CN110634128A - A ball stud pin size detection method, device, computer equipment and storage medium - Google Patents

Abstract

The present invention is applicable to the computer field, and provides a ball stud size detection method, device, computer equipment and storage medium. The ball stud size detection method includes: acquiring a first image, preprocessing the first image, and obtaining The second image, the second image at least includes a contour map of the ball stud; edge detection is performed on the contour map to determine the edge line of the ball stud; straight line detection based on Hough transform is performed on the edge line to determine the straight line part in the edge line, Determine the two endpoints of the straight line, and use the Euclidean distance to measure the distance between the two endpoints, and obtain and output the linear dimension of the ball stud. The method for detecting the size of the ball stud provided by the embodiment of the present invention uses a computer program to identify and measure the image of the ball stud, eliminates image noise through image preprocessing, image edge detection, and straight line detection, and measures the size through the computer program. Fast speed and high efficiency ensure product quality while reducing production costs.

Description

A ball stud pin size detection method, device, computer equipment and storage medium

technical field

The invention belongs to the field of computers, and in particular relates to a ball stud size detection method, device, computer equipment and storage medium.

Background technique

Ball studs are widely used in independent suspension systems to realize the connection between control arms or thrust rods and other components. important detection parameters.

The existing detection of the size and shape of the ball stud is done manually using a jig, but the jig will be worn during use, resulting in errors in the detection of the ball stud, and the manual detection has low efficiency and low accuracy. The shortcomings of poor quality and limited testing quantity, high labor costs, and workers are prone to fatigue, resulting in missed or false detections, resulting in defective or substandard products flowing into the market, causing reputational and economic losses to businesses.

It can be seen that the detection of the size of the existing ball stud basically relies on manual detection, which is costly, low in precision, and low in efficiency, and urgently needs to be improved.

Contents of the invention

The purpose of the embodiments of the present invention is to provide a ball stud size detection method, device, computer equipment and storage medium, aiming to solve the problem that the existing ball stud size detection basically relies on manual detection, which has high cost and low precision, and Inefficient technical issues.

The embodiment of the present invention is achieved in this way, a method for detecting the size of a ball stud, the method comprising:

acquiring a first image, the first image including at least a complete ball stud image;

Preprocessing the first image to obtain a second image, the second image at least including a profile of the ball stud;

performing edge detection on the contour map, and determining the edge line of the ball stud in the contour map;

Carrying out straight line detection based on Hough transform on the edge line, and determining the straight line part in the edge line;

Using the nearest neighbor search algorithm to determine the two end points of the straight line part, and using the Euclidean distance to measure the distance between the two end points in the coordinate system, to obtain and output the straight line size of the ball stud.

Another object of the embodiment of the present invention is a ball stud size detection device, the ball stud size detection device includes:

An image acquisition device, configured to acquire a first image, the first image at least including a complete ball stud image;

An image preprocessing device, configured to preprocess the first image to obtain a second image, the second image at least including the outline of the ball stud;

an edge detection device, configured to perform edge detection on the contour map, and determine the edge line of the ball stud in the contour map;

A straight line detection device, configured to perform straight line detection based on Hough transform on the edge line, and determine a straight line part in the edge line; and

The information output device is used to determine the two end points of the straight line part by using the nearest neighbor search algorithm, and measure the distance between the two end points by using the Euclidean distance in the coordinate system, and obtain and output the straight line of the ball stud size.

Another object of the embodiments of the present invention is a computer device, which is characterized in that it includes a memory and a processor, a computer program is stored in the memory, and when the computer program is executed by the processor, the processor Execute the steps of the method for detecting the size of the ball stud described in the above embodiments.

Another object of the embodiments of the present invention is a computer-readable storage medium, wherein a computer program is stored on the computer-readable storage medium, and when the computer program is executed by a processor, the processor executes The steps of the method for detecting the size of the ball stud described in the above-mentioned embodiments.

A method for detecting the size of a ball stud provided by an embodiment of the present invention collects an image of a ball stud with detection, and performs preprocessing on the image to realize filtering and denoising of the image to obtain a contour map of the ball stud, and then through Edge detection, find out the edge line of the ball stud, perform straight line detection based on Hough transform on the edge line, determine the straight line part in the edge line, then measure the length of the straight line part, determine the size of the ball stud pin, and measure the ball stud through the computer program The image of the head pin is recognized and measured. Through image preprocessing, image edge detection and line detection, image noise is eliminated to ensure the accuracy of size detection. At the same time, the size is measured through computer programs, which is fast and efficient, ensuring product quality. At the same time reduce production costs.

Description of drawings

Fig. 1 shows a flow chart of a ball stud size detection method provided by an embodiment of the present invention;

Fig. 2 shows a schematic diagram of the shape of a ball stud provided by the implementation of the present invention;

Fig. 3 shows a gradient intensity diagram of a pixel provided by an embodiment of the present invention;

Fig. 4 shows a schematic structural diagram of a ball stud size detection device provided by an embodiment of the present invention;

Fig. 5 shows an internal structural diagram of a computer device provided by an embodiment of the present invention.

Detailed ways

In order to make the object, technical solution and advantages of the present invention clearer, the present invention will be further described in detail below in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described here are only used to explain the present invention, not to limit the present invention.

Fig. 1 is a flow chart of a method for detecting the size of a ball stud provided by an embodiment of the present invention. As shown in Figure 1, a method for detecting the size of a ball stud may specifically include the following steps:

Step S102, acquiring a first image, the first image at least including a complete ball stud image.

