CN-122018529-A - Bridge detection method and system based on unmanned aerial vehicle

CN122018529ACN 122018529 ACN122018529 ACN 122018529ACN-122018529-A

Abstract

A bridge detection method and system based on unmanned aerial vehicle belongs to the technical field of bridge detection. The method aims to solve the problems of improving the quality and the identification accuracy of bridge detection images. The method comprises the steps of constructing an image definition scoring function of a single frame image acquired by an unmanned aerial vehicle, constructing a speed-gesture fuzzy coupling risk scoring function of the single frame image acquired by the unmanned aerial vehicle, constructing an image transmission time domain continuity scoring function of continuous images acquired by the unmanned aerial vehicle, constructing an image comprehensive value scoring function, constructing an image main direction deviation vector, combining the image main direction deviation vector and the image comprehensive value scoring function, constructing an image identification suitability weight factor, and designing a dynamic optimization strategy with the recommended flight speed, the recommended gesture angular speed and the recommended flight altitude of the unmanned aerial vehicle based on the image main direction deviation vector and the identification suitability weight factor. The method and the device improve the quality and the identification accuracy of the bridge detection image to a great extent.

Inventors

LIU XING
CHENG GONG
ZHUANG WEIQUN
MENG ANXIN
Ren Bangke
ZHAO HAIYUN

Assignees

深城交科技集团股份有限公司

Dates

Publication Date: 20260512
Application Date: 20260414

Claims (9)

1. The bridge detection method based on the unmanned aerial vehicle is characterized by comprising the following steps of: S1, constructing an image definition scoring function of a single frame image acquired by an unmanned aerial vehicle; s2, constructing a speed-posture fuzzy coupling risk scoring function of the unmanned aerial vehicle for acquiring a single frame image; S3, constructing an image transmission time domain continuity scoring function of continuous images acquired by the unmanned aerial vehicle; s4, constructing an image comprehensive value scoring function based on the definition scoring function of the image obtained in the step S1, the speed-gesture fuzzy coupling risk scoring function obtained in the step S2 and the image graph transmission time domain continuity scoring function obtained in the step S3; s5, constructing an image main direction deviation vector; s6, combining the image main direction deviation vector and the image comprehensive value scoring function to construct an image identification suitability weight factor; And S7, designing a dynamic optimization strategy with the recommended flight speed, the recommended attitude angular speed and the recommended flight altitude of the unmanned aerial vehicle based on the image main direction deviation vector obtained in the step S5 and the identification suitability weight factor obtained in the step S6.
2. The bridge detection method based on the unmanned aerial vehicle according to claim 1, wherein in step S1, the image definition scoring function is obtained by fusing high-frequency energy proportion and airspace edge geometric information in the image frequency domain signal, and the expression is: ; Wherein, the Scoring the image sharpness; As a Fourier weight coefficient, determining by expert experience; The residual error inverse value weight coefficient is determined by expert experience; Is a critical threshold value, and is determined empirically by an expert; representing a high frequency range, u being the two-dimensional frequency coordinates of the image in the frequency domain, an image frame acquired at time t of I t , Fourier transform of I t ; the inverse of the residual was linearly fit to the local edge of I t , calculated from the Sobel gradient.
3. The bridge detection method based on the unmanned aerial vehicle according to claim 2, wherein the speed-gesture fuzzy coupling risk scoring function of the unmanned aerial vehicle acquired single frame images constructed in the step S2 evaluates the influence of the translational speed and yaw gesture change of the unmanned aerial vehicle on imaging, and the expression is as follows: ; Wherein, the Speed-attitude fuzzy coupling risk value; The speed fuzzy sensitivity weight coefficient is determined by expert experience; the fuzzy sensitivity weight coefficient of the angular velocity is determined by expert experience; the speed component is an XY plane speed component and is obtained by an unmanned aerial vehicle monitoring platform; The angular speed of the unmanned aerial vehicle around the Z axis is obtained by an unmanned aerial vehicle monitoring platform; is a speed reference value, and is determined empirically by an expert; Is an angular velocity reference value, and is determined empirically by an expert.
4. The unmanned aerial vehicle-based bridge detection method of claim 3, wherein in step S3, the structural similarity between two frames of images and the inter-frame time interval change amplitude are combined to construct an image graph time domain continuity scoring function, and the expression is: ; Wherein, the Scoring temporal continuity of the image map; the structural similarity of the SSIM is obtained by calling an SSIM calculation module; obtaining the time interval between the current frame image and the previous frame image by camera parameters; 、 The average value and standard deviation of the Slide window are counted by a local cache; is a fixed constant, dimension and And keeping the same, and determining the same by expert experience.
5. The unmanned aerial vehicle-based bridge detection method of claim 4, wherein the image integrated value scoring function constructed in the step S4 has the expression: ; Wherein, the Scoring the composite value of the image.
6. The unmanned aerial vehicle-based bridge inspection method of claim 5, wherein in step S5, in order to evaluate the matching degree of the current camera angle and the structural layout, an image main direction deviation vector is constructed, and the expression is: ; Wherein, the Is an image principal direction deviation vector; the method comprises the steps that an orientation unit vector of a camera lens in panorama is obtained by means of an unmanned aerial vehicle monitoring platform; The image straight line pointing direction is average and is obtained by an unmanned aerial vehicle monitoring platform.
7. The bridge detection method based on the unmanned aerial vehicle according to claim 6, wherein step S6 takes the square of the principal direction deviation vector of the image as a value penalty term, introduces an exponential function to perform nonlinear scaling, combines with the comprehensive value score of the image, calculates the identification suitability weight factor of the image, and has the expression: ; Wherein, the Identifying suitability weighting factors for the image; As the main direction deviation weight factor, it is determined by expert experience.
8. The bridge inspection method based on the unmanned aerial vehicle according to claim 7, wherein the specific implementation method of the step S7 comprises the following steps: S7.1, designing a suggested flight speed of the unmanned aerial vehicle by using a linear response model, wherein the expression is as follows: ; Wherein, the Suggested flight speed for the unmanned aerial vehicle; the current speed of the unmanned aerial vehicle is obtained by an unmanned aerial vehicle monitoring platform; A directional gradient that identifies suitability weighting factors for the image; is a directional gradient weight factor, and is determined by expert experience; s7.2, designing a suggested attitude angular speed of the unmanned aerial vehicle based on the image main direction deviation vector by using a proportional control method, wherein the expression is as follows: ; Wherein, the Suggested attitude angular velocity for the unmanned aerial vehicle; Is an angular velocity weight factor, and is determined by expert experience; S7.3, the overall recognition condition is reflected by using the recognition suitability weight factor of the image, and the overall recognition condition is converted into a proportional adjustment coefficient for designing the recommended flight height of the unmanned aerial vehicle, wherein the expression is as follows: ; Wherein, the Suggested fly height for the unmanned aerial vehicle; the current flight height of the unmanned aerial vehicle is obtained by an unmanned aerial vehicle monitoring platform; Is a flight altitude adjustment factor.
9. A system based on a bridge inspection method of an unmanned aerial vehicle, comprising a processor, a memory and a computer program stored in the memory and executable on the processor, the computer program when run implementing the steps of a bridge inspection method of an unmanned aerial vehicle according to any one of claims 1 to 8.

