A method for measuring the width of expansion joints

Abstract

The present invention discloses a method for measuring the width of an expansion joint. The method calibrates a photographing device to obtain calibration parameters, obtains a first image to be measured by the photographing device, calibrates the first image to be measured based on the calibration parameters to obtain a second image to be measured, identifies the edge of the expansion joint in the second image to be measured, and calculates the width of the expansion joint. The method solves the problems of insufficient detection accuracy and overly complicated operation of the monocular expansion joint width detection method.

Description

Technical Field

[0001] The invention relates to the field of expansion joint measurement, and in particular to a method for measuring the width of an expansion joint. Background Art

[0002] Currently, most of the expansion joint width measurement methods on the market are applied to multi-eye detection systems, which are relatively expensive, and there is a lack of expansion joint width measurement methods suitable for monocular detection systems.

[0003] Some monocular expansion joint width measurement methods have problems such as insufficient measurement accuracy and overly complicated operation methods, which can easily lead to a waste of human resources. Summary of the invention

[0004] The embodiment of the present application provides a method for measuring the width of an expansion joint to solve the problem of insufficient accuracy of current monocular expansion joint measurements.

[0005] According to one aspect of an embodiment of the present application, a method for measuring an expansion joint width is provided, comprising:

[0006] Calibrate the camera to obtain calibration parameters;

[0007] Acquire a first image to be tested by a photographing device;

[0008] Calibrate the first image to be tested based on the calibration parameters to obtain a second image to be tested;

[0009] An edge of the expansion joint in the second image to be measured is identified, and a width of the expansion joint is calculated.

[0010] Preferably, the step of identifying the expansion joint edge in the second image to be measured and calculating the width of the expansion joint includes:

[0011] Preprocessing the second image to be tested to obtain a third image to be tested;

[0012] Performing edge detection processing on the third image to be tested to obtain expansion joint edge data;

[0013] The width of the expansion joint is calculated based on the expansion joint edge data and the calibration parameters.

[0014] Preferably, the preprocessing of the second image to be tested to obtain a third image to be tested includes:

[0015] Binarizing and denoising the second image to be tested, and contrast enhancing the feature part in the image to obtain a third image to be tested;

[0016] The characteristic portion is a portion of the image that includes the edge of the expansion joint.

[0017] Preferably, performing edge detection processing on the third image to be tested to obtain expansion joint edge data includes:

[0018] Identify the expansion joint edge in the third image to be tested based on an edge detection algorithm;

[0019] The expansion joint edge is optimized based on an edge refinement algorithm to obtain the expansion joint edge data.

[0020] Preferably, the identifying the expansion joint edge in the third image to be tested based on an edge detection algorithm includes:

[0021] Identifying the expansion joint edge in the third image to be tested;

[0022] Determine whether the expansion joint edge meets the preset accuracy index. If not, adjust the parameters of the edge detection algorithm and return to the previous step to record the number of returns C. If the number of returns C meets the preset accuracy index, When , the recognition result is output, where C and is a positive integer;

[0023] If yes, the recognition result is output.

[0024] Preferably, the determining whether the expansion joint edge meets a preset accuracy index includes:

[0025] Calculating the ratio of the length of the expansion joint edge to the expected expansion joint edge length;

[0026] It is determined whether the ratio is greater than or equal to a preset ratio.

[0027] Preferably, the determining whether the expansion joint edge meets a preset accuracy index further includes:

[0028] Calculating the maximum offset between the position of the expansion joint edge and the expected expansion joint edge position;

[0029] It is determined whether the maximum offset is less than or equal to a preset maximum offset threshold.

[0030] Preferably, adjusting the parameters of the edge detection algorithm includes:

[0031] According to the formula and / or ,Adjustment and / or The value of the adaptive high threshold and / or adaptive low threshold Make adjustments;

[0032] in, is the high threshold ratio coefficient, M is the median value of the image gradient amplitude, is the low threshold scale factor.

[0033] Preferably, the step of optimizing the expansion joint edge based on an edge refinement algorithm to obtain the expansion joint edge data includes:

[0034] Sampling intensity values ​​of one or more edge pixels in the expansion joint edge along a direction perpendicular to the image edge to create an intensity profile;

[0035] According to the formula , the intensity profile is fitted, where For location The pixel intensity at is the horizontal coordinate of the edge position of the sub-pixel in the fitting coordinate system, for The horizontal coordinate of the pixel at in the fitting coordinate system, a and b are preset fitting parameters;

[0036] The fitted intensity profile data is output to obtain the expansion joint edge data.

