Automatic elevator guide rail measuring system based on binocular vision

Abstract

The invention discloses an elevator guide rail automatic measurement system based on binocular vision, and relates to the technical field of elevator guide rail measurement. The laser device is used for precisely controlling the unmanned aerial vehicle to fly along the guide rail in the elevator shaft, meanwhile, the two cameras collect image data of the guide rail, and the data are transmitted to the visual image processing module through the wireless transmission module; in the image processing stage, the system recognizes and removes noise in an image through a size adjustment technology, optimizes image data through a gray level conversion method, reduces the difficulty of subsequent processing, positions feature points of the guide rail through multiple algorithms, and finely describes surface textures of the guide rail to form complete feature information of the guide rail. The deep learning algorithm is utilized, the system matches and calculates the collected images, accurate measurement of the elevator guide rail is achieved, the result is stored and displayed on the operation terminal in real time, and detection personnel can conveniently and visually know the state of the elevator guide rail.

Description

Technical Field

[0001] The invention relates to the technical field of elevator guide rail measurement, and in particular to an elevator guide rail automatic measurement system based on binocular vision. Background Art

[0002] In the field of modern construction, elevators are key equipment for vertical transportation, and their safety and stability are of vital importance. Elevator guide rails are precise guide components for the operation of elevator cars. The installation accuracy of the guide rails directly determines the smoothness of the elevator operation and the comfort of passengers, and is closely related to the service life and maintenance cost of the elevator.

[0003] Traditional elevator guide rail measurement methods mainly rely on manual operation. Inspectors need to carry calipers, levels and other tools to enter the elevator shaft and measure the guide rails section by section. This manual measurement method has many disadvantages: on the one hand, the elevator shaft environment is usually narrow and dim, and there is a certain amount of dust pollution, which brings great inconvenience to manual operation. Not only is the work intensity high, but the inspectors are also prone to fatigue in this harsh environment for a long time, which increases the risk of measurement errors; on the other hand, manual measurement is seriously affected by subjective factors. Different inspectors have different operating methods and measurement experience, resulting in uneven measurement accuracy, making it difficult to ensure the accuracy and consistency of each measurement result. Moreover, manual measurement efficiency is extremely low. For the measurement of elevator guide rails in high-rise buildings, it often takes a lot of time and manpower, which seriously slows down the overall progress of elevator installation and maintenance.

[0004] Based on the above reasons, the present invention proposes an automatic measurement system for elevator guide rails based on binocular vision. Summary of the invention

[0005] 1. Technical issues to be solved

[0006] In view of the shortcomings of the prior art, the present invention discloses an automatic measurement system for elevator guide rails based on binocular vision. The system collects elevator guide rails through two high-precision industrial cameras installed on a drone, and uses a visual image processing module to analyze and process the data, thereby realizing automatic measurement of various parameters such as the position and size of the elevator guide rails; the present invention uses a laser device to accurately control the drone to fly along the guide rails inside the elevator shaft, and at the same time, the two cameras collect image data of the guide rails, and the data are transmitted to the visual image processing module through a wireless transmission module; in the image processing stage, the system first identifies and removes noise in the image through a size adjustment technology, and optimizes the image data using a grayscale conversion method to reduce the difficulty of subsequent processing, then uses SURF and SIFT algorithms to quickly locate the feature points of the guide rails, and combines the local binary pattern texture descriptor to finely describe the surface texture of the guide rails to form complete guide rail feature information; finally, using a deep learning algorithm, the system matches and calculates the collected images to realize accurate measurement of the elevator guide rails, and stores and displays the results in real time on an operation terminal, so that the inspection personnel can intuitively understand the state of the elevator guide rails, and effectively improve the convenience and accuracy of measuring the elevator guide rails.

[0007] (II) Technical solution

[0008] To achieve the above object, the present invention provides the following technical solution: an automatic measurement system for elevator guide rails based on binocular vision, comprising:

[0009] Binocular vision acquisition module, used to realize the vision-based acquisition operation of elevator guide rail data, consists of two high-precision industrial cameras and is symmetrically installed on the measurement bracket;

[0010] The visual data transmission module is used to transmit the collected elevator guide rail data, realize the connection between the binocular vision acquisition module and the visual image processing module, and ensure the integrity and timeliness of the data;

[0011] The visual image processing module is used to analyze and process the collected binocular elevator guide rail image data, thereby completing the automatic measurement of various data of the elevator guide rail;

[0012] The data storage and output module is used to store the measured elevator guide rail parameters for subsequent query and analysis, and to output key data in real time and display it on the operation terminal so that the inspection personnel can intuitively understand the status of the elevator guide rails.