In the embodiment of the present invention, Fig. 2 shows a schematic diagram of the shape of a ball stud provided by the present invention. As shown in Fig. 2, the ball stud is widely used in the independent suspension system, the control arm or the thrust rod It is often connected to other components through a ball stud located at the end. The main function of the ball stud is to realize the up and down beating and steering movement of the wheel. Wherein, the first image should at least include a complete image of the ball stud, and the image of the ball stud should be able to show the part of the ball stud whose size is to be measured.

In the embodiment of the present invention, the first image may be acquired through a wireless network or a wired network, or directly read the first picture stored in a removable storage medium, which is not limited in the present invention.

The embodiment of the present invention obtains the image of the ball stud in various ways, which is convenient for subsequent size detection, and can be applied to various scenarios at the same time.

Step S104, performing preprocessing on the first image to obtain a second image, the second image at least including a contour map of the ball stud.

In the embodiment of the present invention, the contour map refers to the shape contour map of the ball stud in the image, and the preprocessing refers to the operations performed on the image before the length in the image is detected, including filtering, de-drying, sharpening, enhancement etc.

As an embodiment of the present invention, the preprocessing of the first image may include actual advanced visualization processing and any preparation of image data for subsequent advanced visualization processing, including performing image enhancement processing on the original image, transforming the gray scale of the image, Make the image clearer, enhance the contrast, highlight the edge features, and perform image smoothing on the processed image to remove additive noise, multiplicative noise and quantization noise. As an embodiment of the present invention, grayscale transformation is performed on the original image, the grayscale transformation is to enhance the contrast of the image or stretch the contrast of the image to generate a grayscale enhanced image, and transform the grayscale enhanced image into a histogram Image equalization image: Set the original grayscale of the pixel in the original image as R, the grayscale of the pixel after grayscale transformation as S, and the grayscale transformation function as T(R), then perform grayscale according to the following formula transform:

Among them, 0≤R _j ≤l-1; p _y (R _j ) is the probability of gray level j, n _j is the total number of pixels of gray level j in the image, and l is the total number of gray levels in the image , n is the total number of pixels in the image, and j identifies the gray level, which is consistent with the gray level that can be recognized by the computer.

The image preprocessing method provided by the embodiment of the present invention solves the technical problem of clearing the image acquired by the camera. The present invention adopts image enhancement, smoothing filtering and image sharpening processing methods to effectively eliminate noise and improve image quality , to make the image clearer, which is conducive to better extracting effective information for subsequent analysis.

Step S106, performing edge detection on the outline image, and determining the edge line of the ball stud in the outline image.

In the embodiment of the present invention, edge detection is a basic problem in image processing and computer vision. The purpose of edge detection is to identify points with obvious brightness changes in digital images. Significant changes in image attributes usually reflect important events and changes in attributes. , these include discontinuities in depth, discontinuities in surface orientation, changes in material properties, and changes in scene lighting.

As an embodiment of the present invention, a Canny edge detection operator and/or a Sobel edge detection operator may be used to perform edge detection on the contour map to determine the edge line of the ball stud. As an embodiment of the present invention, the Canny edge detection operator is used to detect the edge of the image. The first step is to convolve the original data with the Gaussian mask, and the obtained image is slightly blurred compared with the original image. In this way, a single pixel noise becomes almost insignificant on the Gaussian smoothed image. Then look for the intensity gradient in the image. The gradient of each pixel in the smoothed image can be obtained by the Sobel operator. First, use the following kernel to obtain the gradient along the horizontal (x) and vertical (y) directions respectively G _x and G _y , such as:

=[-101;-202;-121];=[-1-2-1;000;121]

Then you can use the formula to find the gradient magnitude of each pixel:

Sometimes for the convenience of calculation, the infinite norms of G _x and G _y are used instead of the two norms. Replace each point in the smoothed image with G, and in places with drastic changes (at the boundary), a larger gradient measure value G will be obtained, and the corresponding color will be white; however, these boundaries are usually very thick and difficult to calibrate The real location of the border. In order to do this, the gradient direction must also be stored, with the following formula:

Θ = arctan2(G _y , G _x )

After finding the intensity gradient of the image, approximate its gradient direction to one of the following values (0, 45, 90, 135, 180, 225, 270, 315) and compare the pixel with the gradient strength of the pixel in the positive and negative directions of the gradient direction. If the pixel If the gradient strength is the largest, it is retained, otherwise it is suppressed, as shown in Figure 3, the number in the figure represents the gradient strength of the pixel, and the direction of the arrow represents the gradient direction. Taking the third pixel in the second row as an example, since the gradient direction is upward, compare the intensity (7) of this point with the intensity (5 and 4) of the two pixels above and below, and keep it because the intensity of this point is the largest. Because the gradient direction at the boundary always points to the direction perpendicular to the boundary, that is, the brightest thin line at the boundary will be reserved at the end, and then the edge line of the ball stud will be detected by using the subsequent boundary tracking.

In the embodiment of the present invention, the accurate edge line of the ball stud in the image is obtained through edge detection, which facilitates subsequent recognition of the size of the ball stud.

Step S108, performing straight line detection based on Hough transform on the edge line, and determining a straight line part in the edge line.

In the embodiment of the present invention, the dual property of point and line is used, that is, the collinear points in the image space correspond to the intersecting lines in the parameter space, and all lines intersecting at the same point in the parameter space have a common According to the corresponding characteristics of the points of the line, the problem of line detection in the image space is transformed into the problem of point detection in the parameter space, and the detection task is completed by simple accumulation statistics in the parameter space.

