A tunnel parametric modeling method and system based on visual recognition

Abstract

The present invention provides a tunnel parameterized modeling method and system based on visual recognition, and relates to the technical field of tunnel modeling. The present invention places a drone in a well-lit environment to collect dynamic sampling images taken by the drone, and performs correlation processing on the dynamic sampling images to generate a dynamic fuzzy index for reflecting the degree of influence of the drone's own vibration on the clarity of the sampling images during shooting. At the same time, when the drone is placed in a tunnel, static sampling images collected by the drone without starting but only starting the camera are collected, and correlation analysis is performed on the static sampling images to generate a static fuzzy index for reflecting the average fuzziness of the static sampling images. Correlation analysis is performed on the static fuzzy index and the dynamic fuzzy index to generate a comprehensive fuzzy evaluation index for reflecting the fuzziness of the three-dimensional modeling image taken by the drone, so as to help staff evaluate the accuracy of three-dimensional modeling.

Description

Technical Field

[0001] The present invention relates to the technical field of tunnel modeling, and in particular to a tunnel parameterized modeling method and system based on visual recognition. Background Art

[0002] Tunnel modeling is a technique used to plan, design and analyze tunnel projects, usually involving the application of computer-aided design software and specific tunnel modeling software. Tunnel modeling often uses visual recognition, which is to use drones to take images of the tunnel and use computer software to analyze and process the images to generate a three-dimensional model.

[0003] During the parametric modeling of the tunnel, the drone captures relevant images inside the tunnel. When the drone itself is working, multiple motors will run at the same time, causing itself to vibrate, thus affecting the clarity of the image taken by the camera. At the same time, due to the influence of internal particles or humidity in the tunnel, the light is refracted and reflected, thus affecting the clarity of the image taken by the camera.

[0004] The above information disclosed in this Background section is only for enhancement of understanding of the background of the present disclosure and therefore it may contain information that does not form the prior art that is already known to one of ordinary skill in the art. Summary of the invention

[0005] The purpose of the present invention is to provide a tunnel parameterized modeling method and system based on visual recognition to solve the problems raised in the above background technology.

[0006] To achieve the above object, the present invention provides the following technical solutions:

[0007] A tunnel parameter modeling method based on visual recognition, the specific steps include:

[0008] S1. In standard state, start the drone and collect 60 seconds of camera dynamic sampling images. The camera acquisition frame rate is 60 Hz per second.

[0009] S2, performing correlation analysis on the dynamic sampling image to generate a dynamic sampling image grayscale, where the dynamic sampling image grayscale is used to reflect the grayscale of each pixel point of the dynamic sampling image;

[0010] S3, performing correlation analysis on the grayscale of the dynamic sampling image to generate the second-order pixel value of the dynamic sampling image, and preprocessing the dynamic sampling image to generate the dynamic sampling image blur index, which is used to reflect the blur degree of a single dynamic sampling image, performing correlation analysis on the dynamic sampling image blur index to generate the camera blur index, which is used to reflect the average blur degree of the dynamic sampling image, and further performing correlation analysis on the dynamic sampling image blur index and the camera blur index to generate the camera variance index;

[0011] S4, performing correlation analysis on the camera variance index and the camera blur index to generate a dynamic blur index DM, where the dynamic blur index DM is used to reflect the influence of the drone's own vibration on the clarity of the sampled image during shooting;

[0012] S5, placing the drone in the tunnel, without starting the drone, only starting the camera to collect images, and collecting 60S of camera static sampling images;

[0013] S6, performing correlation analysis on the static sampling image to generate a static sampling image grayscale, where the static sampling image grayscale is used to reflect the grayscale of the pixel points of each frame image;

[0014] S7, performing correlation analysis on the grayscale of the static sampling image to generate a horizontal gradient index and a vertical gradient index, and generating a comprehensive gradient index, the comprehensive gradient index is used to reflect the average blur degree of the corresponding pixel points of the static sampling image, performing correlation analysis on the comprehensive gradient index, generating a static blur index JM, the static blur index JM is used to reflect the average blur degree of the static sampling image;

[0015] S8. Perform correlation analysis on the static fuzzy index JM and the dynamic fuzzy index DM to generate a comprehensive fuzzy evaluation index ZMP. The comprehensive fuzzy evaluation index ZMP is used to reflect the degree of blur of the 3D modeling image taken by the UAV. Set a fuzzy threshold, compare the comprehensive fuzzy evaluation index ZMP with the fuzzy threshold, and output the credibility of tunnel parametric modeling.

