to a system and method for verticality monitoring during the tower crane lifting process based on image recognition. Background Art
[0002] With the expansion of the scale of construction projects and the increase in construction speed, tower cranes (hereinafter referred to as tower cranes) have been used more and more widely. However, due to their high center of gravity and high risk, safety accidents often occur, threatening people's lives and property. With the increasing number of high-rise buildings, the tower bodies under construction are getting higher and higher, and the accompanying dangers are also increasing. In order to improve the safety of tower cranes during construction, the verticality of tower cranes has always been a focus of attention. At present, there are two main methods for monitoring the verticality of tower cranes: the ruler detection method and the theodolite detection method. However, there are certain dangers when using these two methods to detect verticality, and the measurement error is also relatively large. Summary of the invention
[0003] The purpose of the present invention is to propose a verticality monitoring system and method during the tower crane lifting process based on image recognition to solve the technical problems raised in the above background technology, and to be able to safely, quickly and accurately monitor the verticality of the tower crane during the lifting process.
[0004] The technical solution adopted by the present invention to solve the above technical problems is:
[0005] A verticality monitoring system for a tower crane during lifting process based on image recognition, the verticality monitoring system comprising a laser transmitter fixed in the mesh of an iron grid under a tower crane cab, a wireless network signal transmitter fixed on the outside of a main chord of the tower crane, a camera, a laser receiving target placed below the camera, a personal computer, a WIFI router, a wireless network signal transmitter, and a smart phone; the laser receiving target is used to receive the laser rays emitted by the laser transmitter and form a laser spot thereon, the camera is used to shoot the dynamic motion trajectory of the laser spot, the camera is connected to the wireless network signal transmitter via a network cable, the WIFI router is connected to the wireless network signal processor via a network cable, and the smart phone is used to shoot the dynamic process of the tower body during the lifting process of the tower crane.
[0006] Furthermore, the camera is fixed to the upper part of the steel column connecting member through an adjustable fixed support, and the steel column connecting member is fixed on the concrete pedestal.
[0007] Furthermore, the upper portion of the steel column connection member has two bolt openings of the same size, and the lower steel plate of the steel column connection structure has four bolt openings of the same size.
[0008] Furthermore, the laser receiving target has a standard scale and a built-in level bubble calibrator.
[0009] The present invention also provides a method for monitoring verticality during the tower crane lifting process based on image recognition, and the monitoring method comprises the following steps:
[0010] Step 1: Before the tower crane is lifted, the staff turns on the laser transmitter through the remote control switch to ensure that the laser beam is not blocked and the laser point can fall on the laser receiving target. Adjust the zoom adjustment ring at the front end of the laser transmitter so that the laser point can be focused as clear as possible on the laser receiving target.
[0011] Step 2: The smartphone is fixed at a location where it can capture the dynamic process of the tower during the tower crane lifting process, and the time scale in the picture taken by the smartphone is adjusted to be consistent with the time scale in the picture taken by the camera, ensuring that the dynamic motion trajectory of the laser point captured by the camera can reflect the tower crane lifting process captured by the smartphone;
[0012] Step 3: During the lifting process, the time scale in the picture taken by the smartphone is consistent with the time scale in the picture taken by the camera. The smartphone takes pictures of the lifting process of the tower crane. The movement path of the laser point on the laser receiving target reflects the dynamic changes of the tower body offset during the lifting of the tower crane.
[0013] Step 4: Perform image recognition processing on the laser point images taken by the camera. By extracting the pixel coordinates of the center point of the laser point at the same time interval, identify the number of pixel intervals between the laser point and the initial origin in the laser motion path image obtained at the same time interval, that is, extract the coordinates of the center point of the laser point, customize the initial origin of the laser point, and the pixel coordinates of the center of the laser point in the digital image coordinate system are (X 0 , Y 0 ), obtain the laser point motion path image at time t after a certain time interval, and the pixel coordinates of the laser point center in the digital image coordinate system are (X 1 , Y 1 ), the number of pixels between the laser point and the initial origin at the same time interval is Then the actual displacement distance of the tower crane during the lifting process at the corresponding time scale is obtained, and then the verticality of the tower crane at the corresponding moment is calculated.
