A static load heap safety monitoring system and monitoring method based on machine vision

Abstract

The present invention discloses a machine vision-based static load safety monitoring system and method. By dynamically capturing images of the load, a camera equipped with RTK positioning uses a visual algorithm to analyze and calculate the load's settlement change and tilt angle during the test. A backend service module deploys measurement algorithms and data management functions. The backend service module is connected to an image acquisition module via the Internet, and a mobile terminal is bidirectionally connected to the backend service module via the Internet. This invention has the advantages of improving the intelligence and automation level of monitoring, enhancing the convenience and safety of monitoring, improving the convenience of testing, improving the real-time nature of monitoring, and reducing costs.

Description

Technical Field

[0001] The present invention relates to the technical field of static load heaping monitoring, and in particular to a static load heaping safety monitoring system and monitoring method based on machine vision. Background Art

[0002] Currently, static load tests are an essential part of ensuring construction safety in construction projects. During these tests, eccentric loading can occur, leading to system instability and posing a significant safety hazard.

[0003] Therefore, the present invention uses machine vision to achieve safety monitoring of static load and heap load tests and real-time monitoring of the test environment. Current detection methods primarily rely on geodetic leveling, including precision leveling and precision trigonometric height measurement. This method is a widely used settlement detection method both domestically and internationally, but it requires specialized equipment and is expensive to perform by professional technicians, and cannot achieve real-time monitoring.

[0004] To this end, we have developed a static load safety monitoring system and monitoring method based on machine vision, which has practical engineering value. Summary of the Invention

[0005] The purpose of the present invention is to overcome the shortcomings of the existing technology and provide a static load heap loading safety monitoring system and monitoring method based on machine vision, which has the advantages of improving the intelligence and automation level of monitoring, improving the convenience and safety of monitoring, improving the convenience of testing, improving the real-time monitoring, and reducing costs.

[0006] To achieve the above-mentioned purpose, the technical solution adopted by the present invention is: a static load stacking safety monitoring system based on machine vision, including an image acquisition module, a mobile terminal, and a background service module. The image acquisition module includes a camera, a fixed reference object, a stacking platform and a stacking. A light source is provided at each of the four corners of the top of the stacking platform. The stacking is set at the top of the stacking platform. The camera is set on a diagonal of the stacking platform. The fixed reference object is placed vertically on one side of the stacking platform and is set on the perpendicular bisector of the other diagonal of the stacking platform. The camera analyzes and calculates the settlement change and inclination angle of the stacking in the test through visual algorithms. The background service module deploys measurement algorithms and data management functions, and is connected to the image acquisition module through the Internet. The mobile terminal is bidirectionally connected to the background service module through the Internet.

[0007] Preferably, for intelligent monitoring based on machine vision, the cameras are two monocular cameras or one binocular camera with RTK positioning function.

[0008] Preferably, a monitoring method of a static load safety monitoring system based on machine vision comprises the following steps:

[0009] 3.1. Start the camera to collect the image of the loading platform in the static load test in real time. If the environment is dark, turn on the light source on the loading platform to improve real-time performance and reduce costs.

[0010] 3.2. Using RTK positioning technology, the actual distance between the two cameras is measured, and the data is collated and uploaded to the backend service module to be stored as the physical characteristics of the loading platform for the static load test;

[0011] 3.3. The camera uploads the captured image to the backend service module, and analyzes the real-time image through the algorithm encapsulated in the backend service module;

[0012] 3.4. Determine the safety level of the loading platform according to the preset range of settlement and tilt angle of the loading;

[0013] 3.5. Using cloud service technology, monitoring data and safety levels are sent to the mobile terminal, allowing test personnel to promptly view the safety information of the loading platform. If the monitoring value exceeds the safety threshold, the monitoring device will alarm, improving the convenience and safety of monitoring.

[0014] Preferably, the step 3.3 includes the following steps:

[0015] 3.3.1. The fixed reference object and the ground in the image are considered as a fixed measurement coordinate system;

[0016] 3.3.2. Reduce the interference of useless information through image filtering and blurring;

[0017] 3.3.3. Based on the actual distance between the two cameras, the pixel distance between the fixed reference object in the image and the two cameras, and the proportional relationship between the pixel size of similar triangles and the actual distance, the overall physical dimensions of the load are obtained, and the total mass of the load is estimated based on the characteristics of the heavy object;

[0018] 3.3.4. Using frequency domain analysis, extract the bottom edge lines between the stacked objects and the stacked platform;

[0019] 3.3.5. Calculate the line-related data information in the initial image acquisition in the measurement coordinate system, save it to the background service module, and use it as the initial reference position information for monitoring;

[0020] 3.3.6. Compare and analyze the straight line information measured in the images collected at different times with the initial reference position information to obtain the change in distance from the reference position and the angle formed by the change in the straight line, which is the settlement and tilt angle of the loading platform in the static load test.