[0037] Preferably, calculating the width of the expansion joint based on the expansion joint edge data and the calibration parameter comprises:

[0038] Calculating a pixel-to-width conversion scaling factor based on the calibration parameters;

[0039] The distance between the expansion joint edges is calculated according to the expansion joint edge data and the pixel-width conversion ratio factor to obtain the width of the expansion joint.

[0040] Preferably, the step of calibrating the photographing device to obtain calibration parameters includes:

[0041] Set the calibration pattern;

[0042] A photographing device photographs the calibration pattern to obtain a calibration image set, wherein the calibration image set includes one or more images containing the calibration pattern;

[0043] The calibration image set is analyzed and processed to obtain distortion calibration parameters and perspective correction parameters.

[0044] Preferably, the analyzing and processing the calibration image to obtain the distortion calibration parameters and the perspective correction parameters further includes:

[0045] Determine whether the photographing device is perpendicular to the plane of the calibration pattern, and if so, the perspective correction parameters include external parameters, and the external parameters include a rotation matrix and a translation vector;

[0046] If not, the perspective correction parameters include external parameters and a perspective transformation matrix, and the external parameters include a rotation matrix and a translation vector.

[0047] Preferably, the step of identifying the expansion joint edge in the second image to be measured and calculating the width of the expansion joint further includes:

[0048] The second image to be tested also includes multiple frames of continuous sub-images;

[0049] The edge position prediction value of the sub-image in the second image to be tested is predicted and calculated by the following formula:

[0050]

[0051] in, is the edge position value of the sub-image of the t+1th frame, is the edge position prediction value of the sub-image of the t+1th frame, is the actual detection value of the edge position obtained by edge detection of the sub-image of the t+1th frame, K is the gain coefficient, t≥1, and t is a natural number;

[0052] The width of the expansion joint of each frame sub-image is calculated according to the edge position value of each frame sub-image.

[0053] The present invention calibrates the camera to obtain calibration parameters, obtains a first image to be tested through the camera, calibrates the first image to be tested based on the calibration parameters to obtain a second image to be tested, identifies the expansion joint edge in the second image to be tested, and calculates the width of the expansion joint. The problem of insufficient detection accuracy and overly complicated operation of the monocular expansion joint width detection method is solved. BRIEF DESCRIPTION OF THE DRAWINGS

[0054] Figure 1 This is a flow chart of an embodiment of a method for measuring the width of an expansion joint provided in the present application. DETAILED DESCRIPTION

[0055] In order to enable those skilled in the art to better understand the solution of the present application, the technical solution in the embodiments of the present application will be clearly and completely described below in conjunction with the drawings in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, not all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by ordinary technicians in this field without creative work should fall within the scope of protection of the present application.

[0056] It should be noted that the terms "first", "second", etc. in the specification and claims of the present application and the above-mentioned drawings are used to distinguish similar objects, and are not necessarily used to describe a specific order or sequence. It should be understood that the data used in this way can be interchangeable where appropriate, so that the embodiments of the present application described herein can be implemented in an order other than those illustrated or described herein. In addition, the terms "including" and "having" and any of their variations are intended to cover non-exclusive inclusions, for example, a process, method, system, product or device comprising a series of steps or units is not necessarily limited to those steps or units clearly listed, but may include other steps or units that are not clearly listed or inherent to these processes, methods, products or devices.

[0057] The present application is described in detail below with reference to the accompanying drawings and embodiments. The present invention provides a method for measuring the width of an expansion joint, such as Figure 1 As shown, the method includes steps S102-S105, wherein:

[0058] S102: calibrating the photographing device to obtain calibration parameters;

[0059] In the embodiment provided in the present application, a calibration pattern is set, and a camera captures the calibration pattern to obtain a calibration image set, wherein the calibration image set includes one or more images containing the calibration pattern, and the calibration image set is analyzed and processed to obtain distortion calibration parameters and perspective correction parameters. Specifically, a calibration pattern is placed in the field of view of the camera, and the camera captures one or more images containing the calibration pattern from different angles and positions to obtain a calibration image set. The feature points of the calibration pattern are detected using an image processing algorithm to obtain the camera coordinate system coordinates of the feature points. , according to the formula Calculate the intrinsic parameter matrix K of the camera, where u is the horizontal coordinate of the pixel coordinate system, and v is the vertical coordinate of the pixel coordinate system. and is the coordinate of the projection position of the optical axis of the camera in the pixel coordinate system, , is the focal length of the camera, dx is the physical width of a pixel, and dy is the physical height of a pixel. Convert to image coordinate system , and normalize the image coordinate system coordinates to obtain , according to the formula ,in, and is the image coordinate after distortion calibration, r is the radial distance to the feature point, , ,