[0013] Preferably, the binocular vision acquisition module includes the following steps when acquiring data from the elevator guide rail:

[0014] S11: two high-precision industrial cameras are fixedly mounted on the measuring bracket, the measuring bracket is fixedly mounted on the drone, a laser receiving device is installed on the top surface of the measuring bracket, and a laser emitting device is installed at a designated position on the top surface of the elevator shaft;

[0015] S12: During the measurement, the drone is started and placed inside the elevator shaft, and the laser emitted by the laser emitting device is aligned with the laser receiving device, and the flight of the drone is precisely controlled based on the signal feedback from the laser receiving device;

[0016] S13: The drone flies along the elevator rails inside the elevator shaft, allowing two high-precision industrial cameras to complete visual data collection operations on the elevator rails.

[0017] Preferably, the visual data transmission module and the binocular vision acquisition module are connected by wireless transmission, and the visual data transmission module and the visual image processing module are connected by wireless or wired transmission.

[0018] Preferably, the implementation of the visual image processing module comprises the following steps:

[0019] S21: setting an image window of an initial size, performing local window analysis on each pixel in the image, judging the noise situation by counting the grayscale value distribution of the pixels in the window, and gradually expanding the window until a suitable window is found so that the grayscale value distribution of the pixels in the window meets certain uniformity conditions if it is found that the grayscale value difference between the central pixel and the surrounding pixels is too large;

[0020] S22: Convert the elevator guide rail image data into a grayscale image to reduce the data dimension and calculation amount of subsequent processing. For any pixel point (P(x,y)) in the image, accurately assign different weights to its red, green, and blue components, and calculate its grayscale value (Gray(x,y)) by the formula (Gray(x,y)=a×R(x,y)+b×G(x,y)+c×B(x,y)), where (R(x,y)), (G(x,y)), and (B(x,y)) are the red, green, and blue component values ​​of the pixel point, respectively.

[0021] S23: Use the SURF algorithm to quickly screen feature points. Based on the extreme point detection principle of the Hessian matrix determinant, it can quickly locate possible feature points in the image. Its detection speed is greatly improved compared with traditional methods, and it can cover a large area of ​​the image in a short time.

[0022] S24: For the feature points initially screened, the SIFT algorithm is used for precise positioning. By constructing a Gaussian difference pyramid, the feature points are detected in a multi-scale space, which can effectively overcome the influence of changes such as image scaling and rotation, and ensure the accurate position of the feature points. At the same time, for the continuous texture area on the surface of the guide rail, the local binary pattern texture descriptor is used for feature description. By comparing the grayscale values ​​of the central pixel and the neighboring pixels, a binary code is generated to characterize the texture features. It has strong robustness to changes in illumination, thereby comprehensively capturing various features of the guide rail and providing rich and accurate information for subsequent stereo matching.

[0023] S25: Based on deep learning, various parameters of the elevator guide rails are matched and calculated on the image, so that the automatic measurement operation of the elevator guide rails can be accurately completed.

[0024] Preferably, the initial window size in step S21 is set to 4×4, and during the analysis and processing, priority is always given to retaining the edge features of the guide rail.

[0025] Preferably, in step S22, different weights are assigned to the red, green and blue components based on the fact that the human eye is most sensitive to green, followed by red and relatively less sensitive to blue, thereby effectively highlighting the green component to which the human eye is sensitive. The converted grayscale image performs better in retaining the surface texture information of the elevator guide rail, and compared with the traditional average weighted grayscale, it successfully reduces unnecessary color information redundancy.

[0026] Preferably, the green component (G(x, y)) is weighted to (0.6), the red component (R(x, y)) is weighted to (0.25), and the blue component (B(x, y)) is weighted to (0.15).

[0027] Preferably, the data storage and output module outputs key data on the operating terminal including the position of the elevator guide rail, measured items, measured data, collected images and processed images.