[0016] Furthermore, in S1, the red, green, and blue components of the pixel at the i-th row and j-th column in the u-th frame of the dynamic sampling image are respectively expressed as , , Represents, performs correlation analysis on it, and generates the dynamic sampling image grayscale , the formula based on is:

[0017] ;

[0018] Among them, the dynamic sampling image grayscale Used to reflect the grayscale of the pixel in the i-th row and j-th column in the u-th frame of the dynamic sampling image.

[0019] Furthermore, the grayscale of the dynamic sampling image Correlation analysis is performed to generate the second-order pixel value of the dynamic sampling image, based on the formula:

[0020] ;

[0021] Second-order pixel value of dynamic sampling image It is used to reflect the blur degree of each pixel of the dynamic sampling image, where: and The values ​​of are -1, 0, 1, is the operator kernel for calculating the second-order pixel of the image point;

[0022] Second-order pixel value of dynamic sampling image Perform correlation analysis to generate the fuzzy index of the dynamic sampling image , the formula based on is:

[0023] ;

[0024] Blur index of dynamic sampling image It is used to reflect the blur degree of the u-th frame of the dynamic sampling image, where M is the number of rows of pixels in the dynamic sampling image, and N is the number of columns of pixels in the dynamic sampling image.

[0025] Furthermore, the blur index of the dynamic sampling image Perform correlation analysis to generate camera blur index , the formula based on is:

[0026] ;

[0027] The camera blur index SM is used to reflect the average blur level of all dynamic sampling images;

[0028] Blur index of dynamic sampling image Perform correlation analysis with the camera blur index SM to generate the camera variance index , the formula based on is:

[0029] ;

[0030] Camera Variance Index Used to reflect the discrete degree of blur of all dynamic sampling images.

[0031] Furthermore, the camera variance index SF and the camera blur index SM are subjected to correlation analysis to generate the dynamic blur index DM, and the formula used is: ;

[0032] The dynamic blur index DM is used to reflect the impact of the drone's own vibration on the clarity of the sampled image during shooting.

[0033] Furthermore, the red, green, and blue components of the pixel point in the i-th row and j-th column of the p-th frame static sampling image are respectively , , Represents, and performs correlation analysis to generate the static sampling image grayscale , the formula based on is:

[0034] ;

[0035] in, Used to reflect the grayscale of the pixel at the i-th row and j-th column in the p-th frame of static sampling image.

[0036] Furthermore, the grayscale of the statically sampled image Perform correlation analysis to generate a horizontal gradient index and vertical gradient index , the formula based on is:

[0037] ;

[0038] Horizontal gradient index and vertical gradient index Perform correlation analysis to generate a comprehensive gradient index , the formula based on is:

[0039] ;

[0040] Among them, the comprehensive gradient index It is used to reflect the average blur degree of the corresponding pixel points of the static sampling image, where the values ​​of m and n are -1, 0, and 1. is the operator kernel for calculating the horizontal gradient, is the operator kernel for calculating the vertical gradient;

[0041] Comprehensive Gradient Index Correlation analysis is performed to generate the static fuzzy index JM, based on the formula:

[0042] ;

[0043] The static blur index JM is used to reflect the average blur level of static sampling images.

[0044] Furthermore, the static fuzzy index JM and the dynamic fuzzy index DM are subjected to correlation analysis to generate a comprehensive fuzzy evaluation index ZMP, based on the following formula:

[0045] ;

[0046] The comprehensive fuzzy evaluation index ZMP is used to reflect the degree of blur of the 3D modeling images taken by UAVs.

[0047] Furthermore, the fuzzy threshold is set , The value is 79.6. When , the confidence level of the output tunnel parameter modeling is 10%; when When , the confidence level of the output tunnel parameterized modeling is 60%; when When , the confidence level of the output tunnel parameterized modeling is 95%.

