A mobile wind speed measurement method in a construction tunnel combined with video speed measurement

Abstract

The present invention provides a method for measuring wind speed in a construction tunnel using mobile technology combined with video speed measurement to obtain wind speed values at various locations within the tunnel. Hardware placement involves placing a binocular camera on a mobile vehicle, capturing video images of the tunnel wall on the mobile system to calculate the vehicle's motion state. An anemometer is also mounted on the mobile vehicle to obtain the integrated wind speed measured by the mobile system. Time-sequential stereo image pair generation involves performing time synchronization correction on the captured binocular video images to generate time-sequential stereo image pairs. Binocular stereo solution involves obtaining the continuous displacement changes of the mobile measurement vehicle through time-sequential stereo image pair solution, intersection algorithm, and coordinate transformation, thereby calculating the vehicle's moving speed. The wind speed within the tunnel is calculated by measuring the platform's moving speed and correcting the wind speed based on data from the wind speed acquisition device to obtain the continuously changing wind speed within the tunnel.

Description

Technical Field

[0001] The present invention relates to the field of engineering monitoring, and in particular to a mobile wind speed measurement method in a construction tunnel combined with video speed measurement. Background Art

[0002] In the construction industry, especially during tunnel construction, accurate wind speed measurement is crucial for worker safety and the smooth progress of the project. Accurate wind speed measurement can help predict airflow patterns and prevent adverse wind impacts on the construction process and monitoring. Currently, common wind speed measurement methods include real-time sensor-based measurement and the use of meteorological instruments. However, in the mobile measurement environment of construction tunnels, traditional wind speed measurement methods may be limited. For example, the placement and fixation of sensors may be limited by the space within the tunnel, and sensors may be affected by the unique airflow patterns within the tunnel, resulting in inaccurate measurement data.

[0003] Existing ground-based mobile measurement methods use vehicle-mounted GPS systems to achieve spatial positioning and measure vehicle speed. Although GPS can provide accurate position and speed information, the limited signal in the enclosed and sheltered environment of a construction tunnel affects the real-time and accuracy of the measurement results. In addition, vehicle-mounted speed sensors, which measure vehicle speed by sensing wheel rotation, also face severe challenges in terms of sensor accuracy when faced with harsh muddy or uneven road conditions. Furthermore, in the field of vehicle-mounted video speed measurement, the reliance on key points and the real-time nature of large-scale, high-definition video acquisition and processing make system design and performance difficult. Summary of the Invention

[0004] To address the current issues of insufficient vehicle speed and wind speed monitoring accuracy, poor data synchronization, and high equipment maintenance costs in automated tunnel construction, this paper proposes a mobile wind speed measurement method in a construction tunnel combined with video speed measurement. This method leverages the large environmental variations within the construction tunnel to adaptively select key points and track pixel changes at these key points, thereby accurately estimating movement speed. Furthermore, the method uses data from wind speed acquisition equipment to achieve wind speed correction.

[0005] In order to achieve the above technical features, the purpose of the present invention is achieved as follows: a method for measuring wind speed in a construction tunnel combined with video speed measurement, comprising the following steps:

[0006] Step S1: Install a pan-tilt platform at an appropriate position on the surface of the mobile vehicle, and place a video acquisition device at the front end of the pan-tilt platform. The video acquisition device uses two cameras and uses binocular measurement to continuously acquire video image information in the direction of the vehicle's movement;

[0007] Step S2: Using the stereo overlap area between the two cameras, perform time synchronization correction on the video image information, and set a fixed time interval to sample the video to form a frame sequence;

[0008] Step S3: Using filtering technology to remove noise from the image frame, and periodically selecting key points of the same name according to the set feature point extraction rules, and performing enhancement processing on them;

[0009] Step S4: Based on the key points of the same name extracted in S3, track their pixel changes, fit the three-dimensional motion trajectory of the car, and deduce the basis vector movement speed of the car in the set coordinate system;

[0010] Step S5: Install a wind speed acquisition device at an appropriate position on the surface of the mobile vehicle, and obtain the actual wind field conditions in combination with the basis vector movement speed obtained in S4.