[0028] (III) Beneficial effects

[0029] Compared with the prior art, the present invention provides an automatic measurement system for elevator guide rails based on binocular vision, which has the following beneficial effects: The present invention discloses an automatic measurement system for elevator guide rails based on binocular vision, which collects data on elevator guide rails by means of two high-precision industrial cameras installed on a drone, and uses a visual image processing module to analyze and process the data, thereby realizing automatic measurement of various parameters such as the position and size of the elevator guide rails; the present invention uses a laser device to accurately control the drone to fly along the guide rails inside the elevator shaft, and at the same time, two cameras collect image data of the guide rails, and these data are transmitted to the visual image processing module via a wireless transmission module; in FIG. In the image processing stage, the system first uses resizing technology to identify and remove noise from the image, and uses grayscale conversion methods to optimize image data to reduce the difficulty of subsequent processing. Then, the SURF and SIFT algorithms are used to quickly locate the feature points of the guide rail, and the local binary pattern texture descriptor is used to describe the surface texture of the guide rail in detail to form complete guide rail feature information. Finally, using the deep learning algorithm, the system matches and calculates the collected images to achieve accurate measurement of the elevator guide rails, and stores and displays the results in real time on the operation terminal, which makes it convenient for inspection personnel to intuitively understand the status of the elevator guide rails, effectively improving the convenience and accuracy of measuring the elevator guide rails. BRIEF DESCRIPTION OF THE DRAWINGS

[0030] Figure 1 It is a schematic diagram of the system structure of the present invention;

[0031] Figure 2 This is a flow chart of the binocular vision acquisition module of the present invention acquiring data on elevator guide rails;

[0032] Figure 3 This is a flow chart for implementing the visual image processing module of the present invention. DETAILED DESCRIPTION

[0033] In order to better understand the purpose, structure and function of the present invention, and to achieve efficient and accurate measurement of elevator guide rails, the present invention discloses an automatic measurement system for elevator guide rails based on binocular vision. The system collects elevator guide rails by two high-precision industrial cameras installed on a drone, and uses a visual image processing module to analyze and process the data, thereby achieving automatic measurement of various parameters such as the position and size of the elevator guide rails; the present invention uses a laser device to accurately control the drone to fly along the guide rails inside the elevator shaft, and at the same time, the two cameras collect image data of the guide rails, and these data are transmitted to the visual image processing module through a wireless transmission module; in the image processing In the processing stage, the system first uses the size adjustment technology to identify and remove the noise in the image, and uses the grayscale conversion method to optimize the image data to reduce the difficulty of subsequent processing. Then, the SURF and SIFT algorithms are used to quickly locate the feature points of the guide rail, and the local binary pattern texture descriptor is combined to finely describe the surface texture of the guide rail to form complete guide rail feature information. Finally, using the deep learning algorithm, the system matches and calculates the collected images to achieve accurate measurement of the elevator guide rails, and stores and displays the results in real time on the operation terminal, which is convenient for the inspection personnel to intuitively understand the status of the elevator guide rails, and effectively improves the convenience and accuracy of the measurement of the elevator guide rails. The present invention is a further detailed description of an automatic measurement system for elevator guide rails based on binocular vision.

[0034] refer to Figure 1-3 , the present invention: an automatic measurement system for elevator guide rails based on binocular vision, comprising:

[0035] Binocular vision acquisition module, used to realize the vision-based acquisition operation of elevator guide rail data, consists of two high-precision industrial cameras and is symmetrically installed on the measurement bracket;

[0036] Specifically, when the binocular vision acquisition module acquires data from the elevator guide rail, the steps include:

[0037] S11: two high-precision industrial cameras are fixedly mounted on the measuring bracket, the measuring bracket is fixedly mounted on the drone, a laser receiving device is installed on the top surface of the measuring bracket, and a laser emitting device is installed at a designated position on the top surface of the elevator shaft;

[0038] S12: During the measurement, the drone is started and placed inside the elevator shaft, and the laser emitted by the laser emitting device is aligned with the laser receiving device, and the flight of the drone is precisely controlled based on the signal feedback from the laser receiving device;

[0039] S13: The drone flies along the elevator rails inside the elevator shaft, allowing two high-precision industrial cameras to complete visual data collection operations on the elevator rails.

[0040] The visual data transmission module is used to transmit the collected elevator guide rail data, realize the connection between the binocular vision acquisition module and the visual image processing module, and ensure the integrity and timeliness of the data;

[0041] Specifically, the visual data transmission module and the binocular vision acquisition module are connected by wireless transmission, and the visual data transmission module and the visual image processing module are connected by wireless or wired means, which can effectively overcome the inconvenience of data transmission operation in the elevator shaft and the problem of data transmission rate, and ensure the real-time operation of the system.