[0060] , , is the radial distortion coefficient, and according to the formula , calculate the reprojection error under the current radial distortion coefficient, and use the nonlinear least squares optimization method to optimize the radial distortion coefficient, reduce the reprojection error, and obtain the optimal value of the radial distortion coefficient, where N is the number of images of the calibration pattern collected, and M is the number of feature points in each image. is the detected two-dimensional feature point, is the projected two-dimensional feature point calculated based on the current radial distortion coefficient.

[0061] It should be noted that the calibration pattern used in this embodiment may be a checkerboard, a dot matrix or other calibration patterns, which are not restricted here.

[0062] In this embodiment, based on the position of the camera relative to the plane of the calibration pattern, the calculation of the perspective calibration parameters is divided into two cases, one case is that the shooting device is perpendicular to the plane of the calibration pattern, and the other case is that the shooting device is not perpendicular to the plane of the calibration pattern.

[0063] Specifically, when the camera is perpendicular to the plane of the calibration pattern, the perspective calibration parameters are directly calculated. Set a world coordinate system with the expansion joint feature point as the origin, the X axis of the world coordinate system is the length direction of the expansion joint, the Y axis is perpendicular to the expansion joint, and the Z axis is perpendicular to the plane where the expansion joint is located, and obtain the world coordinate system coordinates of the calibration pattern feature point , according to the formula, Calculate the extrinsic matrix , where u is the horizontal coordinate of the pixel coordinate system, v is the vertical coordinate of the pixel coordinate system, K is the intrinsic parameter matrix, R is the rotation matrix, and t is the translation vector.

[0064] When the camera is not perpendicular to the plane of the calibration pattern, the formula , calculate the perspective transformation matrix H, where , is the pixel coordinate system coordinate, , is the world coordinate system coordinate, and the perspective transformation matrix H is used to eliminate the influence of the shooting device not being perpendicular to the plane of the calibration pattern.

[0065] S103: Acquire a first image to be tested by a photographing device;

[0066] In the embodiment provided in the present application, in the expansion joint width measurement scenario applied by the present solution, the basis for expansion joint identification is the image captured by the shooting device.

[0067] Specifically, the first image to be tested is collected by a photographing device, and the first image to be tested may include one or more expansion joint images taken from different angles and directions of the same expansion joint to be tested, or one or more expansion joint images taken from the same angle and different directions, or one or more expansion joint images taken from the same angle and the same direction.

[0068] It should be noted that, in the embodiment provided in the application, the photographing device is a monocular camera.

[0069] S104: calibrating the first image to be tested based on calibration parameters to obtain a second image to be tested;

[0070] In the embodiments provided in the present application, the image in the first image to be measured is calibrated based on calibration parameters such as the intrinsic parameters of the shooting device, distortion coefficients, rotation matrices, and translation vectors to obtain a second image to be measured, and the common distortions produced by the shooting device are eliminated, so that the image can more accurately reflect the real scene and eliminate perspective distortion. The coordinate system can accurately align the physical structure, thereby achieving accurate calculation of the actual measurement unit.

[0071] S105: identifying an edge of the expansion joint in the second image to be measured, and calculating a width of the expansion joint;

[0072] In the embodiment provided in the present application, the second image to be tested is preprocessed to obtain a third image to be tested, the third image to be tested is edge detected to obtain expansion joint edge data, and the width of the expansion joint is calculated based on the expansion joint edge data and calibration parameters. Specifically, the second image to be tested is binarized and converted into a grayscale image, and the grayscale image is denoised using a noise reduction algorithm to eliminate noise, and an adaptive histogram equalization method is used to enhance the contrast of the feature part in the grayscale image. In this embodiment, the feature part refers to the part of the image containing the expansion joint edge, so as to improve the visibility of the edge and the feature.

[0073] It should be noted that the noise reduction algorithms used in this embodiment include Gaussian blur, non-local mean denoising, bilateral filtering, median filtering, chromatic aberration correction and image stabilization algorithm, which are not restricted here.