Description

Bridge detection method and system based on unmanned aerial vehicle Technical Field The invention belongs to the technical field of bridge detection, and particularly relates to a bridge detection method and system based on an unmanned aerial vehicle. Background Along with the continuous increase of the number of bridges on highways and railways, regular detection and state evaluation become the guarantee of the safety operation and maintenance of the infrastructure, traditional bridge detection depends on manual close-range inspection and overhead operation, the safety risk is relatively large, the efficiency is very low, the coverage range is limited, and recently, unmanned aerial vehicles gradually become an emerging tool for bridge detection by virtue of the advantages of flexibility, high efficiency, low cost and the like. The device can perform close-range shooting and video detection under river crossing, high pier and high altitude environments, and can greatly improve the operation efficiency. Although unmanned aerial vehicle presents higher application potential in many scenes, can it still face a lot of challenges when using in actual bridge detection task, and the development is, has the shake in the flight in-process and causes the problem such as image blurring, the view angle deviation causes to shoot out of focus, the discontinuous structure recognition result that produces of image frame sequence, does not have image quality's feedback when unmanned aerial vehicle flies moreover, does not have the image quality's that accomplish in the detection process and shoots again to the poor quality's image. The flight control adjusting mechanism aiming at the image recognition target is required to be constructed, the image quality and recognition value are dynamically integrated with the flight control, the unmanned aerial vehicle is led to adjust the speed, the height and the gesture, the bridge image which is more stable, clear and more beneficial to recognition is obtained, and the support in the aspect of foundation is provided for bridge disease detection and evaluation. Disclosure of Invention The invention aims to solve the problems of improving the quality and the recognition accuracy of bridge detection images and provides a bridge detection method and system based on an unmanned aerial vehicle. In order to achieve the above purpose, the present invention is realized by the following technical scheme: a bridge detection method based on unmanned aerial vehicle comprises the following steps: S1, constructing an image definition scoring function of a single frame image acquired by an unmanned aerial vehicle; s2, constructing a speed-posture fuzzy coupling risk scoring function of the unmanned aerial vehicle for acquiring a single frame image; S3, constructing an image transmission time domain continuity scoring function of continuous images acquired by the unmanned aerial vehicle; s4, constructing an image comprehensive value scoring function based on the definition scoring function of the image obtained in the step S1, the speed-gesture fuzzy coupling risk scoring function obtained in the step S2 and the image graph transmission time domain continuity scoring function obtained in the step S3; s5, constructing an image main direction deviation vector; s6, combining the image main direction deviation vector and the image comprehensive value scoring function to construct an image identification suitability weight factor; And S7, designing a dynamic optimization strategy with the recommended flight speed, the recommended attitude angular speed and the recommended flight altitude of the unmanned aerial vehicle based on the image main direction deviation vector obtained in the step S5 and the identification suitability weight factor obtained in the step S6. Further, in step S1, by fusing the high-frequency energy proportion and the spatial domain edge geometric information in the image frequency domain signal, an image definition scoring function is obtained, where the expression is: Wherein, the Scoring the image sharpness; As a Fourier weight coefficient, determining by expert experience; The residual error inverse value weight coefficient is determined by expert experience; Is a critical threshold value, and is determined empirically by an expert; representing a high frequency range, u being the two-dimensional frequency coordinates of the image in the frequency domain, an image frame acquired at time t of I t, Fourier transform of I t; the inverse of the residual was linearly fit to the local edge of I t, calculated from the Sobel gradient. Further, the speed-gesture fuzzy coupling risk scoring function of the unmanned aerial vehicle for acquiring the single frame image constructed in the step S2 evaluates the influence of the translational speed and yaw gesture change of the unmanned aerial vehicle on imaging, and the expression is as follows: Wherein, the Speed-attitude fuzzy coupling r