[0014] Furthermore, in step 4, image recognition processing is performed on the laser point image taken by the camera, specifically, grayscale processing is performed on the color image information of the laser point obtained by the camera, filtering processing is performed to remove noise, image binarization processing is performed, morphological image processing is performed, connected area marking of the target image is performed, and the pixel coordinate mean centroid method is used to obtain the pixel coordinates of the laser center point, and the distance of the laser center point is calculated.
[0015] Furthermore, the grayscale processing of the color image information of the laser point includes converting the color image into a grayscale image first, and applying a grayscale mapping method to the grayscale image. The grayscale mapping is shown as follows:
[0016]
[0017] Where f is the pixel value of the pixel point coordinates, f max , f min are the upper and lower limits of the transformation threshold respectively.
[0018] Furthermore, the image binarization process includes defining a pixel number calculation formula, for an M×N image pixel matrix, where i is defined as the pixel value, and m and n are row coordinates and column coordinates in a pixel coordinate system:
[0019]
[0020]
[0021] Where 0≤i≤255, traverse the pixels in each row, for all pixel values with count(f;i)>0, the maximum pixel value is recorded as i max , the minimum pixel value is recorded as i min , from the minimum pixel value i min Start to obtain the continuous pixel values of count(f; i)>0 in sequence, and record the maximum pixel value of the first segment of continuous pixel values as i 1-max , and then obtain the continuous pixel values of ount(f; i) = 0 in turn, and record the minimum pixel value of the second segment of continuous pixel values as i 2-max , repeat the above process until i k-max =i max , the obtained i k-min As the segmentation threshold for grayscale image binarization, the most appropriate segmentation threshold is determined based on the effect of the binarized image.
[0022] Furthermore, the target image connected area marking includes traversing the matrix pixels row by row starting from the first row of elements in the pixel matrix. When a white pixel is encountered, the pixel is entered into a pixel queue, and then it is determined whether the pixel queue is empty. If it is not empty, the pixel is taken out, and the eight connected domains are selected to determine the eight connected areas of the pixel. If there are also white pixels, the pixel is marked as a laser point to set a label number, and the pixel coordinates of the element are stored in the pixel coordinate queue, and then the pixel is released from the pixel queue, and the traversal continues, and the above process is repeated. BRIEF DESCRIPTION OF THE DRAWINGS
[0023] Figure 1 is a schematic diagram of a verticality monitoring system based on image recognition according to the present invention;
[0024] Figure 2 A schematic diagram of the laser transmitter, wireless bridge, camera and laser receiving target installed on a tower crane according to the present invention;
[0025] Figure 3 It is a schematic diagram of a device for fixing a camera at the bottom of a tower crane according to the present invention;
[0026] Figure 4 This is a schematic diagram of an image in which a laser point is located at a laser receiving light target in a picture taken by a camera of the present invention;
[0027] Figure 5 This is a flow chart of a method for monitoring the verticality of a structure during the tower crane lifting process according to the present invention;
[0028] Figure 6 The present invention proposes an improved recognition method flow chart based on the basic process of image recognition. DETAILED DESCRIPTION
[0029] The present invention will be further explained below in conjunction with the accompanying drawings and specific embodiments. It should be understood that the following specific embodiments are only used to illustrate the present invention and are not used to limit the scope of the present invention.
[0030] To facilitate understanding of the present invention, the present invention will be described more comprehensively below with reference to the relevant drawings. Several embodiments of the present invention are given in the drawings, but the present invention can be implemented in different forms and is not limited to the embodiments described in the text. On the contrary, these embodiments are provided to make the content disclosed in the present invention more thorough and comprehensive.