As an embodiment of the present invention, after the edge line of the ball stud in the image is extracted, it is necessary to detect the straight portion of the edge line of the ball stud, so as to detect the size of the straight portion of the ball stud. In the embodiment of the present invention, the straight line detection method based on Hough transform is used to detect the straight line on the edge line, and the straight line is detected by performing simple accumulation statistics in the parameter space, and then finding the peak value of the accumulator in the Hough parameter space, and the Hough transform The essence of this method is to cluster the pixels with a certain relationship in the image space, and find the corresponding point of parameter space accumulation that can connect these pixels with a certain analytical form. The specific operation steps will be explained in detail in the subsequent part of the manual. .

In the embodiment of the present invention, the edge line of the ball stud is detected by the straight line detection based on Hough transform, and the straight line part in the ball head image is obtained, which facilitates subsequent detection of the size of the straight line part of the ball stud, with high recognition accuracy and accurate detection results. .

Step S110, using the nearest neighbor search algorithm to determine the two end points of the straight line, and using the Euclidean distance to measure the distance between the two end points in the coordinate system, to obtain and output the straight line size of the ball stud.

In the embodiment of the present invention, the Euclidean distance refers to the Euclidean distance, and refers to the ordinary distance between two points in the Euclidean space, that is, the straight-line distance.

As an embodiment of the present invention, in the straight line part obtained in step S108, through the confirmation of the gray value, when the gray value of a certain pixel area is obviously higher than the gray value of the three directions of the area, and slightly lower For the gray value in the other direction, use the pixel area as an end point, and use the nearest neighbor search algorithm to find the second end point in the straight line part, so as to determine the two end points of the straight line part, and project the two end points into a Cartesian coordinate system, then use the formula:

Calculate the Euclidean distance l between two endpoints, where x ₁ , y ₁ , x ₂ , and y ₂ are the coordinate values of the first endpoint and the second endpoint in the Cartesian coordinate system, respectively. After the distance between the two endpoints is obtained by calculation, directly output l, which is the size of the straight part of the ball stud.

A method for detecting the size of a ball stud provided by an embodiment of the present invention collects an image of a ball stud with detection, and performs preprocessing on the image to realize filtering and denoising of the image to obtain a contour map of the ball stud, and then through Edge detection, find out the edge line of the ball stud, perform straight line detection based on Hough transform on the edge line, determine the straight line part in the edge line, then measure the length of the straight line part, determine the size of the ball stud pin, and measure the ball stud through the computer program The image of the head pin is recognized and measured. Through image preprocessing, image edge detection and line detection, image noise is eliminated to ensure the accuracy of size detection. At the same time, the size is measured through computer programs, which is fast and efficient, ensuring product quality. At the same time reduce production costs.

As an embodiment of the present invention, the preprocessing the first image to obtain the second image includes:

Gradient sharpening is performed on the first image; the gradient sharpening formula is:

Among them, f(x, y) represents the image, G′ _M [f(x, y)] is the final grayscale replacement value,

G′ _M [f(x, y)] _max = 225, T′ ₁ represents the sharpening threshold, when the gradient value is greater than T′ ₁ , the gradient edge is strengthened; T′ ₂ represents the grayscale threshold, when the image grayscale is greater than T ′ ₂ , the gray value minus T ₂ .

Obtain the edge gray value of the ball stud image in the sharpened first image, and use the connected pixels whose gray value difference satisfies the edge judgment condition as the outline of the ball stud to obtain the first Two images.

In the embodiment of the present invention, the existing image sharpening uses a two-way one-time differential operation. After the gradient is calculated, the gradient value is directly used to replace the gray value of the point, and the pixel values in the last row and column of the image are filled with the adjacent gradient values. Experiments have found that directly replacing the gray value with the gradient value will cause the image to lose a large amount of original information. According to the characteristics of the ball stud image, the present invention improves the gradient sharpening by setting a threshold judgment. The specific formula is:

Among them, f(x, y) represents the image, G′ _M [f(x, y)] is the final grayscale replacement value, G′ _M [f(x, y)] _max = 225, T′ ₁ represents the sharpness When the gradient value is greater than T′ ₁ , the gradient edge is strengthened; T′ ₂ represents the grayscale threshold, when the image grayscale is greater than T′ ₂ , the grayscale value is subtracted by T ₂ ; the above gradient sharpening method can be used to retain The high gray value information of the original ball stud image eliminates its influence on the gradient edge at the same time, and the gray value remains unchanged in other cases. The processed image not only enhances the outline of the ball stud but also increases the distinction between the edge information and other backgrounds. . After sharpening the edge, obtain the gray value of the image and analyze the gray value. When the gray value of the image shows a clear boundary line, the boundary line is the edge line of the ball stud.

The embodiment of the present invention sharpens the ball stud image, strengthens the edge gradient, and determines the edge line of the ball stud through the distribution of gray values, and recognizes and measures the straight line portion of the subsequent ball stud with fast speed. The measurements are accurate.

In an embodiment of the present invention, the performing edge detection on the contour map and determining the edge line of the ball stud in the contour map includes:

An edge detection is performed on the contour map by using a Canny edge detection operator and/or a Sobel edge detection operator to determine the edge line of the ball stud.

