A tunnel, tunnel 360-degree panoramic image acquisition control system and method

Abstract

The application provides a tunnel, a 360-degree panoramic image acquisition control system and method, relates to the technical field of image acquisition control and water conservancy engineering detection, and comprises an image acquisition device, a focal length control module, a light supplement control module and an image intelligent feedback control module; the application adopts 12 groups of variable-focus lens arrays with equal-angle distribution, the focal length is adjusted in real time by laser radar detection data at the tail end of the acquisition module, and the application can be suitable for tunnel acquisition of different sections; the application carries out quality evaluation and integrity detection on the collected photo images according to motion blurring degree, defocus blurring degree and illumination balance index; the detection result is used to generate a positioning point cloud map through the control system; different strategies are used for reacquisition after the automatic triggering of point analysis of abnormal reasons during the return journey; the application solves the problems of great difficulty in image full-section acquisition, single application range of the acquisition module and data integrity in the complex tunnel environment, and significantly improves the tunnel panoramic image acquisition efficiency and precision.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of image acquisition control and water conservancy detection, in particular to a tunnel and tunnel 360-degree panoramic image acquisition control system and method. BACKGROUND

[0002] Flood control tunnels and tunnels are important flood control facilities, usually located in complex environments such as dam bodies and underground, used to quickly guide water flow during flood periods and assist in drainage during operation. Therefore, the structural stability and integrity of the tunnel and tunnel are crucial to the safe operation of the tunnel and tunnel system.

[0003] Traditional tunnel and tunnel detection methods: Traditional tunnel and tunnel detection often relies on detection personnel entering the tunnel and tunnel, holding a camera to take pictures of the disease location and record it in the book, which is not only time-consuming and labor-intensive, but also prone to inaccurate positioning, difficult to quantify, and easy to miss. Secondly, the detection data recorded by manual recording is often saved in paper or electronic document form, and the data arrangement, storage and query are relatively cumbersome, which is not conducive to long-term data management and analysis. Moreover, due to the complex internal environment of the tunnel and tunnel, insufficient light, water mist, and reflection often make it difficult for ordinary cameras or acquisition equipment to obtain high-quality images, and image blur, color distortion and other problems will also affect subsequent image analysis and disease identification.

[0004] Traditional mobile detection difficulties: Traditional detection devices such as radar and ultrasonic waves, although their detection accuracy has been continuously improved with the development of technology, are often bulky and difficult to transport, and the complex tunnel and tunnel environment makes it difficult for traditional devices to provide basic working conditions. In the face of long tunnels and tunnels, detection personnel have difficulty carrying traditional detection devices for long periods of work.

[0005] Existing mobile detection difficulties: With the continuous development of technology, many rapid mobile detection equipment has been developed to solve the problem of traditional tunnel rapid acquisition. However, the complex system and fine technology inevitably increase the size of the equipment such as detection vehicles, making it difficult to apply to small-diameter tunnels for detection. For example, mobile and track robots, due to the high requirements of flood control tunnels for safety voltage (generally with an internal voltage of less than 40v), most mobile detection equipment cannot be applied.

[0006] Therefore, there is an urgent need for a tunnel and tunnel 360-degree panoramic image acquisition control system and method to solve the problem of low efficiency of existing flood control tunnel detection technology, heavy and difficult to deploy traditional equipment, and new equipment limited by safety voltage and small hole diameter, which makes it difficult to effectively realize high-precision disease detection in complex environments. SUMMARY

[0007] In view of the above technical problems in the related art, the present application provides a tunnel and tunnel 360-degree panoramic image acquisition control system and method.

[0008] In a first aspect, the present application provides a tunnel and tunnel 360-degree panoramic image acquisition control system, comprising: an image acquisition device, a focal length control module, a light supplement control module, and an image intelligent feedback control module.

[0009] The image acquisition device is configured to receive a shooting control signal to acquire a full-section photo image and a section contour point cloud of a tunnel or a tunnel; the image acquisition device adjusts the shooting focal length of a camera, adjusts the light source signal intensity of a light supplement lamp, and determines whether to shoot according to the shooting control signal; the shooting control signal comprises a camera focal length adjustment signal, a light supplement lamp light source intensity signal, and a camera shooting signal;

[0010] The focal length control module is connected to the image acquisition device, and is configured to calculate the distance between each lens of the image acquisition device and the wall according to the section contour point cloud data of the tunnel or the tunnel, adjust the focal length of each camera in the image acquisition device according to the distance between each lens of the image acquisition device and the wall, and send a camera focal length adjustment signal to the image acquisition device; the camera focal length adjustment signal comprises a camera number and a corresponding focal length;

[0011] The light supplement control module is connected to the image acquisition device, and is configured to send a light supplement lamp light source intensity signal to control the light supplement lamp light source intensity when the image acquisition device is working; the light supplement lamp light source intensity signal comprises a camera number and a corresponding light supplement lamp light source intensity;

[0012] The image intelligent feedback control module is connected to the image acquisition device, the focal length control module, and the light supplement control module, and is configured to evaluate the photo image acquired by the image acquisition device each time according to the motion blur degree, the defocus blur degree, and the illumination balance index to obtain a quality comprehensive score, and add the center coordinates of the image acquisition device to an abnormal coordinate list when the quality comprehensive score is lower than a preset quality score threshold; when returning, a re-acquisition strategy is executed at each coordinate point in the abnormal coordinate list until the tunnel or the tunnel is exited;

[0013] The quality comprehensive score is calculated based on the following formula:

[0014] ;

[0015] wherein, is the quality comprehensive score; is the motion blur degree; is the defocus blur degree; is the illumination balance index; is the motion blur degree weight; is an out-of-focus blur weight; is an illumination balance index weight.

[0016] Specifically, the image acquisition device comprises a ring-shaped acquisition module 1 and a laser radar 6; the laser radar 6 is used to acquire the cross-sectional profile point cloud of the tunnel or the tunnel; the ring-shaped acquisition module 1 is arranged in a ring direction with equal angles by 12 acquisition units 11; the acquisition unit 11 comprises a camera 111, a lens 112, a heat dissipation device 113 and a light supplementing lamp 114.

[0017] The image acquisition device acquires the photos of the tunnel or the tunnel in batches.

[0018] The image acquisition device acquires the photos of the tunnel or the tunnel in batches.

[0019] Firstly, the first camera, the fourth camera, the seventh camera and the tenth camera and the corresponding light supplementing device are controlled to perform the first batch acquisition, after the first batch acquisition is completed, the second camera, the fifth camera, the eighth camera and the eleventh camera and the corresponding light supplementing device are controlled to perform the second batch acquisition, after the second batch acquisition is completed, the third camera, the sixth camera, the ninth camera and the twelfth camera and the corresponding light supplementing device are controlled to perform the third batch acquisition, and the total light supplementing range of the light supplementing device of each batch can cover the whole cross section of the tunnel or the tunnel.

[0020] Specifically, the calculation of the distance between the image acquisition device and the wall according to the cross-sectional profile point cloud data of the tunnel or the tunnel comprises the following steps:

[0021] S21, the laser radar 6 is used to acquire the cross-sectional profile point cloud at the front end of the ring-shaped acquisition module 1 at t seconds. , ; wherein, is the radius of the polar coordinate at t; is the angle of the polar coordinate at t, j is the serial number of the point cloud, and N is the total number of the cross-sectional profile point cloud;

[0022] S22, the coordinates of the cross-sectional profile point cloud are converted into Cartesian coordinates to obtain the cross-sectional profile Cartesian point cloud = , and the conversion process is shown in the following formula:

[0023] ;

[0024] wherein, respectively, are the angles of the laser radar calculation relative to the current cross section that the ring-shaped acquisition module 1 is offset on the X and Y axes at t; wherein, is the X coordinate of the cross-sectional Cartesian point cloud at t, is the Y coordinate of the Cartesian point cloud at t;​​

[0025] S23, extracting the cross-section point cloud slice of the front end of the annular collection module 1 within t seconds :

[0026] ;

[0027] wherein, is the slice thickness; represents the position of the collection unit 11 at t seconds; is the point cloud coordinate set of the cross-section point cloud slice; is the preset scanning time for one round; v is the moving speed of the collection unit 11; is the partial derivative function; represents the partial derivative of ;

[0028] S24, calculating the distance between the lens 112 and the hole wall according to the following formula :

[0029] ;

[0030] wherein, , , represents the point cloud coordinate of the maximum hole wall at t seconds from the i-th lens; represents the position coordinate of the i-th lens at t seconds; i={1, 2,..., 12};

[0031] Specifically, the focal length of each camera in the image collection device is adjusted according to the distance between each lens of the image collection device and the hole wall by the following formula :

[0032] ;

[0033] wherein, represents the focal length of the i-th camera at t seconds, , is the standard focal length of the lens.