[0042] The visual image processing module is used to analyze and process the collected binocular elevator guide rail image data, thereby completing the automatic measurement of various data of the elevator guide rail;

[0043] Specifically, the implementation of the visual image processing module includes the following steps:

[0044] S21: Set the initial size of the image window to 4×4, perform local window analysis on each pixel in the image, and judge the noise situation by counting the gray value distribution of the pixels in the window. If it is found that the gray value difference between the central pixel and the surrounding pixels is too large, gradually expand the window until a suitable window is found so that the gray value distribution of the pixels in the window meets certain uniformity conditions, and always give priority to retaining the edge features of the guide rail during the analysis and processing;

[0045] S22: Convert the elevator guide rail image data into a grayscale image to reduce the data dimension and calculation amount of subsequent processing. For any pixel point (P(x,y)) in the image, accurately assign different weights to its red, green, and blue components, and calculate its grayscale value (Gray(x,y)) by the formula (Gray(x,y)=0.6×R(x,y)+0.25×G(x,y)+0.15×B(x,y)), where (R(x,y)), (G(x,y)), and (B(x,y)) are the red, green, and blue component values ​​of the pixel point, respectively.

[0046] Specifically, based on the fact that human eyes are most sensitive to green, followed by red, and relatively less sensitive to blue, different weights are given to the red, green, and blue components, effectively highlighting the green component to which the human eye is sensitive. The converted grayscale image performs better in retaining the surface texture information of the elevator guide rail, and compared with traditional average weighted grayscale, it successfully reduces unnecessary color information redundancy.

[0047] S23: Use the SURF algorithm to quickly screen feature points. Based on the extreme point detection principle of the Hessian matrix determinant, it can quickly locate possible feature points in the image. Its detection speed is greatly improved compared with traditional methods, and it can cover a large area of ​​the image in a short time.

[0048] S24: For the feature points initially screened, the SIFT algorithm is used for precise positioning. By constructing a Gaussian difference pyramid, the feature points are detected in a multi-scale space, which can effectively overcome the influence of changes such as image scaling and rotation, and ensure the accurate position of the feature points. At the same time, for the continuous texture area on the surface of the guide rail, the local binary pattern texture descriptor is used for feature description. By comparing the grayscale values ​​of the central pixel and the neighboring pixels, a binary code is generated to characterize the texture features. It has strong robustness to illumination changes, thereby comprehensively capturing various features of the guide rail and providing rich and accurate information for subsequent stereo matching.

[0049] S25: Based on deep learning, various parameters of the elevator guide rails are matched and calculated on the image, so that the automatic measurement operation of the elevator guide rails can be accurately completed.

[0050] The data storage and output module is used to store the measured elevator guide rail parameters for subsequent query and analysis, and to output key data in real time and display it on the operation terminal, so that the inspection personnel can intuitively understand the status of the elevator guide rails;

[0051] Specifically, the data storage and output module outputs key data on the operating terminal, including the position of the elevator guide rail, measured items, measured data, collected images and processed images, and can directly and specifically display the measurement results, thereby improving the convenience and accuracy of staff's query and analysis.

[0052] It is to be understood that the present invention is described by some embodiments, and it is known to those skilled in the art that various changes or equivalent substitutions may be made to these features and embodiments without departing from the spirit and scope of the present invention. In addition, under the teachings of the present invention, these features and embodiments may be modified to adapt to specific circumstances and materials without departing from the spirit and scope of the present invention. Therefore, the present invention is not limited by the specific embodiments disclosed herein, and all embodiments falling within the scope of the claims of this application are within the scope of protection of the present invention.

Claims

1. An automatic measurement system for elevator guide rails based on binocular vision, characterized in that: include: Binocular vision acquisition module, used to realize the vision-based acquisition operation of elevator guide rail data, consists of two high-precision industrial cameras and is symmetrically installed on the measurement bracket; The visual data transmission module is used to transmit the collected elevator guide rail data, realize the connection between the binocular vision acquisition module and the visual image processing module, and ensure the integrity and timeliness of the data; The visual image processing module is used to analyze and process the collected binocular elevator guide rail image data, thereby completing the automatic measurement of various data of the elevator guide rail; The data storage and output module is used to store the measured elevator guide rail parameters for subsequent query and analysis, and to output key data in real time and display it on the operation terminal so that the inspection personnel can intuitively understand the status of the elevator guide rails.