[0074] In this embodiment, the Canny edge detection algorithm is used to identify the edge contour of the expansion joint in the third image to be tested, and it is determined whether the edge contour of the expansion joint meets the preset accuracy index. If not, the parameters of the edge detection algorithm are adjusted and the previous step is returned to record the number of returns C. If the number of returns C meets the preset accuracy index, the expansion joint edge contour is judged to be correct. When , output the recognition result, where C is a positive integer. If , output the recognition result.

[0075] Specifically, in this embodiment, judging whether the edge profile of the expansion joint meets the preset accuracy index specifically includes two judgments, one is whether the ratio of the length of the expansion joint edge to the expected expansion joint edge length is less than or equal to the preset ratio, and the other is whether the maximum offset between the position of the expansion joint edge and the expected expansion joint edge position is less than or equal to the preset maximum offset threshold.

[0076] Specifically, a preset model is used to confirm whether the expansion joint edge contour detected in the image is continuous, and the length of the expansion joint edge is calculated to confirm the ratio of the length of the expansion joint edge to the expected expansion joint edge length, and to determine whether the above ratio is greater than or equal to the preset ratio. If so, the expansion joint edge detection meets the preset ratio.

[0077] Specifically, the maximum offset refers to the maximum deviation between the edge contour of the expansion joint identified in the third image to be tested and the expected position.

[0078] Specifically, the expansion joint edge points in the image are converted to the Hough space, and the peak values ​​are found in the Hough accumulator matrix to identify the straight line, so as to extract the expansion joint edge line and obtain the parameters of the expansion joint edge line. ,in, is the distance from the origin to the line in Hough space, is the angle of the normal vector. The number of detected expansion joint edges is determined by the number of peaks. It is confirmed whether the number of detected edges is consistent with the preset value. , calculate the tolerance of the angle of the detected expansion joint edge line ,when ≤ the preset maximum offset threshold, the expansion joint edge detection meets the preset maximum offset threshold.

[0079] It should be noted that the edge contour of the expansion joint is considered to meet the preset standard indicators if and only if the edge detection meets the two preset standard indicators that the ratio of the length of the expansion joint edge to the expected expansion joint edge length is less than or equal to the preset ratio and the maximum offset of the position of the expansion joint edge and the expected expansion joint edge position is less than or equal to the preset maximum offset threshold.

[0080] Specifically, in this embodiment, an adaptive threshold is set for the Canny edge detection algorithm, including an adaptive high threshold and an adaptive low threshold. When the edge contour of the expansion joint does not meet the preset standard indicators, the parameters of the edge algorithm can be automatically adjusted to improve the accuracy of edge detection and reduce the complexity of operation.

[0081] Specifically, according to the formula Calculate the gradient magnitude G of the image, where is the gradient of the image in the x direction, is the gradient of the image in the y direction, and the median M of the gradient magnitude G is obtained, according to the formula and / or ,Adjustment and / or The value of the adaptive high threshold and / or adaptive low threshold Make adjustments, including is the high threshold ratio coefficient, M is the median value of the image gradient amplitude, is the low threshold scale factor.

[0082] This application sets an adaptive threshold so that when the detection results of the Canny edge detection algorithm do not meet the preset requirements, it can automatically adjust the parameters and optimize the detection results. It also limits the maximum number of parameter adjustments to avoid excessive task processing caused by too many parameter adjustments and prevent system crashes.

[0083] In this embodiment, after the expansion joint edge profile is determined, the expansion joint edge profile is optimized based on an edge refinement algorithm to obtain the expansion joint edge data.

[0084] Specifically, one or more edge pixels in the expansion joint edge profile are sampled along the direction perpendicular to the image edge to create an intensity profile. , the intensity profile is fitted, where For location The pixel intensity at is the horizontal coordinate of the edge position of the sub-pixel in the fitting coordinate system, for The horizontal coordinate of the pixel at in the fitting coordinate system, a and b are preset fitting parameters, and the intensity profile data after fitting is output to obtain the expansion joint edge data.

[0085] The initial expansion joint edge profile is optimized through the edge refinement algorithm, which improves the accuracy of edge positioning and reaches the sub-pixel level, greatly improving the accuracy of expansion joint width measurement.

[0086] In this embodiment, after the expansion joint edge data is obtained, the pixel-width conversion scale factor is calculated according to the calibration parameters, and the distance between the expansion joint edges is calculated according to the expansion joint edge data and the pixel-width conversion scale factor to obtain the width of the expansion joint.