[0034] The motion blur degree is calculated based on the following formula:

[0035] ;

[0036] wherein, is the effective edge gradient mean of the collected photo image, is the reference clear image gradient threshold value;

[0037] The defocus blur degree ​The following formula was used to calculate:

[0038] ;

[0039] in, represents the Laplacian operator of image I at pixel (x,y), represents the second derivative of the image at that point, and is the default expression; M represents the height of image I; N represents the width of image I. This represents the variance of the image after Laplacian filtering;

[0040] The light balance index The following formula was used to calculate:

[0041] ;

[0042] Where e is the base of the natural logarithm; This represents the average brightness of the current image. The average brightness of the reference group images; This is the attenuation coefficient.

[0043] Specifically, the re-acquisition strategy includes: re-acquiring at the current coordinate point, then performing a comprehensive quality score on the re-acquisitioned photo image; if the recalculated comprehensive quality score is lower than a preset quality score threshold, a graded adjustment strategy is executed; otherwise, the camera continues to return to the next coordinate point in the abnormal coordinate list; the graded adjustment strategy includes: first, executing a motion blur adjustment strategy, then executing an abnormal lighting adjustment strategy, and finally executing a defocus blur adjustment strategy.

[0044] The re-acquisition refers to the acquisition of photos of the tunnel and tunnel shaft in batches using the image acquisition device.

[0045] Specifically, the graded adjustment also includes: performing a comprehensive quality score on the last re-acquired photo images, determining that the comprehensive quality score of all photo images is greater than or equal to the preset quality score threshold, and then conducting a review.

[0046] Specifically, the motion blur adjustment strategy is as follows: if the motion blur corresponding to the coordinate points in the abnormal location list is... If motion blur occurs, a new image is captured at that coordinate point, and then the motion blur determination is performed again on the image captured by the ring acquisition module 1. The calculation continues until the motion blur of the newly acquired photographic image is reached. Or stop re-collecting when the preset maximum number of retakes is reached; among which, This is the preset motion blur threshold.

[0047] Specifically, the lighting anomaly adjustment strategy is as follows: if the lighting balance index If yes, it is considered that the light is abnormal, and the light brightness is adjusted at the coordinate point when returning, and the image is recollected until the image light is normal or the adjustment times are greater than the preset adjustment times; wherein the light brightness adjustment is according to the following formula:

[0048] ;

[0049] Wherein, is the reference group light balance index, which changes with different application scenarios, and is valued according to the initial reference image; is the light brightness of the light when the light is abnormal, and the first time is is equal to the initial light brightness I, and the initial light brightness I is a fixed value, which is generally unchanged; is a linear adjustment factor, which is when the image is overexposed, and when the image is underexposed;

[0050] If , it is considered that the image is underexposed, and the linear adjustment factor is adjusted as follows:

[0051] ;

[0052] If , it is considered that the image is overexposed, and the linear adjustment factor is adjusted as follows:

[0053] ;

[0054] If , it is considered that the image light is normal, and no adjustment is made.

[0055] Wherein, is the floating difference value; is the current image brightness average; is the reference group image brightness average; is the underexposure adjustment coefficient, ; is the overexposure adjustment coefficient, .

[0056] Specifically, the out-of-focus blur adjustment strategy specifically includes: if the out-of-focus blur degree , it is considered that the out-of-focus blur occurs, and the focal length is adjusted by the focal length adjustment amount at the coordinate point when returning, and then the image is recollected, and then the quality comprehensive score of the recollected photo image is calculated, until the score quality requirement is met, and the recollection is stopped; wherein, is the reference group photo blur degree;

[0057] the focal length adjustment amount The calculation formula is as follows:

[0058]

[0059] wherein, is a distance-focal length conversion coefficient; is the distance between the i-th lens and the tunnel wall; is the optimal distance for the current camera focal length imaging; is the focal length adjustment compensation term sgn is a sign function; is the image comprehensive quality fluctuation error value; is an indicator function, which is 1 if the condition is true, and 0 otherwise; is the quality comprehensive score of the image at the current time t, is the quality comprehensive score of the image at the previous time t-1.

[0060] In a second aspect, the present application provides a tunnel and tunnel 360-degree panoramic image acquisition control method based on the tunnel and tunnel 360-degree panoramic image acquisition control system in the first aspect, comprising the following steps:

[0061] S1, using the image acquisition device to advance along the tunnel axis, and using a preset light compensation strategy to acquire the full cross-section photo image and cross-section contour point cloud of the tunnel and tunnel; the preset light compensation strategy is that the intensity of the coaxial light source of the camera of the image acquisition device is less than the intensity of the adjacent light source of the camera wherein: ; k is a light compensation adjustment coefficient;

[0062] S2, calculating the distance between each lens of the image acquisition device and the tunnel wall according to the cross-section contour point cloud data of the tunnel and tunnel;

[0063] S3, dynamically adjusting the focal length of each camera in the image acquisition device according to the distance between each lens of the image acquisition device and the tunnel wall by the following formula:

[0064] S4, evaluating the photo image acquired each time by the image acquisition device according to the motion blur degree, defocus blur degree and illumination uniformity index to obtain a quality comprehensive score, and adding the center coordinates of the image acquisition device to the abnormal coordinate list when the quality comprehensive score is lower than the preset quality score threshold; when returning, a re-acquisition strategy is performed at each coordinate point in the abnormal coordinate list until leaving the tunnel and tunnel.

[0065] ​The application provides a tunnel and tunnel 360-degree panoramic image acquisition control system, which adopts 12 groups of annular variable-focus lens arrays with equal-angle distribution, the focal length of which is adjusted in real time by laser radar detection data at the tail end of the acquisition module, and can be applicable to tunnel acquisition of different sections, the application performs quality evaluation and integrity detection on the acquired photo images according to motion blurring degree, defocus blurring degree and illumination balance index, the detection result is used to generate a positioning point cloud map by the control system, and different strategies are used for reacquisition after triggering point analysis of abnormal reasons automatically during the return journey, so that the problems of great difficulty in full-section image acquisition in a complex tunnel environment, single application range of the acquisition module and data integrity are solved, and the tunnel panoramic image acquisition efficiency and precision are significantly improved.

[0066] Further, the anti-dazzle structure with the uniform light film on the surface of the light supplementing lamp can effectively eliminate water stain reflection interference during the acquisition. BRIEF DESCRIPTION OF DRAWINGS

[0067] In order to more clearly illustrate the technical solutions in the embodiments of the present application or the prior art, the drawings needed to be used in the embodiments will be briefly introduced below. Obviously, the drawings in the following description only constitute some embodiments of the present application, and other drawings can also be obtained by those skilled in the art without creative labor on the basis of these drawings.