[0031] For examples, please refer to the attached Figures 1 to 3 A structural verticality monitoring system during the tower crane lifting process based on image recognition is mainly composed of a laser transmitter 1; a wireless network signal transmitter 2; a wireless network signal processor 8; a WIFI router 7; a camera 3; a steel column connecting member 4; a personal computer 6; a laser receiving target 5, and a smart phone 9. The laser transmitter 1 is fixed in the mesh of the iron grid under the tower crane cab, the camera 3 is connected to the wireless network signal transmitter 2 through a network cable, the wireless network signal transmitter 2 is fixed to the outside of the main chord of the tower crane at a suitable height, the WIFI router 7 is connected to the wireless network signal processor 8 through a network cable, the camera 3 is fixed to the upper part of the steel column connecting member 4, the base of the steel column connecting member 4 is fixed to the concrete pedestal, the laser receiving target 5 is placed below the camera 3, and the smart phone 9 is fixed at a suitable location to shoot the dynamic process of the tower body during the tower crane lifting process.
[0032] The laser transmitter 1 is a 400mw remote-controlled waterproof green light spot-shaped zoomable laser transmitter, equipped with a matching remote control switch and a power adapter.
[0033] The camera 3 is a Hikvision 400W home box camera.
[0034] The scale on the laser receiving target 5 is 10 mm / grid, and is provided with a horizontal bubble calibrator 5-1
[0035] The camera 3 is fixed to the upper part of the steel column connecting member 4 through an adjustable fixed support 3-1.
[0036] The upper portion of the steel column connecting member 4 has two bolt openings 4 - 1 of the same size, and the lower steel plate of the steel column connecting structure 4 has four bolt openings 4 - 2 of the same size.
[0037] The wired network signal transmitter 2 is directly opposite to the wireless network signal processor 8, and there is no shielding between the two.
[0038] A method for monitoring the verticality of a structure during the jacking process of a tower crane based on image recognition, the method comprising the following steps:
[0039] Step 1: Before the tower crane is lifted, the staff turns on the laser transmitter through the remote control switch to ensure that the laser beam is not blocked and the laser point can fall on the laser receiving target. Adjust the zoom adjustment ring at the front end of the laser transmitter to focus the laser point into a clear point shape as much as possible.
[0040] Step 2: Fix the smartphone at a suitable location to capture the dynamic process of the tower during the tower crane lifting process, adjust the time scale in the smartphone shooting picture to be consistent with the time scale in the camera shooting picture, and ensure that the dynamic motion trajectory of the laser point captured by the camera can reflect the tower crane lifting process captured by the mobile phone.
[0041] Step 3: During the lifting process, the time scale in the picture taken by the smartphone is consistent with the time scale in the picture taken by the camera. The smartphone shoots the lifting process of the tower crane. The movement path of the laser point on the laser receiving target reflects the dynamic changes of the tower body offset during the lifting of the tower crane.
[0042] Step 4: Perform image recognition processing on the laser point images taken by the camera, and extract the pixel coordinates of the center point of the laser point at the same time interval. Identify the number of pixel intervals between the laser point and the initial origin in the laser motion path image obtained at the same time interval, that is, extract the coordinates of the center point of the laser point, customize the initial origin of the laser point, and the pixel coordinates of the center of the laser point in the digital image coordinate system are (X 0 , Y 0 ), obtain the laser point motion path image at time t after a certain time interval, and the pixel coordinates of the laser point center in the digital image coordinate system are (X 1 , Y 1 ), the number of pixels between the laser point and the initial origin at the same time interval is Then the actual displacement distance of the tower crane during the lifting process at the corresponding time scale is obtained, and then the verticality of the tower crane at the corresponding moment is calculated according to the corresponding formula.
[0043] The method for monitoring the verticality of a structure during the tower crane lifting process based on image recognition obtains color image information of a laser point through a camera, performs grayscale processing on the color image information, performs filtering processing to remove noise, performs image binarization processing, performs morphological image processing, marks connected areas of a target image, obtains the pixel coordinates of the laser center point by using the pixel coordinate mean centroid method, and calculates the distance of the laser center point.