[0048] A tunnel parameterized modeling system based on visual recognition, used for executing a tunnel parameterized modeling method based on visual recognition, comprising:

[0049] The dynamic sampling module is used to collect dynamic sampling images of each frame of the camera when the drone starts working for drawing 3D modeling under standard conditions for 60 seconds;

[0050] A dynamic grayscale analysis module is used to perform correlation analysis on the dynamic sampling image and generate the dynamic sampling image grayscale;

[0051] A dynamic blur analysis module is used to perform correlation analysis on the grayscale of a dynamic sampling image, generate a second-order pixel value of the dynamic sampling image, and preprocess the dynamic sampling image to generate a dynamic sampling image blur index. The dynamic sampling image blur index is used to reflect the blur degree of a single dynamic sampling image. A correlation analysis is performed on the dynamic sampling image blur index to generate a camera blur index. A correlation analysis is further performed on the dynamic sampling image blur index and the camera blur index to generate a camera variance index. A correlation analysis is performed on the camera variance index and the camera blur index to generate a dynamic blur index DM.

[0052] The static acquisition module is used to place the drone in the tunnel. The drone is not started, and only the camera is started for acquisition, and the camera static sampling images of 60S are collected;

[0053] The static grayscale analysis module is used to perform correlation analysis on the static sampling image and generate the static sampling image grayscale;

[0054] The static fuzzy analysis module is used to perform correlation analysis on the grayscale of the static sampling image, generate a horizontal gradient index and a vertical gradient index, and generate a comprehensive gradient index, perform correlation analysis on the comprehensive gradient index, and generate a static fuzzy index JM;

[0055] The comprehensive analysis module is used to perform correlation analysis on the static fuzzy index JM and the dynamic fuzzy index DM to generate a comprehensive fuzzy evaluation index ZMP.

[0056] Compared with the prior art, the present invention has the following beneficial effects:

[0057] The present invention collects dynamic sampling images taken by the drone by placing the drone in a well-lit environment, and performs correlation processing on the dynamic sampling images to generate a dynamic blur index that reflects the degree of influence of the drone's own vibration on the clarity of the sampling images during shooting. At the same time, when the drone is placed in a tunnel, static sampling images are collected without the drone being started but only the camera is started, and correlation analysis is performed on the static sampling images to generate a static blur index that reflects the average blur degree of the static sampling images. Correlation analysis is performed on the static blur index and the dynamic blur index to generate a comprehensive blur evaluation index that reflects the blur degree of the three-dimensional modeling image taken by the drone, so as to help staff evaluate the accuracy of three-dimensional modeling. BRIEF DESCRIPTION OF THE DRAWINGS

[0058] Figure 1 It is a schematic diagram of the overall method flow of the present invention;

[0059] Figure 2 It is a schematic diagram of the system module of the present invention. DETAILED DESCRIPTION

[0060] In order to make the objectives, technical solutions and advantages of the present invention more clearly understood, the present invention is further described in detail below in conjunction with specific embodiments.

[0061] It should be noted that, unless otherwise defined, the technical terms or scientific terms used in the present invention should be understood by people with ordinary skills in the field to which the present invention belongs. The words "first", "second" and similar words used in the present invention do not indicate any order, quantity or importance, but are only used to distinguish different components. "Include" or "comprise" and similar words mean that the elements or objects appearing before the word include the elements or objects listed after the word and their equivalents, without excluding other elements or objects. "Connect" or "connected" and similar words are not limited to physical or mechanical connections, but may include electrical connections, whether direct or indirect. "Up", "down", "left", "right" and the like are only used to indicate relative positional relationships. When the absolute position of the described object changes, the relative positional relationship may also change accordingly.

[0062] Example:

[0063] See also Figure 1 , the present invention provides a technical solution:

[0064] A tunnel parametric modeling method based on visual recognition. Generally, in tunnel parametric modeling, a drone is used to shoot relevant images in the tunnel, and computer software is used to analyze and process the images shot by the drone to generate a three-dimensional model. During the drone shooting process, multiple motors of the drone will run simultaneously when the drone itself is working, causing its own vibration, thereby affecting the clarity of the image shot by the camera. At the same time, due to the influence of internal particles or humidity in the tunnel, light refraction and reflection occur, thereby affecting the clarity of the image shot by the camera. The unclear shooting caused by the drone's own shaking is defined as dynamic blur, and the unclear shooting caused by the tunnel light is defined as static blur. Both dynamic blur and static blur will affect the modeling accuracy. The specific steps include:

[0065] Step 1: Start the drone under standard conditions and collect 60 seconds of camera dynamic sampling images. The camera capture frame rate is 60 Hz per second. The standard condition is 20 degrees Celsius with sufficient light.