[0068] Figure 1 A tunnel and tunnel 360-degree panoramic image acquisition control system schematic diagram provided for the embodiments of the present application;

[0069] Figure 2 A structure schematic diagram of an image acquisition device provided for the embodiments of the present application;

[0070] Figure 3 A structure schematic diagram of an annular acquisition module provided for the embodiments of the present application;

[0071] Figure 4 A structure schematic diagram of an acquisition unit provided for the embodiments of the present application;

[0072] Figure 5 A heat dissipation port schematic diagram of a heat dissipation device provided for the embodiments of the present application;

[0073] Figure 6 A structure schematic diagram of a light supplementing lamp provided for the embodiments of the present application;

[0074] Figure 7 A structure schematic diagram of a protection device provided for the embodiments of the present application;

[0075] Figure 8 A structure schematic diagram of a cover plate provided for the embodiments of the present application;

[0076] Figure 9 A schematic diagram of a tunnel, tunnel 360-degree panoramic image acquisition control method provided by the embodiment of the present application is shown in the figure;

[0077] Wherein, 1-ring collection module; 2-focal length control module; 3-light supplement control module; 4-protection device; 5-image intelligent feedback control module; 6-laser radar; 11-acquisition unit; 111-camera; 112-lens; 113-heat dissipation device; 114-light supplement lamp; 1121-cover plate; 1122-radial protection support; 1123-circumferential protection support; 1121-1lens hole; 1121-2wire hole; 1121-3reinforcing rib; 1121-4waterproof groove; 1131-radiator; 1132-fan; 1133-heat dissipation air outlet; 1141-light source lens; 1142-COB light source; 1143-homogenizing film. DETAILED DESCRIPTION

[0078] The present application can be explained in detail by the following examples, and the purpose of the present application is to protect all technical improvements within the scope of the present application. In the description of the present application, it should be understood that the terms "first" and "second" are used for descriptive purposes only, and cannot be understood as indicating or implying relative importance or implicitly indicating the number of the indicated technical features. Therefore, the features with "first" and "second" can explicitly or implicitly include one or more of the features. In the description of the present application, the meaning of "multiple" is two or more, unless otherwise specifically limited.

[0079] In order to make the technical problems, technical solutions and advantages of the present application clearer, the following will be described in detail with reference to the accompanying drawings and specific examples.

[0080] Example 1

[0081] Reference Figure 1 The present embodiment provides a tunnel, tunnel 360-degree panoramic image acquisition control system, which comprises the following modules: image acquisition device, focal length control module, light supplement control module and image intelligent feedback control module;

[0082] Reference Figure 2 The image acquisition device is used to receive a shooting control signal to collect the full-section photo image and section contour point cloud of the tunnel and tunnel; the image acquisition device comprises a ring collection module 1, a protection device 4 and a laser radar 6; the ring collection module 1 and the laser radar 6 are installed on the protection device 4; the shooting control signal comprises a camera focal length adjustment signal, a light source intensity signal of a light supplement lamp and a camera shooting signal; the image acquisition device adjusts the shooting focal length of the camera, adjusts the light source signal intensity of the light supplement lamp and whether to shoot according to the shooting control signal; the laser radar 6 is used to collect the section contour point cloud of the tunnel and tunnel;

[0083] Reference Figure 3 , the ring-shaped acquisition module 1 is arranged by 12 acquisition units 11 at equal angles, the field angle overlap of adjacent acquisition units 11 is greater than 50%, a laser radar 6 is arranged at the tail end for acquiring the cross section profile point cloud of the tunnel and the tunnel, and the laser radar 6 is used for transmitting the cross section profile point cloud to the focal length control module in real time, and the ring-shaped acquisition module 1 is parallel to the cross section of the tunnel and the tunnel during acquisition;

[0084] Reference Figure 4 , the acquisition unit 11 comprises a camera 111, a lens 112, a heat dissipation device 113 and a light supplement lamp 114; the acquisition unit 11 controls the shooting focal length of the camera 111 according to the camera focal length adjustment signal; the acquisition unit 11 controls the light supplement lamp light source intensity of the light supplement lamp 114 during work according to the light supplement lamp light source intensity signal; the acquisition unit 11 controls whether the camera 111 starts shooting according to the camera shooting signal;

[0085] The laser radar 6 is a single-line laser radar or a multi-line laser radar, preferably, the laser radar 6 in the embodiment is a single-line laser radar;

[0086] During acquisition, the laser radar 6 is parallel to the tunnel axis and perpendicular to the cross section of the tunnel and the tunnel;

[0087] The protection device 4 is used for the pose fixation, anti-collision and sealing and waterproof functions of the ring-shaped acquisition module 1 and the laser radar 6, and the ring-shaped acquisition module 1, the protection device 4 and the laser radar 6 are modularized and combined in a detachable manner;

[0088] The camera 111, the lens 112, the heat dissipation device 113 and the light supplement lamp 114 are also modularized and combined in a detachable manner;

[0089] The light supplement control module is connected with the image acquisition device and the image intelligent feedback control module respectively, and is used for sending the light supplement lamp light source intensity signal to control the light supplement lamp light source intensity during the work of the image acquisition device;

[0090] The focal length control module, the light supplement control module and the image intelligent feedback control module can be located in the upper computer, the data acquired by the ring-shaped acquisition module 1 and the laser radar 6 is transmitted to the upper computer in a wireless transmission manner, the upper computer processes the data, and then controls the ring-shaped acquisition module 1 and the laser radar 6, or a chip module is installed in the ring-shaped acquisition module 1, and the ring-shaped acquisition module 1 and the laser radar 6 are controlled after the data is processed locally.

[0091] In the embodiment, Figure 2 2 in the focal length control module, Figure 2 3 in the light supplement control module andFigure 2 The image intelligent feedback control module 5 is located in the computer program of the ESP32 module inside the ring-shaped acquisition module 1, and the program in the ESP32 module controls the focal length of the lens and the light supplement intensity to play the role of adaptive focusing and image intelligent feedback control.

[0092] It can be understood that the computer program can be divided into one or more modules, which are stored in the memory and executed by the ESP32 module to complete the present application. The computer program can be divided into various modules such as the focal length control module, the light supplement control module, and the image intelligent feedback control module in the embodiment. The specific functions of each module are described in the working process of the module in the embodiment, which will not be repeated here.

[0093] Reference Figure 5 The heat dissipation device 113 is composed of a heat sink 1131, a fan 1132, and a heat dissipation air outlet 1133.

[0094] Reference Figure 6 The light supplement lamp 114 is composed of a light source lens 1141, a COB light source 1142, and a light homogenizing film 1143 arranged at the tangential ends of the camera 111 respectively; the heat dissipation device 113 is located at the back of the light supplement lamp 114 for heat dissipation of the light supplement lamp 114.

[0095] As shown in Figure 7 The protection device 4 adopts a cover plate 1121, a radial protection support 1122, and a circumferential protection support 1123 for pose fixation and anti-collision protection.

[0096] As shown in Figure 8 The cover plate 1121 is composed of a lens hole 1121-1, a wire passing hole 1121-2, a reinforcing rib 1121-3, and a waterproof groove 1121-4.

[0097] The heat dissipation device 113, the light supplement lamp 114, and the camera 111 are powered by a 4824DC installed on a mobile device, connected to the ring-shaped acquisition module 1 through a power line;

[0098] The focal length control module is connected with the image acquisition device, used for calculating the distance between each lens of the image acquisition device and the wall according to the cross-sectional contour point cloud data of the tunnel or the tunnel, adjusting the focal length of each camera in the image acquisition device according to the distance between each lens of the image acquisition device and the wall, and sending a camera focal length adjustment signal to the image acquisition device; the camera focal length adjustment signal includes a camera number and a corresponding focal length.

[0099] The laser radar 6 acquires the cross-sectional contour point cloud at the front end of the ring-shaped acquisition module 1 at 0.5 seconds; , ; wherein, is the radius of the polar coordinate at time t; is the angle of the polar coordinate at time t, j is the serial number of the point cloud; N is the total number of cross-section contour point clouds;

[0100] Calculate the pose of the ring-shaped acquisition module 1 based on the laser radar 6 (only consider the tunnel cross-section plane) , wherein, and are the angles of the laser radar calculation relative to the current cross-section (XOY plane), the ring-shaped acquisition module 1 is offset on the X and Y axes at time t;

[0101] It can be understood that, can be directly obtained by the laser radar, and the laser radar can generate the coordinate and angle relationship at any time and the initial position, which is prior art and will not be described here.