[0021] Preferably, in step 3.3.3, or when using a binocular camera, the disparity map of the test load can be solved according to the limit constraint principle between camera images, and the three-dimensional spatial size of the load can be obtained by reconstructing the outer edge contour of the load, thereby estimating the total mass of the load.

[0022] Preferably, in step 3.1, the light source is a point light source or a line light source, and is modulated by a low-frequency digital pulse of less than 1 Hz.

[0023] Preferably, in step 3.2, the background service module is a background server, and the background server receives and stores the data.

[0024] Preferably, in step 3.3.2, the image filtering uses the frequency domain to perform denoising and detect edges.

[0025] Due to the application of the above technical solution, the present invention has the following advantages compared with the prior art:

[0026] 1. The machine vision-based static load safety monitoring system and method of the present invention utilize machine vision to achieve intelligent monitoring of static load test safety, thereby improving the level of automation and intelligence.

[0027] 2. Utilize cloud service technology to efficiently combine backend calculations with mobile data viewing to improve the convenience and security of monitoring;

[0028] 3. Based on machine vision and RTK positioning technology, the total mass of the pile is estimated, providing many conveniences for testing;

[0029] 4. The monitoring system is simple to install, easy to use, improves real-time performance and reduces costs. BRIEF DESCRIPTION OF THE DRAWINGS

[0030] Attachment Figure 1 This is a structural diagram of the static load safety monitoring system based on machine vision according to the present invention. DETAILED DESCRIPTION

[0031] The present invention will be further described in detail below with reference to the accompanying drawings and specific embodiments.

[0032] Attachment Figure 1A machine vision-based static load safety monitoring system includes an image acquisition module, a mobile terminal 20, and a backend service module 30. The image acquisition module includes a camera 10, a fixed reference object 40, a loading platform 55, and a load 60. A light source 50 is provided at each of the four corners of the loading platform 55. The load 60 is located at the top of the loading platform 55. The camera 10 is located on a diagonal of the loading platform 55. The fixed reference object 40 is placed vertically on one side of the loading platform 55 and on the perpendicular bisector of the other diagonal of the loading platform 55. The camera 10 is a two-lens camera or a binocular camera with RTK positioning function.

[0033] A monitoring method for a static load safety monitoring system based on machine vision comprises the following steps:

[0034] 3.1. Start the camera 10 to collect images of the loading platform 55 of the static load test in real time. If the environment is dark, turn on the light source 50 on the loading platform 55. The light source 50 is a point light source or a line light source, and is modulated by a low-frequency digital pulse of less than 1 Hz.

[0035] 3.2. Utilize RTK positioning technology to measure the actual distance between the two cameras 10, and organize and upload the data to the background service module 30 to be stored as the physical characteristics of the loading platform 55 for the static load test; the background service module 30 is a background server that receives and stores the data.

[0036] 3.3. The camera 10 uploads the captured image to the background service module 30 and analyzes the real-time image through the algorithm encapsulated in the background service module 30; the process includes the following steps:

[0037] 3.3.1. The fixed reference object 40 in the image and the ground are considered as a fixed measurement coordinate system.

[0038] 3.3.2. Reduce the interference of useless information through image filtering and blurring. Image filtering uses the frequency domain to denoise and detect edges.

[0039] 3.3.3. Based on the actual distance between the two cameras 10 and the pixel distance between the fixed reference object 40 in the image and the two cameras 10, the overall physical dimensions of the load 60 are obtained using the proportional relationship between the pixel size of similar triangles and the actual distance, and the total mass of the load 60 is estimated based on the characteristics of the heavy object; or when using binocular cameras, the disparity map of the test load 60 can be solved based on the limit constraint principle between the camera images, and the three-dimensional spatial dimensions of the load 60 are obtained by reconstructing the outer edge contour of the load 60, thereby estimating the total mass of the load 60.

[0040] 3.3.4. Use frequency domain analysis to extract the bottom edges between the stacked objects and the stacking platform 55.

[0041] 3.3.5. In the measurement coordinate system, calculate the straight line related data information in the initially collected image, save it to the background service module 30, and use it as the initial reference position information for monitoring.

[0042] 3.3.6. Compare and analyze the straight line information measured in the images collected at different times with the initial reference position information to obtain the change in distance from the reference position and the angle formed by the change in the straight line, which is the settlement and tilt angle of the loading platform 55 in the static load test.