[0011] The lens used in the video capture device in step 1 is a fixed-focus lens with a known focal length; the lens of the video capture device should be a lens with small distortion; the layout of the video capture device should ensure that the field of view falls within the lighting space of the lighting device, and the distance between the two cameras should be measured in advance.

[0012] The time synchronization correction described in step 2 is to add timestamp information to the image captured by each camera and adjust the time axis of the video on the captured image accordingly to keep the video synchronized in time; the frame sequence should be batch-processed in a queue manner under the condition that the storage space of the vehicle's on-board memory allows.

[0013] The feature point extraction rule set in step S3 mainly uses the pre-marked positions of significant features in the tunnel to automatically identify the regular uneven areas inside the tunnel or the front vanishing point positions as the key same-name points in the cycle; the enhancement processing is to use contrast enhancement and sharpening algorithms to make the edges and details of the key same-name points more eye-catching and clear.

[0014] The tracking of pixel changes of key homonymous points in step S4 is to track the pixel scaling and pixel position movement of key homonymous points in the frame sequences of the two cameras for a certain time period and store them in the microprocessor.

[0015] The three-dimensional motion trajectory of the car is fitted in step S4 by using the seven-coordinate method in photogrammetry. By analyzing the pixel changes of the key same-name points of the two cameras within a certain time interval, the displacement of the car from the time the key same-name points are selected to the time the solution is completed is calculated.

[0016] The wind speed collection equipment described in step S5 should be able to collect wind speed data in the environment more accurately, including the magnitude and direction of the wind speed; the actual wind field situation is obtained by projecting the calculation results of the wind speed collection equipment onto the basis vector and combining them with the basis vector movement speed calculated in S4 to achieve a comprehensive analysis of the wind field situation.

[0017] The frames are processed in batches in a queue manner, a buffer is opened in the memory, and when the key homonymous point of a cycle is about to lose its effect, the frames of the previous cycle are moved out of the memory, and the frames in the memory buffer are batch imported into the memory for processing.

[0018] The identification of points with significant features uses artificial intelligence to comprehensively score the positions in the stereo overlapping area that may become key homonymous points based on lighting conditions, clarity in the previous and next frames, and the possibility of confusion with other points, and selects the position with the highest score as the key homonymous point for subsequent processing.

[0019] The present invention has the following beneficial effects:

[0020] 1. This invention uses photogrammetry to automatically select and extract key points of the same name within a construction tunnel to determine vehicle speed. Existing methods for extracting vehicle speed from driving videos either rely too heavily on inherent feature points or lack stability in the complex construction tunnel scene. These methods have significant drawbacks in the construction industry and are unsuitable for determining vehicle speed within tunnels.

[0021] 2. This invention provides a novel approach for calculating vehicle speed from video images using three-segment queue storage and processing. First, the image data acquired by the image acquisition device is denoised and time-synchronized using timestamps. Images within a certain period of time are then collected to form Queue 1. Frame sampling is then performed on the images in Queue 1 to form Queue 2. Finally, key points of the same name are extracted to form Queue 3. These three processes are nested within each other, avoiding the large scale and high storage and processor requirements of traditional data processing, while achieving efficient data processing and information transmission.

[0022] 3. The present invention incorporates the moving speed of the trolley into the consideration of wind speed monitoring in the wind farm, and further corrects the wind speed data monitored by the sensor, which is an innovation of the existing mobile wind speed measurement method.

[0023] 4. Based on binocular photogrammetry theory, this invention proposes a mobile video wind speed measurement method for construction tunnels. This method calculates the vehicle's velocity within the tunnel from video data. This method has broad application prospects in underground engineering construction across multiple industries, including civil engineering, water conservancy, and transportation. It is of great significance for monitoring wind speed on construction sites and improving the quality of the construction environment. BRIEF DESCRIPTION OF THE DRAWINGS

[0024] The present invention will be further described below with reference to the accompanying drawings and examples.

[0025] Figure 1 Flow chart of the method of the present invention.