[0102] The coordinates of the cross-section contour point cloud are converted into the cross-section contour Cartesian point cloud = The conversion process is as follows:

[0103] ;

[0104] wherein, is the X coordinate of the cross-section Cartesian point cloud at time t, is the Y coordinate of the Cartesian point cloud at time t;

[0105] Extract the cross-section point cloud slice of the ring-shaped acquisition module 1 front end within seconds :

[0106] ;

[0107] wherein, is the slice thickness; in this embodiment 0.1m is taken; , indicating the position of the acquisition unit 11 at time t; is the point cloud coordinate set of the cross-section point cloud slice, is the preset scanning time, which is the sum of the minimum time of the laser radar scanning one circle and the time of the camera starting shooting, which can be obtained by early testing or according to the general experience value of 0.15-0.3 seconds; v is the moving speed of the acquisition unit 11; is the partial derivative function; indicates the partial derivative of .

[0108] It can be understood that the annular acquisition module 1 includes 12 acquisition units 11, and the lens 112 is the most important part of the acquisition unit 11, so it can be considered that the pose of the annular acquisition module 1 is composed of the positions of the lenses 112 of the 12 acquisition units 11;

[0109] The embodiment provides a method for numbering the cameras 111 in the acquisition unit 11. The cover plate 1121 of the protection device 4 is parallel to the section of the tunnel or the tunnel, the camera 111 of one of the acquisition units 11 is vertically upward and perpendicular to the ground horizontal line, the camera 111 is set as a starting camera, and the number is set as 1. The camera is named the first camera. Then, the numbers of the other cameras are sequentially set as 2, 3,..., 12 in a clockwise direction, and the cameras are sequentially named the second camera, the third camera,..., and the twelfth camera. The plane formed by the 12 cameras 111 is also parallel to the section of the tunnel or the tunnel, that is, the center coordinates of the cameras 111 of the 12 acquisition units 11 are in the same horizontal plane. The numbering method can correspondingly number the acquisition units 11, the lenses 112 and the light supplement lamps 114, so that the corresponding acquisition unit, the lens and the light supplement lamp of the camera can be easily found subsequently.

[0110] Based on the first camera, the azimuth angle of each camera is calculated , unit: degree; wherein i is the number of the camera, i={1, 2, 3,..., 12};

[0111] Since the position of the lens 112 is fixed compared with the position of the laser radar 6, the distance between the two is L, and the angle is The conversion relationship between the i-th lens 112 and the position of the laser radar 6 is calculated, and the lens position is:

[0112] ;

[0113] wherein, , , indicates the center coordinate of the laser radar 6 at t seconds;

[0114] The distance between each lens 112 and the tunnel wall is calculated ;

[0115] wherein, , , indicates the point cloud coordinate of the largest tunnel wall point of the i-th lens at t seconds, which is obtained through the Cartesian coordinate conversion in the above step indicates the position coordinate of the i-th lens at t seconds; i indicates the camera number, and also represents the number of the camera corresponding to the lens; i={1, 2,..., 12};

[0116] To obtain the Cartesian point cloud of the cross-sectional profile at second t, given the angles of each camera, we can calculate the point cloud along the ray by drawing a ray from the lidar as the starting point in the direction of the camera angle. , Distance from camera lens The distance is taken as the maximum value as the distance between the i-th camera lens and the cave wall. The corresponding point cloud coordinates are used as the point cloud coordinates of the cave wall with the greatest distance from the camera lens. , , ); where the point cloud is the set of point cloud coordinates of a slice of cross-sectional point cloud. ; Calculate the distance between all camera lenses and the cave wall in the same manner as described above;

[0117] Since the image acquisition device, namely the circular acquisition module 1, moves along the tunnel (or tunnel) axis, and the cross-section of the circular acquisition module 1 is perpendicular to this axis, the z-coordinate remains unchanged. Therefore, when calculating the distance from the lens to the tunnel wall, it is not necessary to calculate the z-axis distance. The formula for the distance between each lens 112 and the tunnel wall can be simplified as follows:

[0118] ;

[0119] The lidar 6 acquires data from the front end of the ring acquisition module 1. Point cloud of cross-sectional profile at a second The center coordinates of the lidar will be recorded at that time. , , Coordinates of the center of adjacent lidar , , The center coordinates of adjacent lidars are The center coordinates of the lidar are primarily used to record the acquisition location. Since the acquisition interval and speed are fixed, the distance between the center coordinates of two adjacent acquisitions should be consistent. Considering that deviations may occur when the acquisition device moves, a threshold is set. To avoid floating differences, Generally, 1.2vt is used, which is 1.2 times the sampling interval;

[0120] The distance between the center coordinates of the lidar and the center coordinates of the adjacent lidars The calculation formula is as follows:

[0121] ;

[0122] When the ring acquisition module 1 acquires data, it automatically records the center coordinates of the lidar 6. , , and calculate the distance between the center coordinates of the adjacent laser radars If all of them are less than or equal to a threshold value , it is indicated that the tunnel full-section image acquisition is completed; if any of them is greater than the threshold value , it is indicated that the acquisition interval is too large, and the acquisition needs to be restarted from the coordinate point with a smaller time value. The focal length of each camera 111 in the image acquisition device is adjusted according to the distance between each lens and the tunnel wall by the following formula

[0123] :

[0124] ;

[0125] wherein fi(t) represents the focal length of the i-th camera at t seconds, , , is the standard focal length of the lens, which is a fixed value provided by the hardware and is a fixed value. In this embodiment, the standard focal lengths of the lenses of the 12 cameras are consistent.

[0126] Then, a camera focal length adjustment signal is sent to the image acquisition device, and the camera focal length adjustment signal includes the camera number i and the corresponding focal length f i(t). ;

[0127] The light supplement lamp 114 is composed of COB light sources 1142 arranged at the tangential ends of the camera 111. In order to prevent the mirror reflection caused by the large intensity of the main light when the tunnel wall is filled with water, the preset light supplement strategy is adopted during the acquisition, that is, the intensity of the on-axis light source of the image acquisition device is less than the intensity of the adjacent light source of the camera , wherein: , ; k is a light supplement adjustment coefficient.

[0128] The values of the intensity of the on-axis light source of the camera and the intensity of the adjacent light source of the camera are obtained by setting and obtaining the configuration file.

[0129] Referring to the structure of the acquisition unit shown in Figure 4 , one acquisition unit 11 includes two light supplement lamps 114, that is, the two light supplement lamps 114 are the on-axis light sources of the camera 111, the adjacent light sources of the camera are the light supplement lamps of the two acquisition units closest to the left and right of the acquisition unit where the camera is located, the intensity of the on-axis light source of the camera is the intensity of the light source of the light supplement lamp 114 during the shooting, and the intensity of the adjacent light source is the intensity of the light source of the light supplement lamp of the two acquisition units closest to the left and right of the acquisition unit where the camera is located. The surface of the light supplement lamp 114 is covered with a light uniformity film. ​​​

[0130] The upper limit of the value range of the light supplement adjustment coefficient k needs to be given in combination with the field use conditions, and in the embodiment, the upper limit of the value range of k is 1.5; in the embodiment, the acquisition unit 11 is used in a water tunnel, and the brightness of the main light source, that is, the center light source (the coaxial light source of the camera) during shooting is adjusted to be low, and the light uniformity film can solve the problem of reflection in combination.

[0131] The light uniformity film is covered on the surface of the light supplement lamp 114, that is, covered on the surface of the light source lens 1141, and mainly functions to improve the uniformity of the light source, make the light more soft and uniform, and eliminate the phenomenon of bright spots or uneven brightness; at the same time, the tunnel bottom plate also has accumulated water, so the camera light supplement lamp / light source facing the bottom plate should be adjusted to be minimum, and the light supplement lamps on both sides should be slightly larger, and for the adjacent part of the tunnel side wall and the bottom plate, the light source close to the bottom plate should be smaller, and the light source close to the tunnel wall should be larger.