As an embodiment of the present invention, a Canny edge detection operator and/or a Sobel edge detection operator may be used to perform edge detection on the contour map to determine the edge line of the ball stud. As an embodiment of the present invention, the Canny edge detection operator is used to detect the edge of the image. The first step is to convolve the original data with the Gaussian mask, and the obtained image is slightly blurred compared with the original image. In this way, a single pixel noise becomes almost insignificant on the Gaussian smoothed image. Then look for the intensity gradient in the image. The gradient of each pixel in the smoothed image can be obtained by the Sobel operator. First, use the following kernel to obtain the gradient along the horizontal (x) and vertical (y) directions respectively G _x and G _y , such as:

=[-101;-202;-121];=[-1-2-1;000;121]

Then you can use the formula to find the gradient magnitude of each pixel:

Sometimes for the convenience of calculation, the infinite norms of G _x and G _y are used instead of the two norms. Replace each point in the smoothed image with G, and in places with drastic changes (at the boundary), a larger gradient measure value G will be obtained, and the corresponding color will be white; however, these boundaries are usually very thick and difficult to calibrate The real location of the border. In order to do this, the gradient direction must also be stored, with the following formula:

Θ = arctan2(G _y , G _x )

After finding the intensity gradient of the image, approximate its gradient direction to one of the following values (0, 45, 90, 135, 180, 225, 270, 315) and compare the pixel with the gradient strength of the pixel in the positive and negative directions of the gradient direction. If the pixel If the gradient strength is the largest, it is retained, otherwise it is suppressed, as shown in Figure 3, the number in the figure represents the gradient strength of the pixel, and the direction of the arrow represents the gradient direction. Taking the third pixel in the second row as an example, since the gradient direction is upward, compare the intensity (7) of this point with the intensity (5 and 4) of the two pixels above and below, and keep it because the intensity of this point is the largest. Because the gradient direction at the boundary always points to the direction perpendicular to the boundary, that is, the brightest thin line at the boundary will be reserved at the end, and then the edge line of the ball stud will be detected by using the subsequent boundary tracking.

In the embodiment of the present invention, the accurate edge line of the ball stud in the image is obtained through edge detection, which facilitates subsequent recognition of the size of the ball stud.

As an embodiment of the present invention, the performing straight line detection based on Hough transform on the edge line, and determining the straight line part in the edge line include:

Projecting the edge line into a plane Cartesian coordinate system, and obtaining coordinates of sample points on the edge line;

Through the polar coordinate operation, the sample point is transformed into the polar coordinate system; wherein, the transformation equation is as follows:

Wherein, x and y represent the abscissa and ordinate of the sample point in the plane Cartesian coordinate system,

ρ represents the polar diameter of the sample point, and θ is the polar angle of the sample point;

Detecting the number of collinear points of the sample points under the polar coordinates, and judging that a straight line is detected when the number of collinear points is greater than a preset value.

In the embodiment of the present invention, a point in the original image space corresponds to a sinusoid in the new parameter space, that is, a point-sine curve dual. The specific process of detecting a straight line is to let θ take all possible values, then calculate the value of ρ, and then accumulate the accumulation array according to the values of θ and ρ, so as to obtain the number of collinear points. When the line rotates counterclockwise from the coincidence with the x-axis, the value of θ starts to increase from 0° to 180°, that is, the value range of θ is 0°～180°. From the polar coordinate equation of a straight line, we know that:

So if and only if both x and y reach the maximum and θ+Φ=±90°, there are:

That is, the value range of ρ is:

From the value range of θ and ρ and their resolution, the size of the Hough transform accumulator can be determined, and the points in the Hough transform accumulator whose cumulative value is less than the threshold value are cleared according to the threshold value, that is, these points are considered not to correspond to the image a line in the domain until a line is detected.

In the embodiment of the present invention, the straight line portion of the outline of the ball stud is detected through a straight line detection algorithm based on the Hough transform, so as to ensure the accuracy of the line detection and the accuracy of the dimension measurement of the straight line portion of the ball stud.

As an embodiment of the present invention, the nearest neighbor search algorithm is used to determine the two endpoints of the straight line, and the distance between the two endpoints is measured by the Euclidean distance in the coordinate system to obtain and output the spherical Before the linear dimensions of the header pins, also include:

其中K为相机的像素尺寸，N为所述工件在图像上所占像素的个数；Determine the magnification of the optical system by statically calibrating the workpiece with known size L

Wherein K is the pixel size of the camera, and N is the number of pixels occupied by the workpiece on the image;

计算所述光学系统调节后的放大倍数β₁，其中，δ为振动位移，μ为静态标定时所述光学系统的物距；by formula

Calculate the adjusted magnification β ₁ of the optical system, where δ is the vibration displacement, and μ is the object distance of the optical system during static calibration;

计算所述球头销的尺寸；其中，l为进行误差补偿后的所述球头销的尺寸，N₁为所述球头销在所述第二图像上所占像素点的个数use the formula

Calculating the size of the ball stud; wherein, _l is the size of the ball stud after error compensation, and N1 is the number of pixels occupied by the ball stud on the second image

In the embodiment of the present invention, during the high-speed online detection process, the transmission mechanism drives the workpiece to run, and the workpiece and the surface of the transmission mechanism will vibrate, which can be decomposed into two directions: vertical vibration perpendicular to the direction of movement of the transmission mechanism, along the Vibration back and forth in the direction of movement of the transmission mechanism. Vertical vibration will cause changes in the object distance in the optical imaging system, change the magnification of the imaging system, and cause size measurement errors. Back and forth vibration means that the transmission mechanism runs at a certain speed during the online detection process, resulting in relative motion between the workpiece and the transmission mechanism. Image blur caused by motion blur in the online detection process will make the details of the object surface unclear, directly affect the clarity of the image, and make it difficult to distinguish the surface features of the workpiece image. Motion blur will enlarge the edge of the image and affect the size measurement precision.