[0026] Figure 2 Schematic diagram of the functions of the three queues of the present invention.

[0027] In the figure: 1 video stream queue; 2 frame sequence queue; 3 tracking of key points of the same name in the previous cycle; 4 tracking of key points of the same name in the next cycle. DETAILED DESCRIPTION

[0028] The embodiments of the present invention will be further described below with reference to the accompanying drawings.

[0029] Example 1:

[0030] See also Figure 1-2 A method for measuring wind speed in a construction tunnel using video velocity measurement includes the following steps:

[0031] Step S1: Install a pan-tilt platform at an appropriate position on the surface of the mobile car, and place a video acquisition device at the front end of the pan-tilt platform. The video acquisition device uses two cameras and adopts binocular measurement to continuously obtain video image information in the direction of the car's movement.

[0032] Furthermore, the lens used in the video capture device is a fixed-focus lens with a known focal length; the lens of the video capture device should be a lens with small distortion; the layout of the video capture device should ensure that the field of view falls within the lighting space of the lighting device, and the distance between the two cameras should be measured in advance.

[0033] Step S2: Using the stereo overlap area between the two cameras, perform time synchronization correction on the video image information, and set a fixed time interval to sample the video to form a frame sequence.

[0034] Furthermore, the time synchronization correction is to add timestamp information to the image captured by each camera and adjust the time axis of the video on the captured image accordingly to keep the video synchronized in time; the frame sequence should be processed in batches in a queue manner under the condition that the storage space of the vehicle's onboard memory allows;

[0035] Furthermore, the frames are processed in batches in a queue manner, a buffer is opened in the memory, and when the key homonymous point of a cycle is about to lose its effect, the frames of the previous cycle are moved out of the memory, and the frames in the memory buffer are batch imported into the memory for processing.

[0036] Step S3: Use filtering technology to remove noise on the image frame, and at the same time, periodically select key points of the same name according to the set feature point extraction rules and perform enhancement processing on them.

[0037] Furthermore, the set feature point extraction rules mainly use the pre-marked positions of significant features in the tunnel to automatically identify the regular uneven areas inside the tunnel or the front vanishing point positions as the key homonymous points within the cycle; the enhancement processing is to use contrast enhancement and sharpening algorithms to make the edges and details of the key homonymous points more eye-catching and clear.

[0038] Furthermore, the points with significant features are identified using artificial intelligence to comprehensively score the positions in the stereo overlapping area that may become key homonymous points based on lighting conditions, clarity in the previous and next frames, and the possibility of confusion with other points, and the positions with the highest scores are selected as key homonymous points for subsequent processing.

[0039] Step S4: Based on the key points of the same name extracted in S3, track their pixel changes, fit the three-dimensional motion trajectory of the car, and deduce the basis vector movement speed of the car in the set coordinate system.

[0040] Furthermore, the tracking of pixel changes of key homonymous points is to track the pixel scaling and pixel position movement of key homonymous points in the frame sequences of the two cameras for a certain time period and store them in the microprocessor; the fitting of the three-dimensional motion trajectory of the car is to adopt the seven-coordinate method in photogrammetry, and calculate the displacement of the car from the time the key homonymous points are selected to the time the calculation is completed by analyzing the pixel changes of the key homonymous points of the two cameras within a certain time interval.

[0041] Step S5: Install a wind speed acquisition device at an appropriate position on the surface of the mobile vehicle, and obtain the actual wind field conditions in combination with the basis vector movement speed obtained in S4.

[0042] Furthermore, the wind speed collection equipment should be able to collect wind speed data in the environment more accurately, including the magnitude and direction of the wind speed; the actual wind field situation is obtained by projecting the calculation results of the wind speed collection equipment onto the basis vector, and combining it with the basis vector movement speed calculated by S4 to achieve a comprehensive analysis of the wind field situation.