[0132] The value of the intensity of the coaxial light source of the camera and the intensity of the light source adjacent to the camera is mainly used to solve the problem of reflection caused by accumulated water, and needs to be manually adjusted and determined, and through the above processing, it can be ensured that the subsequent image quality evaluation is not affected by the reflection.

[0133] In the water tunnel, due to the long time in the humid environment in the tunnel, the water film will be formed on the tunnel wall due to the accumulated water, and if the brightness of the center light source during shooting is kept normal, the incident angle of the light source will be equal to the reflection angle, because the center light source is installed at the coaxial position of the camera, it will be reflected back to the camera by the water film on the tunnel wall in a mirror reflection manner, resulting in the consequence that the intensity of the reflected light is much higher than that of the scattered light, and a high-brightness light spot will be formed in the imaging area; by adjusting the brightness of the center light source to be low, the incident light energy is directly weakened, and the mirror reflection light energy is also weakened, but directly reducing the brightness will lead to the reduction of the overall brightness and the deterioration of the image quality, therefore, the light angle reconstruction is performed by using the light uniformity film, the light uniformity film can scatter a concentrated light beam into a wide-angle diffuse light, so that the light spot effect caused by the high brightness of the reflected light can be avoided, and the overall brightness change of the image caused by the low light can also be avoided.

[0134] In order to avoid the conflict between the light supplement lamps in the two adjacent acquisition units, therefore, in the embodiment, the photos of the tunnel and the tunnel are collected in batches, first, the first camera, the fourth camera, the seventh camera and the tenth camera and the corresponding light supplement equipment (the coaxial light source of the camera and the light source adjacent to the camera) are controlled to perform the first batch of collection, after the first batch of collection is completed, the second camera, the fifth camera, the eighth camera and the eleventh camera and the corresponding light supplement equipment are controlled to perform the second batch of collection, after the second batch of collection is completed, the third camera, the sixth camera, the ninth camera and the twelfth camera and the corresponding light supplement equipment are controlled to perform the third batch of collection, and the total light supplement range of each batch of light supplement equipment can cover the whole section of the tunnel and the tunnel;

[0135] The embodiment adopts 12 cameras, lenses and light supplement lamps arranged in an equidistant ring shape, and the focal length of each lens is adjusted in real time according to the detection distance of the tail end laser radar during the collection process (the distance between each camera and the tunnel wall after t seconds is extracted in real time by the laser radar, and then the focal length and light supplement are dynamically adjusted), which can adapt to different cross-section tunnels and ensure the image quality.

[0136] The image intelligent feedback control module is connected with the image acquisition device, the focal length control module and the light supplement control module, is used for evaluating the quality comprehensive score of the photo image collected by the ring-shaped collection module 1 of the image acquisition device each time according to the motion blur, the defocus blur and the light balance index, and adding the center coordinates of the image acquisition device to the abnormal coordinate list when the quality comprehensive score is lower than the preset quality score threshold, and executing the re-collection strategy at each coordinate point in the abnormal coordinate list when returning, until leaving the tunnel and the tunnel;

[0137] The center coordinates of the image acquisition device are the center coordinates of the laser radar 6;

[0138] The focal length control module is connected with the image intelligent feedback control module, receives the camera focal length adjustment signal feedback by the image intelligent feedback control module, controls the focal length of the camera in the image acquisition device and then re-collects;

[0139] The light supplement control module is connected with the image intelligent feedback control module, receives the light source intensity signal feedback by the image intelligent feedback control module, controls the light source intensity of the camera in the image acquisition device and then re-collects;

[0140] The quality comprehensive score is calculated based on the following formula:

[0141] ;

[0142] Wherein, is the quality comprehensive score; is the motion blur; is the defocus blur; is the light balance index; is the motion blur weight; is the defocus blur weight; is the light balance index weight;

[0143] In the embodiment 0.4 is taken, 0.3 is taken, 0.3 is taken, The value of the motion blur weight can also be changed according to actual needs;

[0144] The motion blur is calculated based on the following formula:

[0145] ;

[0146] wherein, is the effective edge gradient mean value of the collected photo image, is the reference clear image gradient threshold value;

[0147] is the average value of all gradient amplitudes greater than or equal to the effective edge point;

[0148] ;

[0149] wherein, is the sum of gradient amplitudes of all effective edge points, is the total number of pixels with gradient amplitudes greater than or equal to the effective edge point threshold value; the effective edge point threshold value is 1 / 2 of the maximum value of the gradient amplitude of the image;

[0150] traverse the gradient amplitudes of all pixels in the collected photo image , find the maximum value of the gradient amplitude, and then 1 / 2 of the maximum value of the gradient amplitude will be used as the effective edge point threshold value;

[0151] Then traverse the gradient amplitudes of all pixel points in the collected photo image I again, record the coordinates of all pixel points with gradient amplitudes greater than or equal to the effective edge point and their gradient amplitudes;

[0152] Finally, the gradient amplitudes of all pixel points with gradient amplitudes greater than the effective edge point are accumulated and divided by the total number of all pixel points with gradient amplitudes greater than or equal to the effective edge point to obtain ;

[0153] The gradient amplitude is calculated based on the following formula:

[0154] ;

[0155] wherein, represents the gradient amplitude at pixel coordinates (x, y), is the horizontal gradient; is the vertical gradient;

[0156] The horizontal gradient ;

[0157] The vertical gradient: ;

[0158] Wherein, x represents the pixel horizontal coordinate of the photo image, y represents the pixel vertical coordinate of the photo image; I(x+a, y+b) represents the gray value of the photo image I at the pixel coordinate (x+a, y+b); wherein The template in the horizontal direction is generally [-1 01]; The template in the vertical direction is generally [1 2 1];

[0159] Taking I(x+a, y+b) as the center point, the peripheral 8-neighborhood pixel points are distributed as follows:

[0160] ;

[0161] Then

[0162]

[0163] ;

[0164]

[0165] = ;

[0166] The reference clear image gradient threshold According to the initial shooting reference group image, the value is taken according to different application scenarios; before each new scene application, artificial calibration is first carried out, and the qualified photo is artificially controlled as the reference group image.

[0167] The defocus blur degree The defocus blur degree is calculated based on the following formula:

[0168] ;

[0169] Wherein, Indicates the Laplacian operator of the image I at the pixel point (x, y), indicates the second-order derivative of the image at the point, and is the default expression; M represents the height of the photo image I; N represents the width of the photo image I; Indicates the variance of the Laplacian filtered image;

[0170] The average value of the square of the Laplacian operator of the image on the whole image is calculated to measure the smoothness of the image, so as to represent the defocus blur degree of the image;

[0171] The light balance index The light balance index is calculated based on the following formula:

[0172] ;

[0173] Wherein, e is the base of natural logarithm; This represents the average brightness of the current image. The average brightness of the reference group images varies with different application scenarios and is determined based on the initial captured baseline image. The attenuation coefficient has a value range of [0.02, 0.05]. In this embodiment... Take 0.03;

[0174] in, and The average luminance is calculated using the formula, which is as follows:

[0175] ;

[0176] Where M represents the height of photo image I; N represents the width of photo image I; and I(x,y) represents the gray value of photo image I at pixel coordinates (x,y).

[0177] And when the overall quality score is lower than the preset quality score threshold, the center coordinates of LiDAR 6 will be added to the abnormal coordinates list, specifically including:

[0178] The ring-shaped acquisition module 1 has 12 cameras. The images captured by each camera need to be evaluated for overall quality. If the overall quality score of all images is greater than or equal to the preset quality score threshold, then the image captured by the ring-shaped acquisition module 1 at the current coordinate point is considered to meet the shooting requirements. If the overall quality score of any image is lower than the preset quality score threshold, then the image captured by the ring-shaped acquisition module 1 at the current coordinate point is considered to be lower than the preset quality score threshold, and the image captured by the ring-shaped acquisition module 1 at the current coordinate point is considered to not meet the shooting requirements. The center coordinates of the LiDAR 6 at that time are then added to the abnormal position list.