As an embodiment of the present invention, the vibration error compensation is divided into vertical error compensation and horizontal vibration error compensation. In the vertical error compensation, the online size detection of the workpiece must establish the corresponding relationship between the image pixels and the actual size. Use a workpiece of known size L for static calibration to determine the imaging magnification. The optical system magnification is β, and the camera pixel size is K. The number N of pixels occupied by the workpiece on the image is obtained by processing the workpiece image. Then the magnification is:

Do not change the parameters of the online image acquisition device to detect the online size of the workpiece, regardless of the vertical vibration compensation, the number of pixels occupied by the workpiece on the image, set the target workpiece size detection value to N ₁ , get N from the above formula as:

In the process of online size detection, the vertical vibration error compensation of the workpiece size is carried out. Under the condition _of not changing the focal length, the magnification β1 is calculated by the optical Gaussian formula as:

Where δ is the vibration displacement, β is the magnification during static calibration, and the object distance μ of the optical imaging system during static calibration is calculated from β and the focal length.

Consider the influence of the vertical vibration factor on the dimension measurement accuracy, and compensate the error caused by the vibration to the detection dimension according to the above _formula , so the dimension detection value L after compensation is:

Among them, N ₁ is the number of pixels occupied by the workpiece on the image during online detection. The method corrects the magnification of the optical system in real time through the online measurement of the vertical vibration displacement of the workpiece, and compensates the size detection error caused by the vertical vibration.

For the horizontal vibration error, when the speed v of the transmission mechanism is constant, and the exposure time of the camera is set to t, then the motion blur value of the measured object relative to the imaging system is:

x=vtβ

When v and β are determined, x is affected by t. In practical applications, as long as it is online detection, motion blur cannot be eliminated due to the existence of speed, so it can only be controlled and reduced to a size that does not affect the measurement accuracy. The commonly used method is to control the motion blur value by reducing the exposure time, which is beneficial to reduce the influence of the speed factor and front and rear vibration on the size measurement value in the online detection. However, the reduction of motion blur needs to be at the cost of reducing the exposure time. The reduction of exposure time will make the image brightness insufficient and difficult to recognize. The light source will increase the detection cost, and the high-brightness light source generates a lot of heat, which is not conducive to long-term work. A large aperture reduces the depth of field of the lens. Therefore, it is necessary to comprehensively consider reducing the exposure time, controlling the size of the clear aperture, and adjusting the brightness of the light source to achieve motion blur control.

In the embodiment of the present invention, by performing vertical vibration error compensation and horizontal vibration error compensation on the image, the influence of errors on the measurement results is reduced as much as possible, the accuracy of the size measurement results is improved, and the product accuracy and yield are guaranteed.

A method for detecting the size of a ball stud provided by an embodiment of the present invention collects an image of a ball stud with detection, and performs preprocessing on the image to realize filtering and denoising of the image to obtain a contour map of the ball stud, and then through Edge detection, find out the edge line of the ball stud, perform straight line detection based on Hough transform on the edge line, determine the straight line part in the edge line, then measure the length of the straight line part, determine the size of the ball stud pin, and measure the ball stud through the computer program The image of the head pin is recognized and measured. Through image preprocessing, image edge detection and line detection, image noise is eliminated to ensure the accuracy of size detection. At the same time, the size is measured through computer programs, which is fast and efficient, ensuring product quality. At the same time reduce production costs.

Fig. 4 shows a schematic structural diagram of a ball stud size detection device provided by an embodiment of the present invention. As shown in Fig. 4, the ball stud size detection device provided by an embodiment of the present invention includes:

The image acquiring device 410 is configured to acquire a first image, where the first image at least includes a complete image of the ball stud.

In the embodiment of the present invention, Fig. 2 shows a schematic diagram of the shape of a ball stud provided by the present invention. As shown in Fig. 2, the ball stud is widely used in the independent suspension system, the control arm or the thrust rod It is often connected to other components through a ball stud located at the end. The main function of the ball stud is to realize the up and down beating and steering movement of the wheel. Wherein, the first image should at least include a complete image of the ball stud, and the image of the ball stud should be able to show the part of the ball stud whose size is to be measured.

In the embodiment of the present invention, the first image may be acquired through a wireless network, or through a wired network, or through a local area network, or directly read the first picture stored in a removable storage medium, The present invention is not limited.

The embodiment of the present invention obtains the image of the ball stud in various ways, which is convenient for subsequent size detection, and can be applied to various scenarios at the same time.

The image preprocessing device 420 is configured to preprocess the first image to obtain a second image, and the second image includes at least a contour map of the ball stud.

In the embodiment of the present invention, the contour map refers to the shape contour map of the ball stud in the image, and the preprocessing refers to the operations performed on the image before the length in the image is detected, including filtering, de-drying, sharpening, enhancement etc.

As an embodiment of the present invention, the preprocessing of the first image may include actual advanced visualization processing and any preparation of image data for subsequent advanced visualization processing, including performing image enhancement processing on the original image, transforming the gray scale of the image, Make the image clearer, enhance the contrast, highlight the edge features, and perform image smoothing on the processed image to remove additive noise, multiplicative noise and quantization noise. As an embodiment of the present invention, grayscale transformation is performed on the original image, the grayscale transformation is to enhance the contrast of the image or stretch the contrast of the image to generate a grayscale enhanced image, and transform the grayscale enhanced image into a histogram Image equalization image: Set the original grayscale of the pixel in the original image as R, the grayscale of the pixel after grayscale transformation as S, and the grayscale transformation function as T(R), then perform the grayscale according to the following formula transform:

Among them, 0≤R _j ≤l-1; p _y (R _j ) is the probability of gray level j, n _j is the total number of pixels of gray level j in the image, and l is the total number of gray levels in the image , n is the total number of pixels in the image, and j identifies the gray level, which is consistent with the gray level that can be recognized by the computer.