2. The automatic elevator guide rail measurement system based on binocular vision according to claim 1 is characterized in that: When the binocular vision acquisition module acquires data from the elevator guide rail, the binocular vision acquisition module includes the following steps: S11: two high-precision industrial cameras are fixedly mounted on the measuring bracket, the measuring bracket is fixedly mounted on the drone, a laser receiving device is installed on the top surface of the measuring bracket, and a laser emitting device is installed at a designated position on the top surface of the elevator shaft; S12: During the measurement, the drone is started and placed inside the elevator shaft, and the laser emitted by the laser emitting device is aligned with the laser receiving device, and the flight of the drone is precisely controlled based on the signal feedback from the laser receiving device; S13: The drone flies along the elevator rails inside the elevator shaft, allowing two high-precision industrial cameras to complete visual data collection operations on the elevator rails.

3. The automatic measurement system for elevator guide rails based on binocular vision according to claim 2 is characterized in that: The visual data transmission module and the binocular vision acquisition module are connected by wireless transmission, and the visual data transmission module and the visual image processing module are connected by wireless or wired transmission.

4. The automatic elevator guide rail measurement system based on binocular vision according to claim 3 is characterized in that: The implementation of the visual image processing module includes the following steps: S21: setting an image window of an initial size, performing local window analysis on each pixel in the image, judging the noise situation by counting the grayscale value distribution of the pixels in the window, and gradually expanding the window until a suitable window is found so that the grayscale value distribution of the pixels in the window meets certain uniformity conditions if it is found that the grayscale value difference between the central pixel and the surrounding pixels is too large; S22: Convert the elevator guide rail image data into a grayscale image to reduce the data dimension and calculation amount of subsequent processing. For any pixel point (P(x,y)) in the image, accurately assign different weights to its red, green, and blue components, and calculate its grayscale value (Gray(x,y)) by the formula (Gray(x,y)=a×R(x,y)+b×G(x,y)+c×B(x,y)), where (R(x,y)), (G(x,y)), and (B(x,y)) are the red, green, and blue component values ​​of the pixel point, respectively. S23: Use the SURF algorithm to quickly screen feature points. Based on the extreme point detection principle of the Hessian matrix determinant, it can quickly locate possible feature points in the image. Its detection speed is greatly improved compared with traditional methods, and it can cover a large area of ​​the image in a short time. S24: For the feature points initially screened, the SIFT algorithm is used for precise positioning. By constructing a Gaussian difference pyramid, the feature points are detected in a multi-scale space, which can effectively overcome the influence of changes such as image scaling and rotation, and ensure the accurate position of the feature points. At the same time, for the continuous texture area on the surface of the guide rail, the local binary pattern texture descriptor is used for feature description. By comparing the grayscale values ​​of the central pixel and the neighboring pixels, a binary code is generated to characterize the texture features. It has strong robustness to illumination changes, thereby comprehensively capturing various features of the guide rail and providing rich and accurate information for subsequent stereo matching. S25: Based on deep learning, various parameters of the elevator guide rails are matched and calculated on the image, so that the automatic measurement operation of the elevator guide rails can be accurately completed.

5. The automatic elevator guide rail measurement system based on binocular vision according to claim 4 is characterized in that: The initial window size in step S21 is set to 4×4, and during the analysis process, priority is always given to retaining the edge features of the guide rail.

6. The automatic elevator guide rail measurement system based on binocular vision according to claim 5, characterized in that: In step S22, different weights are assigned to the red, green and blue components based on the fact that the human eye is most sensitive to green, followed by red and relatively less sensitive to blue, thereby effectively highlighting the green component to which the human eye is sensitive. The converted grayscale image performs better in retaining the surface texture information of the elevator guide rail, and compared with the traditional average weighted grayscale, it successfully reduces unnecessary color information redundancy.

7. The automatic elevator guide rail measurement system based on binocular vision according to claim 6, characterized in that: The weight assigned to the green component (G(x, y)) is set to (0.6), the weight assigned to the red component (R(x, y)) is set to (0.25), and the weight assigned to the blue component (B(x, y)) is set to (0.15).

8. The automatic elevator guide rail measurement system based on binocular vision according to claim 7, characterized in that: The data storage and output module outputs key data on the operation terminal, including the position of the elevator guide rail, the measured items, the measured data, the collected images and the processed images.

Images