[0043] 3.4. Determine the safety level of the loading platform 55 based on the preset range of settlement and tilt angle of the loading 60.

[0044] 3.5. Using cloud service technology, the monitoring data and safety level are sent to the mobile terminal 20, so that the test personnel can check the safety information of the loading platform 55 in time. If the monitoring value exceeds the safety threshold, the monitoring device will alarm.

[0045] The camera 10 uses visual algorithm analysis to calculate the settlement change and tilt angle of the load 60 in the test. The background service module 30 deploys measurement algorithms and data management functions, and is connected to the image acquisition module through the Internet. The mobile terminal 20 is bidirectionally connected to the background service module 30 through the Internet.

[0046] The above are only specific application examples of the present invention and do not constitute any limitation on the scope of protection of the present invention. Any technical solutions formed by equivalent transformation or equivalent replacement shall fall within the scope of protection of the present invention.

Claims

1. A monitoring method for a static load safety monitoring system based on machine vision, characterized by: The static load heap safety monitoring system based on machine vision includes an image acquisition module, a mobile terminal, and a background service module. The image acquisition module includes a camera, a fixed reference object, a loading platform, and a load. A light source is provided at each of the four corners of the top of the loading platform. The load is provided at the top of the loading platform. The camera is provided on a diagonal of the loading platform. The fixed reference object is placed vertically on one side of the loading platform and is provided on the perpendicular bisector of the other diagonal of the loading platform. The camera calculates the settlement change and tilt angle of the load in the test through visual algorithm analysis. The background service module deploys measurement algorithms and data management functions and is connected to the image acquisition module via the Internet. The mobile terminal is bidirectionally connected to the background service module via the Internet. The cameras are two monocular cameras or one binocular camera with RTK positioning function; A monitoring method for a static load safety monitoring system based on machine vision comprises the following steps: 3.

1. Start the camera to capture the image of the loading platform during the static load test in real time. If the environment is dark, turn on the light source on the loading platform. 3.

2. Using RTK positioning technology, the actual distance between the two cameras is measured, and the data is collated and uploaded to the backend service module to be stored as the physical characteristics of the loading platform for the static load test; 3.

3. The camera uploads the captured image to the backend service module, and analyzes the real-time image through the algorithm encapsulated in the backend service module; 3.3.

1. The fixed reference object and the ground in the image are considered as a fixed measurement coordinate system; 3.3.

2. Reduce the interference of useless information through image filtering and blurring; 3.3.

3. Based on the actual distance between the two cameras, the pixel distance between the fixed reference object in the image and the two cameras, and the proportional relationship between the pixel size of similar triangles and the actual distance, the overall physical dimensions of the load are obtained, and the total mass of the load is estimated based on the characteristics of the heavy object; 3.3.

4. Using frequency domain analysis, extract the bottom edge lines between the stacked objects and the stacked platform; 3.3.

5. Calculate the line-related data information in the initial image acquisition in the measurement coordinate system, save it to the background service module, and use it as the initial reference position information for monitoring; 3.3.

6. Compare and analyze the straight line information measured in the images collected at different times with the initial reference position information to obtain the change in distance from the reference position and the angle formed by the change in the straight line, which is the settlement and tilt angle of the loading platform in the static load test; 3.

4. Determine the safety level of the loading platform according to the preset range of settlement and tilt angle of the loading; 3.

5. Using cloud service technology, the monitoring data and safety level are sent to the mobile terminal, so that the test personnel can check the safety information of the loading platform in time. If the monitoring value exceeds the safety threshold, the monitoring device will alarm.

2. The monitoring method of the static load safety monitoring system based on machine vision according to claim 1 is characterized in that: In step 3.3.3, or when using a binocular camera, the disparity map of the test load can be solved based on the limit constraint principle between camera images, and the three-dimensional spatial size of the load can be obtained by reconstructing the outer edge contour of the load, thereby estimating the total mass of the load.

3. The monitoring method of the static load safety monitoring system based on machine vision according to claim 1 is characterized in that: In step 3.1, the light source is a point light source or a line light source, and is modulated by a low-frequency digital pulse of less than 1 Hz.

4. The monitoring method of the static load safety monitoring system based on machine vision according to claim 1 is characterized in that: In step 3.2, the background service module is a background server, and the background server receives and stores data.

5. The monitoring method of the static load safety monitoring system based on machine vision according to claim 1 is characterized in that: In step 3.3.2, the image filtering uses the frequency domain to perform denoising and detect edges.

Images