[0066] Step 2: perform correlation analysis on the dynamic sampling image to generate a dynamic sampling image grayscale, where the dynamic sampling image grayscale is used to reflect the grayscale of each pixel point of the dynamic sampling image;

[0067] A total of 3600 frames of dynamic sampling images are collected, and the red, green, and blue components of the pixel point in the i-th row and j-th column of the u-th dynamic sampling image are respectively , , Represents, performs correlation analysis on it, and generates the dynamic sampling image grayscale , the formula based on is:

[0068] ;

[0069] Among them, the dynamic sampling image grayscale Used to reflect the grayscale of the pixel in the i-th row and j-th column in the u-th frame of the dynamic sampling image.

[0070] Step 3: perform correlation analysis on the grayscale of the dynamic sampling image to generate the second-order pixel value of the dynamic sampling image, and preprocess it to generate the dynamic sampling image blur index, which is used to reflect the blur degree of a single dynamic sampling image. Perform correlation analysis on the dynamic sampling image blur index to generate the camera blur index, which is used to reflect the average blur degree of the dynamic sampling image. Further perform correlation analysis on the dynamic sampling image blur index and the camera blur index to generate the camera variance index.

[0071] Dynamic sampling image grayscale Correlation analysis is performed to generate the second-order pixel value of the dynamic sampling image, based on the formula:

[0072] ;

[0073] Second-order pixel value of dynamic sampling image It is used to reflect the blur degree of each pixel of the dynamic sampling image, where: and The values ​​of are -1, 0, 1; is the operator kernel for calculating the second-order pixel of the image point, The value is based on:

[0074] ;

[0075] Second-order pixel value of the dynamic sampling image Perform correlation analysis to generate the fuzzy index of the dynamic sampling image , the formula based on is:

[0076] ;

[0077] Blur index of dynamic sampling image It is used to reflect the blur degree of the u-th frame of the dynamic sampling image, where M is the number of rows of pixels in the dynamic sampling image, and N is the number of columns of pixels in the dynamic sampling image.

[0078] Blur index of dynamic sampling image Perform correlation analysis to generate camera blur index , the formula based on is:

[0079] ;

[0080] The camera blur index SM is used to reflect the average blur level of all dynamic sampling images;

[0081] Blur index of dynamic sampling image Perform correlation analysis with the camera blur index SM to generate the camera variance index , the formula based on is:

[0082] ;

[0083] Camera Variance Index Used to reflect the discrete degree of blur of all dynamic sampling images.

[0084] Step 4: Perform correlation analysis on the camera variance index and the camera blur index to generate a dynamic blur index DM. The dynamic blur index DM is used to reflect the influence of the drone's own vibration on the clarity of the sampled image during shooting.

[0085] The camera variance index SF and the camera blur index SM are correlated to generate the dynamic blur index DM, based on the formula:

[0086] ;

[0087] The dynamic blur index DM is used to reflect the impact of the drone's own vibration on the clarity of the sampled image during shooting.

[0088] Step 5: Place the drone in the tunnel without starting the drone, and only start the camera to collect static sampling images for 60 seconds;

[0089] Step 6: Perform correlation analysis on the static sampling image to generate a static sampling image grayscale, where the static sampling image grayscale is used to reflect the grayscale of the pixel points of each frame image;

[0090] The red, green, and blue components of the pixel point in the i-th row and j-th column of the p-th frame static sampling image are respectively expressed as , , Represents, and performs correlation analysis to generate the static sampling image grayscale , the formula based on is:

[0091] ;

[0092] in, Used to reflect the grayscale of the pixel at the i-th row and j-th column in the p-th frame of static sampling image.

[0093] Step 7: Perform correlation analysis on the grayscale of the static sampling image to generate a horizontal gradient index and a vertical gradient index, and generate a comprehensive gradient index. The comprehensive gradient index is used to reflect the average blur degree of the corresponding pixel points of the static sampling image. Perform correlation analysis on the comprehensive gradient index to generate a static blur index JM. The static blur index JM is used to reflect the average blur degree of the static sampling image.