[0043] Example 2:

[0044] A mobile wind speed measurement method in a construction tunnel combined with video speed measurement. This method provides a more suitable method and equipment selection for speed monitoring and wind speed measurement in a construction tunnel. Furthermore, a three-stage queue algorithm is designed to accurately and automatically measure the speed of a vehicle moving in a construction tunnel. The mobile video speed measurement and wind speed correction method in a construction tunnel provided by this invention innovates traditional mobile speed monitoring and wind speed measurement methods, improving the efficiency and accuracy of wind speed measurement data during movement.

[0045] The present invention will take a mobile vehicle equipped with a pan-tilt platform as an example and use it as a carrier for mobile wind speed measurement in an example.

[0046] This example implements a method for mobile video velocity measurement and wind speed correction in a construction tunnel. The specific steps include:

[0047] (1) An image acquisition device is set at an appropriate position on the vehicle's pan-tilt platform. In addition to the basic motor drive module and power supply module, the vehicle should also have a pan-tilt platform and a lighting system on its surface. Two cameras with a known distance between them are arranged on the pan-tilt platform, and the lighting system provides sufficient lighting intensity so that the imaging range of the camera falls within the space where the light of the lighting system does not diverge. The image acquisition device should include a storage end and a processing end. Due to the limited performance of the memory on the vehicle, it is necessary to obtain the vehicle's three-dimensional basis vector movement speed within a certain time period and then delete the image information of the previous period.

[0048] (2) Perform time synchronization correction on the video image information and sample it to form a frame sequence. By obtaining the current system time at the moment the camera captures the image, and embedding the time information into the image as a timestamp, the image time axis is adjusted to eliminate the inaccuracy caused by the slight time difference between the devices or sensors; if the two cameras do not have a common system time, the video stream is interpolated to obtain the frame images in two adjacent system times. Then, based on the set time interval, the image frames are extracted from the continuous video stream and arranged in order to form a frame sequence. To save storage resources, the frame sequence is stored in a queue, and the storage period is the effective duration of the key points of the same name within the period.

[0049] (3) Filter to remove noise on the image frame, and select key points of the same name for enhancement processing. By adopting an appropriate filtering algorithm such as an adaptive histogram equalization algorithm combined with a Canny operator, the unstable fluctuations on the image caused by vehicle vibration, circuit fluctuations, dust interference, etc. are reduced, making the key information more prominent while weakening the influence of background noise. The key points of the same name are selected in advance, such as regular uneven areas inside the tunnel or the front vanishing point position, and the artificial intelligence algorithm is collected to identify the possible key points of the same name within the cycle; the possible key points of the same name are comprehensively scored according to the lighting conditions, the clarity of the previous and next frames, and the possibility of confusion with other points, and the position with the highest score is selected as the key point of the same name for subsequent processing. The enhancement processing is to use contrast enhancement and sharpening algorithms to make the edges and details of the key points of the same name more eye-catching and clear;

[0050] (4) Track the pixel changes of key homonymous points, fit the three-dimensional motion trajectory of the car, and deduce the basis vector velocity of the car in the set coordinate system. The coordinates of a key homonymous point in a single frame can be derived by the following formula:

[0051]

[0052] The actual coordinates of the object point derived from the formula are

[0053] Where: x is the x-coordinate of the object in a single frame, y is the y-coordinate of the object in a single frame, z is the z-coordinate of the object in a single frame, w is the scale magnification factor, x l is the x coordinate of the object on the image plane, y l is the y coordinate of the object on the image plane, d is the parallax, c x is the offset in the x direction from the principal point of the left camera to the center of the image plane, c y is the offset in the y direction from the camera principal point to the center of the image plane, f is the focal length of the camera, T x is the center distance between the left and right cameras, c′ x The offset in the x direction from the right camera principal point to the center of the image plane.

[0054] Based on the pixel changes of key points of the same name between frames, the actual coordinate changes of the car are deduced, and the three-dimensional motion trajectory of the car is fitted. Combined with the sampling interval of the frame sequence, the three-dimensional speed of the car is calculated and projected onto the set coordinate axis to obtain the basis vector speed.