The image preprocessing method provided by the embodiment of the present invention solves the technical problem of clearing the image acquired by the camera. The present invention adopts image enhancement, smoothing filtering and image sharpening processing methods to effectively eliminate noise and improve image quality , to make the image clearer, which is conducive to better extracting effective information for subsequent analysis.

The edge detection device 430 is configured to perform edge detection on the contour map, and determine the edge line of the ball stud in the contour map.

In the embodiment of the present invention, edge detection is a basic problem in image processing and computer vision. The purpose of edge detection is to identify points with obvious brightness changes in digital images. Significant changes in image attributes usually reflect important events and changes in attributes. , these include discontinuities in depth, discontinuities in surface orientation, changes in material properties, and changes in scene lighting.

As an embodiment of the present invention, a Canny edge detection operator and/or a Sobel edge detection operator may be used to perform edge detection on the contour map to determine the edge line of the ball stud. As an embodiment of the present invention, the Canny edge detection operator is used to detect the edge of the image. The first step is to convolve the original data with the Gaussian mask, and the obtained image is slightly blurred compared with the original image. In this way, a single pixel noise becomes almost insignificant on the Gaussian smoothed image. Then look for the intensity gradient in the image. The gradient of each pixel in the smoothed image can be obtained by the Sobel operator. First, use the following sum to obtain the gradient along the horizontal (x) and vertical (y) directions respectively G _x and G _y , such as:

=[-101;-202;-121];=[-1-2-1;000;121]

Then you can use the formula to find the gradient magnitude of each pixel:

Sometimes for the convenience of calculation, the infinite norms of G _x and G _y are used instead of the two norms. Replace each point in the smoothed image with G, and in places with drastic changes (at the boundary), a larger gradient measure value G will be obtained, and the corresponding color will be white; however, these boundaries are usually very thick and difficult to calibrate The real location of the border. In order to do this, the gradient direction must also be stored, with the following formula:

Θ = arctan2(G _y , G _x )

After finding the intensity gradient of the image, approximate its gradient direction to one of the following values (0, 45, 90, 135, 180, 225, 270, 315) and compare the pixel with the gradient strength of the pixel in the positive and negative directions of the gradient direction. If the pixel If the gradient strength is the largest, it is retained, otherwise it is suppressed, as shown in Figure 2, the numbers in the figure represent the gradient strength of the pixel, and the direction of the arrow represents the gradient direction. Taking the third pixel in the second row as an example, since the gradient direction is upward, compare the intensity (7) of this point with the intensity (5 and 4) of the two pixels above and below, and keep it because the intensity of this point is the largest. Because the gradient direction at the boundary always points to the direction perpendicular to the boundary, that is, the brightest thin line at the boundary will be reserved at the end, and then the edge line of the ball stud will be detected by using the subsequent boundary tracking.

In the embodiment of the present invention, the accurate edge line of the ball stud in the image is obtained through edge detection, which facilitates subsequent recognition of the size of the ball stud.

The straight line detecting device 440 is configured to perform straight line detection based on Hough transform on the edge line, and determine the straight line part in the edge line.

In the embodiment of the present invention, the dual property of point and line is used, that is, the collinear points in the image space correspond to the intersecting lines in the parameter space, and all lines intersecting at the same point in the parameter space have a common According to the corresponding characteristics of the points of the line, the problem of line detection in the image space is transformed into the problem of point detection in the parameter space, and the detection task is completed by simple accumulation statistics in the parameter space.

As an embodiment of the present invention, after the edge line of the ball stud in the image is extracted, it is necessary to detect the straight portion of the edge line of the ball stud, so as to detect the size of the straight portion of the ball stud. In the embodiment of the present invention, the straight line detection method based on Hough transform is used to detect the straight line on the edge line, and the straight line is detected by performing simple accumulation statistics in the parameter space, and then finding the peak value of the accumulator in the Hough parameter space, and the Hough transform The essence of this method is to cluster the pixels with a certain relationship in the image space, and find the corresponding point of parameter space accumulation that can connect these pixels with a certain analytical form. The specific operation steps will be explained in detail in the subsequent part of the manual. .

In the embodiment of the present invention, the edge line of the ball stud is detected by the straight line detection based on Hough transform, and the straight line part in the ball head image is obtained, which facilitates subsequent detection of the size of the straight line part of the ball stud, with high recognition accuracy and accurate detection results. .

The information output device 450 is configured to use the nearest neighbor search algorithm to determine the two end points of the straight line, and use the Euclidean distance to measure the distance between the two end points in the coordinate system, and obtain and output the distance between the two end points of the ball stud. linear dimension.

In the embodiment of the present invention, the Euclidean distance refers to the Euclidean distance, and refers to the ordinary distance between two points in the Euclidean space, that is, the straight-line distance.