[0094] Static sampling image grayscale Perform correlation analysis to generate a horizontal gradient index and vertical gradient index , the formula based on is:

[0095] ;

[0096] Horizontal gradient index And the vertical gradient index Perform correlation analysis to generate a comprehensive gradient index , the formula based on is:

[0097] ;

[0098] Among them, the comprehensive gradient index It is used to reflect the average blur degree of the corresponding pixel points of the static sampling image, where the values ​​of m and n are -1, 0, and 1. is the operator kernel for calculating the horizontal gradient, is the operator kernel for calculating the vertical gradient, and The formula for determining the value of is:

[0099] ;

[0100] Among them, m and n are used for indexing, m+2 represents the number of rows, and n+2 represents the number of columns;

[0101] Comprehensive Gradient Index Correlation analysis is performed to generate the static fuzzy index JM, based on the formula:

[0102] ;

[0103] The static blur index JM is used to reflect the average blur level of static sampling images.

[0104] Step 8: Perform correlation analysis on the static fuzzy index JM and the dynamic fuzzy index DM to generate a comprehensive fuzzy evaluation index ZMP. The comprehensive fuzzy evaluation index ZMP is used to reflect the degree of blur of the 3D modeling image taken by the UAV. Set a fuzzy threshold, compare the comprehensive fuzzy evaluation index ZMP with the fuzzy threshold, and output the credibility of tunnel parametric modeling.

[0105] The static fuzzy index JM and the dynamic fuzzy index DM are correlated and the comprehensive fuzzy evaluation index ZMP is generated based on the following formula:

[0106] ;

[0107] The comprehensive fuzzy evaluation index ZMP is used to reflect the degree of blur of the 3D modeling images taken by UAVs.

[0108] Set blur threshold , The value is 79.6. When the output tunnel parametric modeling credibility is 10%, it is recommended to abandon the tunnel parametric modeling and re-model and re-evaluate; when When , the output tunnel parametric modeling credibility is 60%, which means that the accuracy of tunnel parametric modeling is average. If the visible environment in the tunnel is poor, tunnel parametric modeling can be used; when When , the output tunnel parametric modeling confidence is 95%, which means that the tunnel parametric modeling accuracy is high. The higher the tunnel parametric modeling confidence, the higher the accuracy of the tunnel parametric modeling.

[0109] Reference Figure 2 The present invention also provides a tunnel parameter modeling system based on visual recognition, which is used to execute a tunnel parameter modeling method based on visual recognition, including:

[0110] The dynamic sampling module is used to collect dynamic sampling images of each frame of the camera when the drone starts working for drawing 3D modeling under standard conditions for 60 seconds;

[0111] A dynamic grayscale analysis module is used to perform correlation analysis on the dynamic sampling image and generate the dynamic sampling image grayscale;

[0112] A dynamic blur analysis module is used to perform correlation analysis on the grayscale of a dynamic sampling image, generate a second-order pixel value of the dynamic sampling image, and preprocess the dynamic sampling image to generate a dynamic sampling image blur index. The dynamic sampling image blur index is used to reflect the blur degree of a single dynamic sampling image. A correlation analysis is performed on the dynamic sampling image blur index to generate a camera blur index. A correlation analysis is further performed on the dynamic sampling image blur index and the camera blur index to generate a camera variance index. A correlation analysis is performed on the camera variance index and the camera blur index to generate a dynamic blur index DM.

[0113] The static acquisition module is used to place the drone in the tunnel. The drone is not started, and only the camera is started for acquisition, and the camera static sampling images of 60S are collected;

[0114] The static grayscale analysis module is used to perform correlation analysis on the static sampling image and generate the static sampling image grayscale;

[0115] The static fuzzy analysis module is used to perform correlation analysis on the grayscale of the static sampling image, generate a horizontal gradient index and a vertical gradient index, and generate a comprehensive gradient index, perform correlation analysis on the comprehensive gradient index, and generate a static fuzzy index JM;

[0116] The comprehensive analysis module is used to perform correlation analysis on the static fuzzy index JM and the dynamic fuzzy index DM to generate a comprehensive fuzzy evaluation index ZMP.