[0044] The grayscale processing of the color image information includes converting the color image into a grayscale image first, and applying a grayscale mapping method to the grayscale image, where f is the pixel value of the pixel point coordinate, and f max , f min are the upper and lower limits of the transformation threshold respectively, and the grayscale mapping is defined as follows:
[0045]
[0046] The image subjected to the noise reduction processing is subjected to image binarization processing, including defining a pixel number calculation formula, for an M×N image pixel matrix, where i is defined as the pixel value size, and m and n are row coordinates and column coordinates in a pixel coordinate system:
[0047]
[0048]
[0049] Where 0≤i≤255, traverse the pixels in each row, for all pixel values of ount(f;i)>0, the maximum pixel value is recorded as i max , the minimum pixel value is recorded as i min , from the minimum pixel value i min Start to obtain the continuous pixel values of ount(f; i)>0 in sequence, and record the maximum pixel value of the first segment of continuous pixel values as i 1-max , and then obtain the continuous pixel values of ount(f; i) = 0 in turn, and record the minimum pixel value of the second segment of continuous pixel values as i 2-max , repeat the above process until i k-max =i max , the obtained i k-min As the segmentation threshold for grayscale image binarization, the most appropriate segmentation threshold is determined based on the effect of the binarized image.
[0050] The marking algorithm includes traversing the matrix pixels row by row starting from the first row of elements in the pixel matrix. When a white pixel is encountered, the pixel is put into the pixel queue, the pixel is taken out, the eight-connected domain is selected, and the eight-connected area of the pixel is determined. If there is also a white pixel, the pixel is marked as a laser point and a label number is set. The pixel coordinates of the element are stored in the pixel coordinate queue, and then the pixel is released from the pixel queue, and the traversal continues, and the above process is repeated.
[0051] The above is only an embodiment of the present invention. Although the present invention describes several embodiments, it is not intended to be limited to these embodiments. For those skilled in the art, the present invention can be improved without departing from the principles of the present invention, and all improvements are included in the scope of the claims of the present invention.
Claims
1. A method for monitoring verticality during the lifting process of a tower crane based on image recognition, and a system for monitoring verticality during the lifting process of a tower crane based on image recognition, the verticality monitoring system comprising a laser transmitter (1) fixed in a mesh of an iron grid under a tower crane cab, a wireless network signal transmitter (2) fixed on the outside of a main chord of the tower crane, a camera (3), a laser receiving target (5) placed below the camera (3), a personal computer (6), a WIFI router (7), a wireless network signal processor (8), and a smart phone (9); the laser receiving target (5) is used to receive laser rays emitted by the laser transmitter (1) and form a laser point (1-1) thereon, the camera (3) is used to shoot the dynamic motion trajectory of the laser point (1-1), the camera (3) is connected to the wireless network signal transmitter (2) via a network cable, the WIFI router (7) is connected to the wireless network signal processor (8) via a network cable, and the smart phone (9) is used to shoot the dynamic process of the tower body during the lifting process of the tower crane; The camera (3) is fixed to the upper part of the steel column connecting member (4) via an adjustable fixed support (3-1), and the steel column connecting member (4) is fixed to the concrete pedestal; The upper portion of the steel column connecting member (4) has two bolt openings (4-1) of the same size, and the lower steel plate of the steel column connecting member (4) has four bolt openings (4-2) of the same size; The laser receiving target (5) has a standard scale and a built-in horizontal bubble calibration instrument (5-1), and is characterized in that: The monitoring method comprises the following steps: Step 1: Before the tower crane is lifted, the staff turns on the laser transmitter (1) through the remote control switch to ensure that the laser beam is not blocked and the laser point (1-1) can fall on the laser receiving target (5). The zoom adjustment ring at the front end of the laser transmitter (1) is adjusted so that the laser point (1-1) is focused as clearly as possible in a dot shape on the laser receiving target (5); Step 2: The smart phone (9) is fixed at a location where it can capture the dynamic process of the tower body during the tower crane lifting process, and the time scale in the picture captured by the smart phone (9) is adjusted to be