As an embodiment of the present invention, in the straight line part obtained in step S108, through the confirmation of the gray value, when the gray value of a certain pixel area is obviously higher than the gray value of the three directions of the area, and slightly lower For the gray value in the other direction, use the pixel area as an end point, and use the nearest neighbor search algorithm to find the second end point in the straight line part, so as to determine the two end points of the straight line part, and project the two end points into a Cartesian coordinate system, then use the formula:

Calculate the Euclidean distance l between two endpoints, where x ₁ , y ₁ , y ₂ , and y ₂ are the coordinate values of the first endpoint and the second endpoint in the Cartesian coordinate system, respectively. After the distance between the two endpoints is obtained by calculation, directly output l, which is the size of the straight part of the ball stud.

The ball stud size detection device provided by the embodiment of the present invention collects the image of the ball stud with the detection, and preprocesses the image to realize the filtering and denoising of the image, and obtains the contour map of the ball stud, and then passes Edge detection, find out the edge line of the ball stud, perform straight line detection based on Hough transform on the edge line, determine the straight line part in the edge line, then measure the length of the straight line part, determine the size of the ball stud pin, and measure the ball stud through the computer program The image of the head pin is recognized and measured. Through image preprocessing, image edge detection and line detection, image noise is eliminated to ensure the accuracy of size detection. At the same time, the size is measured through computer programs, which is fast and efficient, ensuring product quality. At the same time reduce production costs.

Figure 5 shows a diagram of the internal structure of a computer device in one embodiment. As shown in FIG. 5 , the computer equipment includes a processor, a memory, a network interface, an input device, and a display screen connected through a system bus. Wherein, the memory includes a non-volatile storage medium and an internal memory. The non-volatile storage medium of the computer device stores an operating system and also stores a computer program, and when the computer program is executed by the processor, the processor can realize the ball stud size detection method. A computer program may also be stored in the internal memory, and when the computer program is executed by the processor, the processor may execute the method for detecting the size of the ball stud. The display screen of the computer equipment may be a liquid crystal display screen or an electronic ink display screen, and the input device of the computer equipment may be a touch layer covered on the display screen, or a button, a trackball or a touch pad provided on the casing of the computer equipment, or It can be an external keyboard, touchpad or mouse.

Those skilled in the art can understand that the structure shown in Figure 5 is only a block diagram of a part of the structure related to the solution of this application, and does not constitute a limitation to the computer equipment on which the solution of this application is applied. The specific computer equipment can be More or fewer components than shown in the figures may be included, or some components may be combined, or have a different arrangement of components.

In one embodiment, a computer device is provided, the computer device includes a memory, a processor and a computer program stored on the memory and executable on the processor, the processor executes the computer The following steps are implemented in the program:

acquiring a first image, the first image including at least a complete ball stud image;

Preprocessing the first image to obtain a second image, the second image at least including a profile of the ball stud;

performing edge detection on the contour map, and determining the edge line of the ball stud in the contour map;

Carrying out straight line detection based on Hough transform on the edge line, and determining the straight line part in the edge line;

Using the nearest neighbor search algorithm to determine the two end points of the straight line part, and using the Euclidean distance to measure the distance between the two end points in the coordinate system, to obtain and output the straight line size of the ball stud.

In one embodiment, a computer-readable storage medium is provided. A computer program is stored on the computer-readable storage medium. When the computer program is executed by a processor, the processor is made to perform the following steps:

acquiring a first image, the first image including at least a complete ball stud image;

Preprocessing the first image to obtain a second image, the second image at least including a profile of the ball stud;

performing edge detection on the contour map, and determining the edge line of the ball stud in the contour map;

Carrying out straight line detection based on Hough transform on the edge line, and determining the straight line part in the edge line;

Using the nearest neighbor search algorithm to determine the two end points of the straight line part, and using the Euclidean distance to measure the distance between the two end points in the coordinate system, to obtain and output the straight line size of the ball stud.

It should be understood that although the various steps in the flow charts of the embodiments of the present invention are shown sequentially according to the arrows, these steps are not necessarily executed sequentially in the order indicated by the arrows. Unless otherwise specified herein, there is no strict order restriction on the execution of these steps, and these steps can be executed in other orders. Moreover, at least some of the steps in each embodiment may include multiple sub-steps or multiple stages, these sub-steps or stages are not necessarily executed at the same time, but may be executed at different times, the sub-steps or stages The order of execution is not necessarily performed sequentially, but may be performed alternately or alternately with at least a part of other steps or sub-steps or stages of other steps.

Those of ordinary skill in the art can understand that all or part of the processes in the methods of the above embodiments can be implemented through computer programs to instruct related hardware, and the programs can be stored in a non-volatile computer-readable storage medium When the program is executed, it may include the processes of the embodiments of the above-mentioned methods. Wherein, any references to memory, storage, database or other media used in the various embodiments provided in the present application may include non-volatile and/or volatile memory. Nonvolatile memory can include read only memory (ROM), programmable ROM (PROM), electrically programmable ROM (EPROM), electrically erasable programmable ROM (EEPROM), or flash memory. Volatile memory can include random access memory (RAM) or external cache memory. By way of illustration and not limitation, RAM is available in many forms such as Static RAM (SRAM), Dynamic RAM (DRAM), Synchronous DRAM (SDRAM), Double Data Rate SDRAM (DDRSDRAM), Enhanced SDRAM (ESDRAM), Synchronous Chain Synchlink DRAM (SLDRAM), memory bus (Rambus) direct RAM (RDRAM), direct memory bus dynamic RAM (DRDRAM), and memory bus dynamic RAM (RDRAM), etc.

The various technical features of the above-mentioned embodiments can be combined arbitrarily. To make the description concise, all possible combinations of the various technical features in the above-mentioned embodiments are not described. However, as long as there is no contradiction in the combination of these technical features, should be considered as within the scope of this specification.