[0179] The preset quality score threshold Take 0.8 or the overall quality score calculated based on the reference group images;

[0180] The re-acquisition strategy specifically includes:

[0181] Re-acquire at the current coordinate point, and then perform a comprehensive quality score on the re-acquired photo image. If the recalculated comprehensive quality score is lower than the preset quality score threshold, a graded adjustment strategy is executed; otherwise, continue to return to the next coordinate point in the abnormal coordinate list.

[0182] The graded adjustment strategy includes: first, executing a motion blur adjustment strategy; then, executing an abnormal lighting adjustment strategy; and finally, executing a defocus blur adjustment strategy.

[0183] Further, the hierarchical adjustment further comprises: performing quality comprehensive scoring on the last re-acquired photo image, judging that the quality comprehensive score of the photo image is greater than or equal to the preset quality score threshold, and performing review.

[0184] The re-acquisition is performed by adopting batch-by-batch acquisition of the photo of the tunnel or the tunnel hole by the ring-shaped acquisition module 1 of the image acquisition device.

[0185] It can be understood that, during the initial shooting, that is, the shooting when entering the hole, due to special design of the tunnel or the tunnel hole, such as aluminum electric box equipment with reflective material or water accumulation on the hole wall, which will cause reflection, blur or focal length change (branch hole or fire-fighting equipment storage window) will be involved during acquisition. Since the position of the above-mentioned equipment is not fixed, if the situation is encountered, the focal length and light compensation intensity will be adjusted, which will affect the subsequent shooting parameters. Since most of the tunnel or tunnel structure is continuous, we generally make fine adjustments when entering the hole, and do not make large changes to the subsequent shooting parameters. However, when returning, we have enough time to adjust the parameters, and the quality of the position that does not meet the requirements is changed and adjusted. When returning to the coordinate point in the abnormal coordinate list, first, keep all parameters unchanged, re-acquire, and calculate the quality comprehensive score. Since the motion blur is generated during motion shooting, it is an accidental factor, and the phenomenon may not occur during re-acquisition when returning. If it still does not meet the requirements, the motion blur adjustment strategy is performed again. In addition, the motion blur and light abnormality may occur at the same time, so after solving the motion blur, the light abnormality may still occur. Therefore, it is necessary to make multi-faceted judgments, preferentially process the motion blur, because the static acquisition can solve this problem, then process the light, and finally process the defocus blur. After adjusting the three, the quality comprehensive score of the re-acquired photo image is calculated again, that is, the quality comprehensive score of the photo image is greater than or equal to the preset quality score threshold, and the re-acquisition of the current coordinate point is exited to go to the next coordinate point.

[0186] The motion blur adjustment strategy specifically comprises: if the motion blur degree of the coordinate point in the abnormal position list is greater than the preset motion blur degree threshold , it is considered that the motion blur occurs, re-acquisition is performed at the coordinate point, and then the motion blur degree of the photo image acquired by the ring-shaped acquisition module 1 is calculated again until the motion blur degree of the newly acquired photo image is less than or equal to the preset motion blur degree threshold or the preset maximum re-shooting number is reached, and the re-acquisition is stopped.

[0187] The preset motion blur degree threshold is 0.3 in the embodiment , which can be changed according to actual needs.

[0188] Because the motion blur is generally generated during the movement, the problem can be solved by stopping at the abnormal position point to take a picture (static collection) during the return; the preset maximum number of rephotographing times is 3 in the embodiment, which can be changed according to actual requirements;

[0189] The light abnormality adjustment strategy specifically comprises: If the light balance index of the reference group is greater than the light balance index of the current image, it is considered that the image is overexposed, the linear adjustment factor is adjusted to be 1, and the current image is re-collected.

[0190] The light brightness of the supplementary lamp is adjusted according to the following formula:

[0191] ;

[0192] Wherein, is the light balance index of the reference group, which changes with different application scenarios and is valued according to the reference image initially shot; is the light brightness of the supplementary lamp when the light is abnormal, which is equal to the initial light brightness I of the supplementary lamp when appearing for the first time; The initial light brightness I of the supplementary lamp is a fixed value, which is generally unchanged; is a linear adjustment factor, which is equal to 1 when the image is overexposed, and is equal to 0 when the image is underexposed.

[0193] If the light balance index of the reference group is less than the light balance index of the current image, it is considered that the image is underexposed, and the linear adjustment factor is adjusted to be 0.

[0194] ;

[0195] If the light balance index of the reference group is greater than the light balance index of the current image, it is considered that the image is overexposed, and the linear adjustment factor is adjusted to be 1.

[0196] ;

[0197] If the light balance index of the reference group is equal to the light balance index of the current image, it is considered that the image light is normal, and no adjustment is performed.

[0198] Wherein, is a floating difference value, which represents a certain degree of change is considered to be reasonable by default, and is generally taken as 10; is the average brightness of the current image; is the average brightness of the reference group image; the preset adjustment number is 3, which can be changed according to specific requirements; is an underexposure adjustment coefficient, ; is an overexposure adjustment coefficient,​​ ;

[0199] It can be understood that the above-mentioned returning to the point for supplementing the light brightness adjustment is to adjust all light brightness, and only represent the light brightness after adjustment and before adjustment, and has nothing to do with whether it is a coaxial light source and the intensity of adjacent light sources, and all need to be adjusted.

[0200] The defocus blur adjustment strategy specifically comprises: if the defocus blur degree is greater than the defocus blur degree threshold, it is considered that defocus blur occurs, and the focal length adjustment amount is added to the focal length f of the camera at the coordinate point, and then the image is reacquired, and then the quality comprehensive score of the reacquired image is calculated, until the score quality requirement is met, and the reacquisition is stopped.

[0201] Wherein, is the defocus blur degree of the reference group photograph, which is obtained according to the reference group photograph;

[0202] The focal length adjustment amount is calculated according to the following formula:

[0203] ;

[0204] Wherein, is a distance-focal length conversion coefficient, which is calibrated by experiment, and is different for different types of cameras ; is the distance between the ith lens and the hole wall; is the current camera focal length imaging best distance, which is obtained by experiment calibration; is the focal length adjustment compensation item, generally taking 0.1-0.3 times of the focal length ; is the focal length of the ith camera at t seconds, , and sgn is a sign function; is the image comprehensive quality fluctuation error value, the value range is (0, 0.1), and in the embodiment, 0.05 is taken; is an indicator function (1 when the condition is true, otherwise 0); is the quality comprehensive score of the image at the current time t, is the quality comprehensive score of the image at the previous time t-1;

[0205] At the coordinate point, the focal length adjustment amount is added to the focal length f of the camera, and then the image is reacquired, and then the quality comprehensive score of the image reacquired by the ring-shaped acquisition module 1 is calculated, until the score quality requirement is met, and the reacquisition is stopped. The specific steps are as follows:

[0206] A1, on the basis of the focal length f, the focal length adjustment amount After reacquiring, then re-evaluating the quality of the reacquired photo image; the focal length f is the focal length of the camera at the point coordinate when returning;

[0207] A2, if the quality score of the current batch of photo images is greater than the quality score of the previous batch of photo images, it proves that the current adjustment direction is correct, and jump to A3; otherwise, it means that the current adjustment direction is incorrect, and jump to A4;

[0208] A3, verify whether the quality score of the current batch of photo images meets the quality score requirement, if yes, continue to return; if not, increase the focal length adjustment amount by 1 / 2 based on the current focal length After reacquiring, then re-evaluating the quality of the reacquired photo image, if the quality score of the current batch of photo images is less than the quality score of the previous batch of photo images, decrease the focal length adjustment amount by 1 / 2 based on the current focal length After reacquiring, until the quality score of the current batch of photo images is greater than the quality score of the previous batch of photo images and the quality score of the current batch of photo images meets the quality score requirement;