consistent with the time scale in the picture captured by the camera, so as to ensure that the dynamic motion trajectory of the laser point captured by the camera (3) can reflect the tower crane lifting process captured by the smart phone (9); Step 3: During the lifting process, the time scale in the picture taken by the smart phone (9) is consistent with the time scale in the picture taken by the camera (3), the smart phone (9) takes pictures of the tower crane lifting process, and the movement path of the laser point (1-1) on the laser receiving target (5) reflects the dynamic change of the tower body displacement during the tower crane lifting process as the tower crane is working; Step 4: Perform image recognition processing on the laser point image taken by the camera, extract the pixel coordinates of the center point of the laser point at the same time interval, identify the number of pixel intervals between the laser point and the initial origin in the laser motion path image obtained at the same time interval, that is, extract the coordinates of the center point of the laser point, customize the initial origin of the laser point, the pixel coordinates of the center of the laser point in the digital image coordinate system are (X0, Y0), obtain the laser point motion path image at time t after a certain time interval, the pixel coordinates of the center of the laser point in the digital image coordinate system are (X1, Y1), and the number of pixel intervals between the laser point and the initial origin at the same time interval is Then the actual displacement distance of the tower crane during the lifting process at the corresponding time scale is obtained, and then the verticality of the tower crane at the corresponding moment is calculated.
2. A method for monitoring verticality during the tower crane lifting process based on image recognition as described in claim 1, characterized in that, in step 4, the laser point picture taken by the camera is subjected to image recognition processing, specifically, the color image information of the laser point obtained by the camera is grayed, filtered to remove noise, image binarized, morphological image processed, connected area of the target image is marked, pixel coordinate mean centroid method is used to obtain the pixel coordinates of the laser center point, and the distance of the laser center point is calculated.
3. A method for monitoring verticality of a tower crane during its lifting process based on image recognition as claimed in claim 2, characterized in that: The grayscale processing of the color image information of the laser point includes converting the color image into a grayscale image first, and applying a grayscale mapping method to the grayscale image. The grayscale mapping formula is as follows: Where f is the pixel value of the pixel point coordinates, f max , f min are the upper and lower limits of the transformation threshold respectively.
4. A method for monitoring verticality of a tower crane during its lifting process based on image recognition as claimed in claim 3, characterized in that: The image binarization process includes defining a pixel number calculation formula, for an M×N image pixel matrix, where i is defined as the pixel value, and m and n are the row coordinates and column coordinates in the pixel coordinate system: Where 0≤i≤255, traverse the pixels in each row, for all pixel values with count(f;i)>0, the maximum pixel value is recorded as i max , the minimum pixel value is recorded as i min , from the minimum pixel value i min Start to obtain the continuous pixel values of count(f; i)>0 in sequence, and record the maximum pixel value of the first segment of continuous pixel values as i 1-max , and then obtain the continuous pixel values of count(f; i) = 0 in turn, and record the minimum pixel value of the second segment of continuous pixel values as i 2-max , repeat the above process until i k-max =i max , the obtained i k-min As the segmentation threshold for grayscale image binarization, the most appropriate segmentation threshold is determined based on the effect of the binarized image.
5. A method for monitoring verticality of a tower crane during its lifting process based on image recognition as claimed in claim 4, characterized in that: The target image connected area marking includes traversing the matrix pixels row by row starting from the first row of elements in the pixel matrix. When a white pixel is encountered, the pixel is entered into a pixel queue, and then it is determined whether the pixel queue is empty. If it is not empty, the pixel is taken out, and the eight connected domains are selected to determine the eight connected areas of the pixel. If there are also white pixels, the pixel is marked as a laser point setting label number, and the coordinates of the pixel are stored in the pixel coordinate queue. Then, the pixel is released from the pixel queue, and the traversal continues, and the above process is repeated.