[0117] The above formulas are all dimensionless and numerical calculations. The formula is a formula for the most recent real situation obtained by collecting a large amount of data and performing software simulation. The preset parameters in the formula are set by technicians in this field according to actual conditions.

[0118] The above embodiments may be implemented in whole or in part by software, hardware, firmware or any other combination thereof. When implemented by software, the above embodiments may be implemented in whole or in part in the form of a computer program product. Those skilled in the art may appreciate that the units and algorithm steps of each example described in conjunction with the embodiments disclosed herein may be implemented by electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are performed by hardware or software methods depends on the specific application and design constraints of the technical solution.

[0119] The units described as separate components may or may not be physically separated, and the components shown as units may or may not be physical units, and may be located in one place or distributed on multiple network units. Some or all of the units may be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0120] The above are only specific implementation methods of the present application, but the protection scope of the present application is not limited thereto. Any technician familiar with the technical field can easily think of changes or substitutions within the technical scope disclosed in the present application, which should be covered by the protection scope of the present application.

Claims

1. A tunnel parameter modeling method based on visual recognition, characterized in that: The specific steps include: S1. In standard state, start the drone and collect 60 seconds of camera dynamic sampling images. The camera acquisition frame rate is 60 Hz per second. S2, performing correlation analysis on the dynamic sampling image to generate the dynamic sampling image grayscale; S3, performing correlation analysis on the grayscale of the dynamic sampling image to generate the second-order pixel value of the dynamic sampling image, and preprocessing it to generate the dynamic sampling image blur index, performing correlation analysis on the dynamic sampling image blur index to generate the camera blur index, and further performing correlation analysis on the dynamic sampling image blur index and the camera blur index to generate the camera variance index; S4, performing correlation analysis on the camera variance index and the camera blur index to generate a dynamic blur index DM; S5, placing the drone in the tunnel, without starting the drone, only starting the camera to collect images, and collecting 60S of camera static sampling images; S6, performing correlation analysis on the static sampling image to generate a static sampling image grayscale; S7, performing correlation analysis on the grayscale of the static sampling image to generate a horizontal gradient index and a vertical gradient index, and generating a comprehensive gradient index, performing correlation analysis on the comprehensive gradient index to generate a static fuzzy index JM; S8. Perform correlation analysis on the static fuzzy index JM and the dynamic fuzzy index DM, generate a comprehensive fuzzy evaluation index ZMP, set a fuzzy threshold, compare the comprehensive fuzzy evaluation index ZMP with the fuzzy threshold, and output the tunnel parameterized modeling credibility.

2. A tunnel parameter modeling method based on visual recognition according to claim 1, characterized in that: In S1, the red, green, and blue components of the pixel point in the i-th row and j-th column of the u-th frame dynamic sampling image are respectively expressed as , , Represents, performs correlation analysis on it, and generates the dynamic sampling image grayscale , the formula based on is: ； Among them, the dynamic sampling image grayscale Used to reflect the grayscale of the pixel in the i-th row and j-th column in the u-th frame of the dynamic sampling image.

3. A tunnel parameter modeling method based on visual recognition according to claim 2, characterized in that: Dynamic sampling image grayscale Correlation analysis is performed to generate the second-order pixel value of the dynamic sampling image, based on the formula: ； Second-order pixel value of dynamic sampling image It is used to reflect the blur degree of each pixel of the dynamic sampling image, where: and The values ​​of are -1, 0, 1, is the operator kernel for calculating the second-order pixel of the image point; Second-order pixel value of dynamic sampling image Perform correlation analysis to generate the fuzzy index of the dynamic sampling image , the formula based on is: ； Blur index of dynamic sampling image It is used to reflect the blur degree of the u-th frame of the dynamic sampling image, where M is the number of rows of pixels in the dynamic sampling image, and N is the number of columns of pixels in the dynamic sampling image.

4. A tunnel parameter modeling method based on visual recognition according to claim 3, characterized in that: Blur index of dynamic sampling image Perform correlation analysis to generate camera blur index , the formula based on is: ； The camera blur index SM is used to reflect the average blur level of all dynamic sampling images; Blur index of dynamic sampling image Perform correlation analysis with the camera blur index SM to generate the camera variance index , the formula based on is: ； Camera Variance Index Used to reflect the discrete degree of blur of all dynamic sampling images.