[0087] Specifically, according to the formula , calculate the pixel-width conversion scale factor s, where K is the intrinsic matrix, R is the rotation matrix, t is the translation vector, X, Y, Z are the coordinates of the world coordinate system, and u, v are the coordinates of the pixel coordinate system. Calculate the vertical distance between one or more pixels on the edge of the expansion joint and the pixels on the edge of another expansion joint , and based on the pixel-width conversion scale factor s, the resulting vertical distance Convert to width to get the width of the expansion joint.

[0088] In another embodiment provided in the present application, the first image to be tested photographed by the photographing device may include multiple frames of continuous images. After the first image to be tested is calibrated based on the calibration parameters, the second image to be tested also includes multiple frames of continuous images.

[0089] When the second image to be measured includes multiple frames of continuous images, the position of the expansion joint edge in a frame of image can be estimated by the position of the expansion joint edge in a subsequent frame of image.

[0090] Specifically, according to the formula Calculate the edge position prediction value of the sub-image in the second test image ,in, is the frame interval, is the edge position value of the sub-image of the tth frame, is the edge position change rate of the sub-image in the t-th frame, t≥1, and t is a natural number.

[0091] Among them, the edge position change rate of the sub-image of the tth frame is is the speed of edge position change. In this embodiment, the speed of edge position change is defined as the rate of change of the moving distance of the edge position in two adjacent frame sub-images in the same coordinate system relative to the frame interval time.

[0092] According to the formula Calculate the edge position value of the sub-image in, is the actual detected position of the expansion joint edge in the first image, and K is the gain coefficient.

[0093] Among them, the gain coefficient K is adjusted by updating the error covariance matrix based on the measured value of the noise and the predicted value of the noise and according to the formula Iterate, where is the covariance matrix of the t+1 frame sub-image, is the transformation of the observation matrix, is the noise covariance matrix of the t+1 frame sub-image.

[0094] It should be noted that, in this embodiment, the edge position value of the sub-image It is the edge position result obtained after detection by the edge detection algorithm.

[0095] By predicting the edge position of the expansion joint and taking the influence of noise into account when updating the measurement, the edge position fluctuation caused by noise can be effectively filtered out, so that the edge detection remains stable in the presence of noise, thereby improving the detection accuracy. At the same time, when calculating the edge position of the expansion joint of each frame sub-image, the information of the previous frame is incorporated into the prediction of the current frame, ensuring that the change of the edge position is smooth in time and avoiding sudden jumps.

[0096] By predicting the position of the expansion joint edge in the image, the change of the expansion joint edge position can be stabilized in the presence of noise or small jitter in the system, thereby improving the accuracy and stability of the measurement.

[0097] When the image to be tested is a plurality of continuous frames, by predicting the position of the edge of the expansion joint in the image to be tested, the need for complete recognition of each frame of the image is reduced, thereby reducing the amount of calculation and improving processing efficiency.

Claims

1. A method for measuring the width of an expansion joint, characterized in that: The method comprises: Calibrate the camera to obtain calibration parameters; Acquire a first image to be tested by a photographing device; Calibrate the first image to be tested based on the calibration parameters to obtain a second image to be tested; Identify the edge of the expansion joint in the second image to be measured, and calculate the width of the expansion joint; The second image to be tested also includes multiple frames of continuous sub-images, and the edge position prediction value of the t+1th frame sub-image in the second image to be tested is predicted and calculated. ; According to the formula Calculate the edge position value of the sub-image, where is the edge position value of the sub-image of the t+1th frame, is the actual detection value of the edge position obtained by edge detection of the sub-image of the t+1th frame, K is the gain coefficient, t≥1, and t is a natural number; The width of the expansion joint of each frame sub-image is calculated according to the edge position value of each frame sub-image.

2. A method for measuring expansion joint width according to claim 1, characterized in that: The step of identifying the expansion joint edge in the second image to be measured and calculating the width of the expansion joint further includes: Preprocessing the second image to be tested to obtain a third image to be tested; Performing edge detection processing on the third image to be tested to obtain expansion joint edge data; The width of the expansion joint is calculated based on the expansion joint edge data and the calibration parameters.