[0209] A4, decrease the focal length adjustment amount by 1 / 2 based on the focal length f After reacquiring, verify whether the quality score of the current batch of photo images meets the quality score requirement, if yes, continue to return; if not, decrease the focal length adjustment amount by 1 / 2 based on the current focal length After reacquiring, then re-evaluating the quality of the reacquired photo image, if the quality score of the current batch of photo images is less than the quality score of the previous batch of photo images, increase the focal length adjustment amount by 1 / 2 based on the current focal length After reacquiring, until the quality score of the current batch of photo images is greater than the quality score of the previous batch of photo images and the quality score of the current batch of photo images meets the quality score requirement;

[0210] The quality score requirement is that the quality score of the current batch of photo images is greater than or equal to the preset quality score threshold or reaches the image quality stability condition; it is worth noting that the focal length adjustment amount is recalculated after each reacquisition And the quality score of the current batch of photo images that meets the quality score requirement is taken as the new preset quality score threshold;

[0211] The image quality stability condition is: After reaching this condition, the focal length does not need to be adjusted;

[0212] It can be understood that the focal length control module is connected with the image intelligent feedback control module, when the focal length of the camera needs to be changed in the image intelligent feedback control module, the focal length adjustment signal of the camera is sent to the focal length control module, and the focal length control module receives the focal length adjustment signal of the camera feedback by the image intelligent feedback control module to control the focal length of the camera in the image acquisition device and then reacquire;

[0213] It can be understood that the light supplement control module is connected with the image intelligent feedback control module, when the light source intensity of the light supplement lamp needs to be changed in the image intelligent feedback control module, the light source intensity signal of the light supplement lamp is sent to the light supplement control module, and the light supplement control module receives the light source intensity signal of the light supplement lamp feedback by the image intelligent feedback control module to control the light source intensity of the light supplement lamp in the image acquisition device and then reacquire;

[0214] In addition, in order to prevent the mirror reflection caused by the large intensity of the main light when the water accumulates on the wall of the tunnel, the intensity of the coaxial light source of the camera is smaller than the intensity of the adjacent light source, the single-chip microcomputer firstly controls the first group of cameras (numbers 1, 4, 7 and 10) and the corresponding light supplement equipment to collect, the second group of cameras (numbers 2, 5, 8 and 11) and the corresponding light supplement equipment to collect subsequently, and the third group of cameras (numbers 3, 6, 9 and 12) and the corresponding light supplement equipment to collect lastly, and the total light supplement range of each batch of light supplement equipment can cover the whole section of the tunnel; the quality evaluation and integrity detection are performed by judging whether the image is blurred or the illumination is abnormal, the detection result generates a positioning point cloud map through the control system, and the positioning point reacquisition is automatically triggered when returning.

[0215] In the embodiment, 12 groups of variable-focus lens arrays are distributed at equal angles, the focal length is adjusted in real time according to the detection data of the laser radar at the tail end of the acquisition module, and the acquisition module can be applied to tunnels with different sections. The light supplement lamps adopt a time-sharing and grouping driving mechanism, the lenses and the corresponding light supplement lamps are equally divided into four groups at equal intervals, and the range of the four groups of lenses and light supplement lamps can completely cover the whole section of the tunnel. Meanwhile, the light supplement lamps are covered with anti-glare structures with light-film membranes, which can effectively eliminate the water stain reflection interference in the process of acquisition. The quality evaluation and integrity detection are performed on the collected photo images according to the motion blur, defocus blur and illumination balance index, the detection result generates a positioning point cloud map through the control system, and the positioning point reacquisition is automatically triggered when returning.

[0216] The embodiment solves the problems of great difficulty in image full-section acquisition in a complex tunnel environment, single application range of an acquisition module and data integrity, and significantly improves the tunnel panoramic image acquisition efficiency and precision.

[0217] Embodiment two

[0218] refer to Figure 9 This embodiment provides a method for acquiring and controlling 360-degree panoramic images of tunnels and culverts, based on the acquisition and control system for 360-degree panoramic images of tunnels and culverts described in Embodiment 1, including the following steps:

[0219] S1. Using an image acquisition device, advance along the tunnel axis and acquire photographic images and point clouds of the tunnel's full cross-section using a preset supplementary lighting strategy; the preset supplementary lighting strategy refers to the intensity of the coaxial light source of the camera in the image acquisition device. Less than the intensity of the light source adjacent to the camera ,in: k is the fill light adjustment coefficient;

[0220] S2. Calculate the distance between each lens of the image acquisition device and the tunnel wall based on the point cloud data of the cross-sectional contour of the tunnel.

[0221] S21. Acquire data from the front end of the ring acquisition module 1 using the lidar 6. Point cloud of cross-sectional profile at a second , ;in, Let be the radius of the polar coordinates at time t; t represents the polar coordinate angle at time t, j represents the point cloud index, and N represents the total number of point clouds in the cross-sectional profile.

[0222] S22. Draw point cloud of cross-sectional outline Cartesian coordinate transformation is performed on the coordinates to obtain the cross-sectional profile Cartesian point cloud. = The conversion process is shown in the following formula:

[0223] ;

[0224] in, , These are the angles of offset on the X and Y axes of the ring acquisition module 1 at time t, calculated by the lidar relative to the current cross-section.

[0225] S23, Extraction Cross-sectional point cloud slices at the front end of the ring acquisition module 1 within seconds :

[0226] ;

[0227] in, The thickness of the slice; , indicating the position of acquisition unit 11 at time t; It is the set of point cloud coordinates for a slice of cross-sectional point cloud; is a preset scanning time; v is the moving speed of the acquisition unit 11; is a partial derivative function;

[0228] S24, the distance of the lens 112 from the hole wall at t seconds is calculated according to the following formula :

[0229] ;

[0230] wherein, , , represents the point cloud coordinate of the farthest hole wall point from the i-th lens at t seconds; represents the position coordinate of the i-th lens at t seconds; i={1, 2,..., 12};

[0231] S3, the focal length of each camera in the image acquisition device is dynamically adjusted according to the distance of each lens of the image acquisition device from the hole wall by the following formula

[0232] ;

[0233] wherein, represents the focal length of the i-th camera at t seconds, , is the standard focal length of the lens;

[0234] S4, the quality comprehensive score is obtained by evaluating the photo image collected each time by the image acquisition device according to the motion blur degree, the defocus blur degree and the illumination balance index, and the center coordinate of the image acquisition device is added to the abnormal coordinate list when the quality comprehensive score is lower than a preset quality score threshold; the re-collection strategy is executed at each coordinate point in the abnormal coordinate list when returning, until leaving the tunnel or the cave.

[0235] The parts of the present application not described in detail are prior art, and it is obvious for those skilled in the art that the present application is not limited to the details of the above exemplary embodiments, and the present application can be realized in other specific forms without departing from the spirit or essential characteristics of the present application; therefore, from any point of view, the embodiments should be regarded as exemplary and non-limiting, and all changes falling within the meaning and scope of the equivalent elements are intended to be included in the present application.

Claims

1. A 360-degree panoramic image acquisition and control system for tunnels and culverts, characterized in that, It includes an image acquisition device, a focus control module, a fill light control module, and an image intelligent feedback control module; The image acquisition device is used to receive shooting control signals to acquire photographic images of the entire cross-section of the tunnel and the cross-sectional contour point cloud; the image acquisition device adjusts the camera's shooting focal length, adjusts the light source signal intensity of the supplementary light, and decides whether to shoot according to the shooting control signals; the shooting control signals include camera focal length adjustment signals and supplementary light source intensity signals. The focal length control module is connected to the image acquisition device and is used to calculate the distance between each lens of the image acquisition device and the tunnel wall based on the cross-sectional contour point cloud data of the tunnel, adjust the focal length of each camera in the image acquisition device according to the distance between each lens of the image acquisition device and the tunnel wall, and send the camera focal length adjustment signal to the image acquisition device. The supplementary lighting control module is connected to the image acquisition device and is used to send a supplementary lighting light source intensity signal to control the supplementary lighting light source intensity when the image acquisition device is working; The image intelligent feedback control module is connected to the image acquisition device, the focal length control module and the supplementary lighting control module. It is used to evaluate the photos acquired by the image acquisition device each time based on motion blur, defocus blur and illumination balance index to obtain a comprehensive quality score. When the comprehensive quality score is lower than the preset quality score threshold, the center coordinates of the image acquisition device are added to the abnormal coordinate list. When returning, a re-acquisition strategy is executed at each coordinate point in the abnormal coordinate list until the tunnel is left. The overall quality score The following formula was used to calculate: ； in, A comprehensive quality score; Motion blur; Defocus blur; The light balance index; For motion blur weights; Weights for out-of-focus blur. Weights for the light balance index; The motion blur The following formula was used to calculate: ； in, The effective edge gradient mean of the acquired photographic image. Use a clear image gradient threshold as a benchmark; The defocus blur The following formula was used to calculate: ； in, Let M represent the Laplacian operator at pixel (x,y) of image I; M represents the height of image I; and N represents the width of image I. This represents the variance of the image after Laplacian filtering; The light balance index The following formula was used to calculate: ； Where e is the base of the natural logarithm; This represents the average brightness of the current image. The average brightness of the reference group images; This is the attenuation coefficient.