[0055] (5) Install a wind speed acquisition device at an appropriate location on the surface of the mobile vehicle and combine it with the base vector velocity obtained in S4 to obtain the actual wind field conditions. Based on the collected wind speed magnitude and direction, project it onto the set coordinate axis and combine it with the base vector velocity to obtain more comprehensive wind field information that takes into account the influence of velocity.

Claims

1. A method for measuring wind speed in a construction tunnel using video velocity measurement, characterized in that: The following steps are involved: Step S1: Install a pan-tilt platform at an appropriate position on the surface of the mobile vehicle, and place a video acquisition device at the front end of the pan-tilt platform. The video acquisition device uses two cameras and uses binocular measurement to continuously acquire video image information in the direction of the vehicle's movement; Step S2: Using the stereo overlap area between the two cameras, perform time synchronization correction on the video image information, and set a fixed time interval to sample the video to form a frame sequence; Step S3: Using filtering technology to remove noise from the image frame, and periodically selecting key points of the same name according to the set feature point extraction rules, and performing enhancement processing on them; Step S4: Based on the key points of the same name extracted in S3, track their pixel changes, fit the three-dimensional motion trajectory of the car, and deduce the basis vector movement speed of the car in the set coordinate system; Step S5: Install a wind speed acquisition device at an appropriate position on the surface of the mobile vehicle, and obtain the actual wind field conditions by combining the basis vector movement speed obtained in S4; The feature point extraction rule set in step S3 is to use the pre-marked positions of significant features in the tunnel to automatically identify the regular uneven areas inside the tunnel or the front vanishing point positions as the key homonymous points within the cycle; perform enhancement processing by using contrast enhancement and sharpening algorithms to make the edges and details of the key homonymous points more eye-catching and clear; identify points with significant features, and use artificial intelligence to comprehensively score the positions in the three-dimensional overlapping area that may become key homonymous points based on lighting conditions, clarity in the previous and next frames, and the possibility of confusion with other points, and select the position with the highest score as the key homonymous point for subsequent processing.

2. The method for measuring wind speed in a construction tunnel using video velocity measurement according to claim 1, wherein: The lens used by the video capture device in step S1 is a fixed-focus lens with a known focal length; the lens of the video capture device should be a lens with small distortion; the layout of the video capture device should ensure that the field of view falls within the lighting space of the lighting device, and the distance between the two cameras should be measured in advance.

3. The method for measuring wind speed in a construction tunnel using video velocity measurement according to claim 1, characterized in that: The time synchronization correction in step S2 is to add timestamp information to the image captured by each camera and adjust the time axis of the video on the captured image accordingly to keep the video synchronized in time; the frame sequence should be batch-processed in a queue manner under the condition that the storage space of the vehicle's on-board memory allows.

4. The method for measuring wind speed in a construction tunnel using video velocity measurement according to claim 1, wherein: The tracking of pixel changes of key homonymous points in step S4 is to track the pixel scaling and pixel position movement of key homonymous points in the frame sequences of the two cameras for a certain time period and store them in the microprocessor.

5. The method for measuring wind speed in a construction tunnel using video velocity measurement according to claim 4, characterized in that: The three-dimensional motion trajectory of the car is fitted in step S4 by using the seven-coordinate method in photogrammetry. By analyzing the pixel changes of the key same-name points of the two cameras within a certain time interval, the displacement of the car from the time the key same-name points are selected to the time the solution is completed is calculated.

6. The method for measuring wind speed in a construction tunnel using video velocity measurement according to claim 1, characterized in that: The wind speed collection equipment described in step S5 should be able to collect wind speed data in the environment more accurately, including the magnitude and direction of the wind speed; the actual wind field situation is obtained by projecting the calculation results of the wind speed collection equipment onto the basis vector and combining them with the basis vector movement speed calculated in S4 to achieve a comprehensive analysis of the wind field situation.

7. The method for measuring wind speed in a construction tunnel using video velocity measurement according to claim 3, characterized in that: The frames are processed in batches in a queue manner, a buffer is opened in the memory, and when the key homonymous point of a cycle is about to lose its effect, the frames of the previous cycle are moved out of the memory, and the frames in the memory buffer are batch imported into the memory for processing.

Images