3. A method for measuring the width of an expansion joint according to claim 2, characterized in that: The preprocessing of the second image to be tested to obtain a third image to be tested includes: Binarizing and denoising the second image to be tested, and contrast enhancing the feature part in the image to obtain a third image to be tested; The characteristic portion is a portion of the image that includes the edge of the expansion joint.

4. A method for measuring expansion joint width according to claim 2, characterized in that: The performing edge detection processing on the third image to be tested to obtain expansion joint edge data includes: Identify the edge contour of the expansion joint in the third image to be tested based on an edge detection algorithm; The edge profile of the expansion joint is optimized based on an edge refinement algorithm to obtain the edge data of the expansion joint.

5. A method for measuring the width of an expansion joint according to claim 4, characterized in that: The step of identifying the expansion joint edge in the third image to be tested based on an edge detection algorithm includes: Identifying the edge contour of the expansion joint in the third image to be tested; Determine whether the edge profile of the expansion joint meets the preset accuracy index. If not, adjust the parameters of the edge detection algorithm and return to the previous step to record the number of returns C. If the number of returns C meets the preset accuracy index, When , the recognition result is output, where C and is a positive integer; If yes, the recognition result is output.

6. A method for measuring the width of an expansion joint according to claim 5, characterized in that: The determining whether the edge profile of the expansion joint meets a preset accuracy index includes: Calculating the ratio of the length of the expansion joint edge to the expected expansion joint edge length; It is determined whether the ratio is greater than or equal to a preset ratio.

7. A method for measuring the width of an expansion joint according to claim 5, characterized in that: The step of determining whether the edge profile of the expansion joint meets a preset accuracy index further includes: Calculating the maximum offset between the position of the expansion joint edge and the expected expansion joint edge position; It is determined whether the maximum offset is less than or equal to a preset maximum offset threshold.

8. The method for measuring the width of an expansion joint according to claim 5, characterized in that: The adjusting the parameters of the edge detection algorithm comprises: According to the formula and / or ,Adjustment and / or The value of the adaptive high threshold and / or adaptive low threshold Make adjustments; in, High threshold proportionality coefficient, M is the median value of the image gradient amplitude, is the low threshold scale factor.

9. A method for measuring the width of an expansion joint according to claim 4, characterized in that: The step of optimizing the expansion joint edge based on an edge thinning algorithm to obtain the expansion joint edge data includes: Sampling intensity values ​​of one or more edge pixels in the expansion joint edge along a direction perpendicular to the image edge to create an intensity profile; According to the formula , the intensity profile is fitted, where For location The pixel intensity at is the horizontal coordinate of the edge position of the sub-pixel in the fitting coordinate system, for The horizontal coordinate of the pixel at in the fitting coordinate system, a and b are preset fitting parameters; The fitted intensity profile data is output to obtain the expansion joint edge data.

10. A method for measuring expansion joint width according to claim 2, characterized in that: Calculating the width of the expansion joint based on the expansion joint edge data and the calibration parameters includes: Calculating a pixel-to-width conversion scaling factor based on the calibration parameters; The distance between the expansion joint edges is calculated according to the expansion joint edge data and the pixel-width conversion ratio factor to obtain the width of the expansion joint.

11. A method for measuring expansion joint width according to claim 1, characterized in that: The step of calibrating the photographing device to obtain calibration parameters includes: Set the calibration pattern; A photographing device photographs the calibration pattern to obtain a calibration image set, wherein the calibration image set includes one or more images containing the calibration pattern; The calibration image set is analyzed and processed to obtain distortion calibration parameters and perspective correction parameters.

12. A method for measuring the width of an expansion joint according to claim 11, characterized in that: The analyzing and processing the calibration image set to obtain distortion calibration parameters and perspective correction parameters also includes: Determine whether the photographing device is perpendicular to the plane of the calibration pattern, and if so, the perspective correction parameters include external parameters, and the external parameters include a rotation matrix and a translation vector; If not, the perspective correction parameters include external parameters and a perspective transformation matrix, and the external parameters include a rotation matrix and a translation vector.

13. The method for measuring the width of an expansion joint according to claim 1, characterized in that: The prediction calculation is to calculate the edge position prediction value of the t+1th frame sub-image in the second image to be tested , also includes: The edge position prediction value of the sub-image in the second image to be tested is predicted and calculated by the following formula: ,in, is the edge position prediction value of the sub-image of the t+1th frame, is the edge position value of the sub-image of the tth frame, is the edge position change rate of the sub-image in the tth frame, is the frame interval, t≥1, and t is a natural number.

Images