The above-mentioned embodiments only express several implementation modes of the present invention, and the description thereof is relatively specific and detailed, but should not be construed as limiting the patent scope of the present invention. It should be pointed out that those skilled in the art can make several modifications and improvements without departing from the concept of the present invention, and these all belong to the protection scope of the present invention. Therefore, the protection scope of the patent for the present invention should be based on the appended claims.

The above descriptions are only preferred embodiments of the present invention, and are not intended to limit the present invention. Any modifications, equivalent replacements and improvements made within the spirit and principles of the present invention should be included in the protection of the present invention. within range.

Claims (10)

1. A ball stud size detection method, characterized in that the method comprises: acquiring a first image, the first image including at least a complete ball stud image; Preprocessing the first image to obtain a second image, the second image at least including a profile of the ball stud; performing edge detection on the contour map, and determining the edge line of the ball stud in the contour map; Carrying out straight line detection based on Hough transform on the edge line, and determining the straight line part in the edge line; Using the nearest neighbor search algorithm to determine the two end points of the straight line part, and using the Euclidean distance to measure the distance between the two end points in the coordinate system, to obtain and output the straight line size of the ball stud. 2. The method according to claim 1, wherein said preprocessing said first image to obtain a second image comprises: Gradient sharpening is performed on the first image; the gradient sharpening formula is:

Among them, f(x, y) represents the image, G′ _M [f(x, y)] is the final grayscale replacement value, G′ _M [f(x, y)] _max = 225, T′ ₁ represents the sharpness threshold, when the gradient value is greater than T′ ₁ , the gradient edge is strengthened; T′ ₂ represents the grayscale threshold, and when the image grayscale is greater than T′ ₂ , the grayscale value minus T ₂ ; Obtain the edge gray value of the ball stud image in the sharpened first image, and use the connected pixels whose gray value difference satisfies the edge judgment condition as the outline of the ball stud to obtain the first Two images. 3. The method according to claim 1, wherein the performing edge detection on the contour map and determining the edge line of the ball stud in the contour map comprises: An edge detection is performed on the contour map by using a Canny edge detection operator and/or a Sobel edge detection operator to determine the edge line of the ball stud. 4. The method according to claim 1, wherein the said edge line is carried out based on the straight line detection of Hough transform, and determining the straight line part in the edge line comprises: Projecting the edge line into a plane Cartesian coordinate system, and obtaining coordinates of sample points on the edge line; Through the polar coordinate operation, the sample point is transformed into the polar coordinate system; wherein, the transformation equation is as follows:

Wherein, x and y represent the abscissa and ordinate of the sample point in the plane Cartesian coordinate system, ρ represents the polar diameter of the sample point, and θ is the polar angle of the sample point; Detecting the number of collinear points of the sample points under the polar coordinates, and judging that a straight line is detected when the number of collinear points is greater than a preset value. 5. The method according to claim 4, wherein the detection of the number of collinear points of the sample point under the polar coordinates, when the number of the collinear points is greater than a preset value, Then it is judged that a straight line is detected, including: According to the preset step size, adjust the value of θ from 0 to 180, and calculate Determine the value range of ρ, where

According to the value of each group (ρ, θ), determine the value of the Hough transform accumulator corresponding to each group (ρ, θ), and the corresponding (ρ) of the Hough transform accumulator greater than the preset threshold , θ) correspond to the sample points as the collinear points, and when the number of the collinear points is greater than a preset value, it is determined that a straight line is detected. 6. The method according to claim 1, wherein said adopting the nearest neighbor search algorithm to determine the two endpoints of the straight line portion, and adopting the Euclidean distance to measure the distance between the two endpoints in the coordinate system , before obtaining and outputting the linear dimension of the ball stud, also include: Determine the magnification of the optical system by statically calibrating the workpiece with known size L Wherein K is the pixel size of the camera, and N is the number of pixels occupied by the workpiece on the image; by formula

Calculate the adjusted magnification β ₁ of the optical system, where δ is the vibration displacement, and μ is the object distance of the optical system during static calibration; use the formula

Calculate the size of the ball stud; wherein, l is the size of the ball stud after error compensation, and N ₁ is the number of pixels occupied by the ball stud on the second image. 7. The method according to claim 1, wherein said preprocessing said first image to obtain a second image further comprises: The first picture is processed by a morphological processing method to remove noise in the first image; wherein the morphological processing method includes at least dilation and erosion of the image. 8. A ball stud size detection device, characterized in that the ball stud size detection device comprises: An image acquisition device, configured to acquire a first image, the first image at least including a complete ball stud image; An image preprocessing device, configured to preprocess the first image to obtain a second image, the second image at least including the outline of the ball stud; an edge detection device, configured to perform edge detection on the contour map, and determine the edge line of the ball stud in the contour map; A straight line detection device, configured to perform straight line detection based on Hough transform on the edge line, and determine a straight line part in the edge line; and The information output device is used to determine the two end points of the straight line part by using the nearest neighbor search algorithm, and measure the distance between the two end points by using the Euclidean distance in the coordinate system, and obtain and output the straight line of the ball stud size. 9. A computer device, characterized in that it comprises a memory and a processor, wherein a computer program is stored in the memory, and when the computer program is executed by the processor, the processor executes the process according to claims 1 to 7. The steps of the ball stud size detection method according to any one of the claims. 10. A computer-readable storage medium, wherein a computer program is stored on the computer-readable storage medium, and when the computer program is executed by a processor, the processor is made to perform any of claims 1-7. The steps of the ball stud size detection method described in one claim.

Images