2. The acquisition and control system for 360-degree panoramic images of tunnels and culverts according to claim 1, characterized in that, The image acquisition device includes a ring acquisition module (1) and a lidar (6); the lidar (6) is used to acquire point clouds of the cross-sectional contours of tunnels and tunnels; the ring acquisition module (1) consists of 12 acquisition units (11) arranged at equal intervals in the ring direction; the acquisition unit (11) includes a camera (111), a lens (112) and a fill light (114); the camera focal length adjustment signal includes the camera number and the corresponding focal length; the fill light source intensity signal includes the camera number and the corresponding fill light source intensity; the image acquisition device acquires photos of tunnels and tunnels in batches.

3. The acquisition and control system for 360-degree panoramic images of tunnels and culverts according to claim 2, characterized in that, The calculation of the distance between the image acquisition device and the tunnel wall based on the cross-sectional contour point cloud data of the tunnel includes the following steps: S21. Acquire the data from the front end of the ring acquisition module (1) using the lidar (6). Point cloud of cross-sectional profile at a second , ;in, Let be the radius of the polar coordinates at time t; t represents the polar coordinate angle at time t, j represents the point cloud index, and N represents the total number of point clouds in the cross-sectional profile. S22. Draw point cloud of cross-sectional outline Cartesian coordinate transformation is performed on the coordinates to obtain the cross-sectional profile Cartesian point cloud. = The conversion process is shown in the following formula: ； in, These are the angles of the annular acquisition module (1) offset on the X and Y axes at time t, calculated by the lidar relative to the current cross-section. Let be the X-coordinate of the Cartesian point cloud of the cross-section at time t. Let be the Y-coordinate of the Cartesian point cloud at time t; S23, Extraction Cross-sectional point cloud slices at the front end of the ring acquisition module (1) within seconds : ； in, The thickness of the slice; , indicating the position of the acquisition unit (11) at time t; It is the set of point cloud coordinates for a slice of cross-sectional point cloud; The preset scanning time is one cycle; v is the moving speed of the acquisition unit (11); It is a partial derivative function; Indicates to Find the partial derivatives; S24. Calculate the distance between the lens (112) and the cave wall according to the following formula. : ； in, , , This represents the coordinates of the point cloud on the cave wall that is furthest from the i-th camera at time t. Let represent the position coordinates of the i-th camera at time t; i = {1, 2, ..., 12}.

4. The acquisition and control system for 360-degree panoramic images of tunnels and culverts according to claim 3, characterized in that, The following formula is used to determine the distance between each lens of the image acquisition device and the cave wall. Adjust the focal length of each camera in the image acquisition device : ； in, This represents the focal length of the i-th camera at second t. , This is the standard focal length of the lens.

5. The acquisition and control system for 360-degree panoramic images of tunnels and culverts according to claim 1, characterized in that, The re-acquisition strategy includes: re-acquiring at the current coordinate point, then performing a comprehensive quality score on the re-acquisitioned image; if the recalculated comprehensive quality score is lower than a preset quality score threshold, a tiered adjustment strategy is executed; otherwise, the system continues to return to the next coordinate point in the abnormal coordinate list. The tiered adjustment strategy includes: first, executing a motion blur adjustment strategy; then, executing an abnormal lighting adjustment strategy; and finally, executing a defocus blur adjustment strategy. The re-acquisition is performed by using an image acquisition device to acquire images of the tunnel and tunnel in batches.

6. The acquisition and control system for 360-degree panoramic images of tunnels and culverts according to claim 5, characterized in that, The motion blur adjustment strategy specifically involves: if the motion blur degree corresponding to the coordinate points in the abnormal location list is... If motion blur occurs, a new image is captured at that coordinate point, and then the motion blur of the image captured by the ring acquisition module (1) is recalculated. The calculation continues until the motion blur of the newly acquired photographic image is reached. Or stop re-collecting when the preset maximum number of retakes is reached; among which, This is the preset motion blur threshold.

7. The acquisition and control system for 360-degree panoramic images of tunnels and culverts according to claim 5, characterized in that, The specific strategy for adjusting abnormal lighting is as follows: if the lighting balance index... If the illumination is abnormal, the supplementary lighting brightness will be adjusted at that coordinate point during the return trip, and image acquisition will be stopped until the image illumination is normal or the number of adjustments exceeds the preset number. The supplementary lighting brightness adjustment is performed according to the following formula: ； in, The light balance index is used as a reference group. It refers to the brightness of supplemental lighting when the lighting is abnormal; As a linear adjustment factor, when the image is overexposed When the image is underexposed , like If the image is underexposed, then the linear adjustment factor is used. Adjusted to: ； like If the image is overexposed, the linear adjustment factor is used. Adjusted to: ； like If the image lighting is normal, no adjustments will be made. in, This is a floating difference; This represents the average brightness of the current image. The average brightness of the reference group images; This is the underexposure adjustment coefficient; This is the overexposure adjustment factor.

8. The acquisition and control system for 360-degree panoramic images of tunnels and culverts according to claim 5, characterized in that, The defocus blur adjustment strategy specifically includes: if the defocus blur degree If out-of-focus blur occurs, the focal length will be adjusted at that coordinate point during the return trip. The focus is adjusted, and then the image is re-acquired. The re-acquired images are then subjected to a comprehensive quality score until the quality requirements are met, at which point re-acquisition stops. The blur level of the reference group photos; The focal length adjustment amount The calculation formula is as follows: ； in, This is the distance-focal length conversion factor; Let be the distance between the i-th camera and the cave wall; This is the optimal imaging distance for the current camera focal length; For focus adjustment compensation item sgn is a sign function; This represents the overall image quality fluctuation error value. This is an indicator function; it returns 1 if the condition is true, and 0 otherwise. The overall quality score of the image at time t. The overall quality score of the image at the previous time step t-1.

9. A method for acquiring and controlling 360-degree panoramic images of tunnels and culverts, based on the acquisition and control system for 360-degree panoramic images of tunnels and culverts as described in any one of claims 1-8, characterized in that, Includes the following steps: S1. Using an image acquisition device, advance along the tunnel axis and acquire photographic images and point clouds of the tunnel's full cross-section using a preset supplementary lighting strategy; the preset supplementary lighting strategy refers to the intensity of the coaxial light source of the camera in the image acquisition device. Less than the intensity of the light source adjacent to the camera ,in: k is the fill light adjustment coefficient; S2. Calculate the distance between each lens of the image acquisition device and the tunnel wall based on the point cloud data of the cross-sectional contour of the tunnel. S3. Dynamically adjust the focal length of each camera in the image acquisition device according to the distance between each lens of the image acquisition device and the cave wall. S4. The image acquisition device is evaluated based on motion blur, defocus blur and illumination balance index to obtain a comprehensive quality score. When the comprehensive quality score is lower than the preset quality score threshold, the center coordinates of the image acquisition device are added to the abnormal coordinate list. When returning, a re-acquisition strategy is executed at each coordinate point in the abnormal coordinate list until the tunnel is left.

Images