5. A tunnel parameter modeling method based on visual recognition according to claim 4, characterized in that: The camera variance index SF and the camera blur index SM are correlated to generate the dynamic blur index DM, based on the formula: ； The dynamic blur index DM is used to reflect the impact of the drone's own vibration on the clarity of the sampled image during shooting.

6. A tunnel parameter modeling method based on visual recognition according to claim 1, characterized in that: The red, green, and blue components of the pixel point in the i-th row and j-th column of the p-th frame static sampling image are respectively expressed as , , Represents, and performs correlation analysis to generate the static sampling image grayscale , the formula based on is: ； in, Used to reflect the grayscale of the pixel at the i-th row and j-th column in the p-th frame of static sampling image.

7. A tunnel parameter modeling method based on visual recognition according to claim 6, characterized in that: Static sampling image grayscale Perform correlation analysis to generate a horizontal gradient index and vertical gradient index , the formula based on is: ； Horizontal gradient index And the vertical gradient index Perform correlation analysis to generate a comprehensive gradient index , the formula based on is: ； Among them, the comprehensive gradient index It is used to reflect the average blur degree of the corresponding pixel points of the static sampling image, where the values ​​of m and n are -1, 0, and 1. is the operator kernel for calculating the horizontal gradient, is the operator kernel for calculating the vertical gradient; Comprehensive Gradient Index Correlation analysis is performed to generate the static fuzzy index JM, based on the formula: ； The static blur index JM is used to reflect the average blur degree of the static sampling image, where M is the number of rows of pixels in the static sampling image, and N is the number of columns of pixels in the static sampling image.

8. A tunnel parameter modeling method based on visual recognition according to claim 1, characterized in that: The static fuzzy index JM and the dynamic fuzzy index DM are correlated and the comprehensive fuzzy evaluation index ZMP is generated based on the following formula: ； The comprehensive fuzzy evaluation index ZMP is used to reflect the degree of blur of the 3D modeling images taken by UAVs.

9. A tunnel parameter modeling method based on visual recognition according to claim 8, characterized in that: Set blur threshold , The value is 79.

6. When , the confidence level of the output tunnel parameterized modeling is 10%; when When , the confidence level of the output tunnel parameterized modeling is 60%; when When , the confidence level of the output tunnel parameterized modeling is 95%.

10. A tunnel parametric modeling system based on visual recognition, used to execute a tunnel parametric modeling method based on visual recognition according to claim 1, characterized in that: include: The dynamic sampling module starts the drone in the standard state and collects 60 seconds of dynamic sampling images from the camera. The camera’s acquisition frame rate is 60 Hz per second. A dynamic grayscale analysis module is used to perform correlation analysis on the dynamic sampling image and generate the dynamic sampling image grayscale; A dynamic blur analysis module is used to perform correlation analysis on the grayscale of a dynamic sampling image, generate a second-order pixel value of the dynamic sampling image, and preprocess the dynamic sampling image to generate a dynamic sampling image blur index. The dynamic sampling image blur index is used to reflect the blur degree of a single dynamic sampling image. A correlation analysis is performed on the dynamic sampling image blur index to generate a camera blur index. A correlation analysis is further performed on the dynamic sampling image blur index and the camera blur index to generate a camera variance index. A correlation analysis is performed on the camera variance index and the camera blur index to generate a dynamic blur index DM. The static acquisition module is used to place the drone in the tunnel. The drone is not started, and only the camera is started for acquisition, and the camera static sampling images of 60S are collected; The static grayscale analysis module is used to perform correlation analysis on the static sampling image and generate the static sampling image grayscale; The static fuzzy analysis module is used to perform correlation analysis on the grayscale of the static sampling image, generate a horizontal gradient index and a vertical gradient index, and generate a comprehensive gradient index, perform correlation analysis on the comprehensive gradient index, and generate a static fuzzy index JM; The comprehensive analysis module is used to perform correlation analysis on the static fuzzy index JM and the dynamic fuzzy index DM to generate a comprehensive fuzzy evaluation index ZMP.

Images