A quality inspection device and method for ultrasonic equipment data line

Abstract

The present invention relates to the field of image processing technology, and in particular to a quality inspection device and method for ultrasonic equipment data lines, comprising: performing a first edge detection on image data to be detected to determine a suspected crack area; obtaining pixel points in the suspected crack area to determine the inhibitory results of the pixel points in the edge detection inhibition process; using the inhibitory results, performing a second edge detection on the suspected crack area to determine the crack edge area of ​​the ultrasonic equipment data line. That is, the present invention improves the existing edge detection algorithm to accurately distinguish between the reflective area and the crack edge area, thereby improving the accuracy of detection.

Description

Technical Field

[0001] The present invention relates to the technical field of image processing, and in particular to a quality inspection device and method for a data line of an ultrasonic device. Background Art

[0002] Charging cables for ultrasonic equipment often suffer from damaged outer skin, shell shedding, loose braids, broken wires, cracks, and splits due to prolonged use or improper use. These damages may expose the internal wires, which may cause fires when in contact with other metal parts. Therefore, it is crucial to detect appearance defects of charging cables in a timely and effective manner.

[0003] When processing data line images, existing edge detection algorithms have difficulty in accurately distinguishing between reflective areas and crack edge areas due to the suppression of non-maximum pixels. This limitation reduces the accuracy of detection. Summary of the invention

[0004] In order to solve the technical problem that when processing data line images, the existing edge detection algorithm is difficult to accurately distinguish between the reflective area and the crack edge area due to the suppression of non-maximum pixels, resulting in low detection accuracy, the purpose of the present invention is to provide a quality inspection device and method for the data line of an ultrasonic device, which improves the existing edge detection algorithm to accurately distinguish between the reflective area and the crack edge area and improve the detection accuracy.

[0005] The technical solution adopted is specifically as follows: a quality inspection method for the data line of an ultrasonic device is provided, comprising: performing a first edge detection on the image data to be inspected to determine the suspected crack area; obtaining pixel points in the suspected crack area to determine the inhibition results of the pixel points in the edge detection inhibition process; using the inhibition results, performing a second edge detection on the suspected crack area to determine the crack edge area of ​​the data line of the ultrasonic device.

[0006] In one embodiment of the present invention, the first edge detection is performed on the image data to be detected to determine the suspected crack area, including: acquiring the image data to be detected, and performing the first edge detection on the image data to determine the edge detection image; determining the suspected crack area from the edge detection image.

[0007] In one embodiment of the present invention, the step of obtaining pixel points in the suspected crack area and determining the inhibitory results of the pixel points in the edge detection inhibition process includes: obtaining edge features of the suspected crack area, and using the edge features to determine the suspected crack edge area in the suspected crack area; obtaining protective data and assignment data of the pixel points in the suspected crack edge area, and then using the protective data and the assignment data to determine the inhibitory data corresponding to the suspected crack edge area.

[0008] In one embodiment of the present invention, it also includes: obtaining a data line area in the edge detection image and a first edge pixel point sequence corresponding to the data line area; obtaining a first proportion of the target pixel point in the sliding window to the data line width, and obtaining a second proportion of the target pixel point in the sliding window; wherein the sliding window is established along the extension direction of the data line in the data line area; and using the first proportion and the second proportion, determining the positional relationship of the suspected crack area in the edge detection image.

[0009] In one embodiment of the present invention, the step of obtaining edge features of the suspected crack area and determining the suspected crack edge area in the suspected crack area using the edge features includes: obtaining a first edge feature, a second edge feature, and a third edge feature corresponding to the suspected crack area, wherein the edge features include the first edge feature, the second edge feature, and the third edge feature; and determining the suspected crack edge area in the suspected crack area using the first edge feature, the second edge feature, and the third edge feature.

[0010] In one embodiment of the present invention, a target pixel point and a second edge pixel point sequence in the suspected crack area are obtained, and a first distance from each of the target pixel points to the second edge pixel point sequence is determined, wherein the first distance is the first edge feature; the first suspected crack area is determined using the first distance; multiple first pixel point sequences of the first suspected crack area are obtained, and vector angles corresponding to adjacent first pixel point sequences are obtained, and then the sum of changes in all the vector angles is determined, wherein the sum of changes is the second edge feature; the second suspected crack area is determined using the sum of changes; multiple second pixel point sequences of the second suspected crack area are obtained, and the suspected crack edge area in the suspected crack area is determined using the distance change value between the second pixel point sequences, wherein the distance change value is the third edge feature.

[0011] In one embodiment of the present invention, the method of obtaining protective data and assignment data of pixel points in the suspected crack edge area includes: obtaining the sum of changes in all vector angles of the first pixel sequence, and obtaining distance change values ​​between each pair of the second pixel sequence; determining the protective data of the pixel points in the suspected crack edge area using the sum of changes and the distance change values; obtaining a first number of target pixel points, and obtaining a gradient value ranking of the target pixel points; and determining the assignment data of the pixel points using the first number and the gradient value ranking.

[0012] In one embodiment of the present invention, the use of the inhibitory result to perform a second edge detection on the suspected crack area to determine the crack edge area of ​​the ultrasound equipment data line includes: obtaining the inhibitory results of all pixel points of the suspected crack area; comparing the inhibitory result with the inhibitory threshold to screen the crack edge area of ​​the ultrasound equipment data line.

[0013] In one embodiment of the present invention, the method of comparing the inhibitory result with the inhibition threshold to screen the crack edge area of ​​the ultrasound device data line includes: taking the area corresponding to the pixel points smaller than the inhibition threshold as the crack edge area of ​​the ultrasound device data line.

[0014] To solve the above technical problems, another technical solution adopted in the present application is: a quality inspection device for an ultrasonic equipment data line, comprising: a first detection module, used to perform a first edge detection on the image data to be detected, and determine the suspected crack area; a determination module, used to obtain pixel points in the suspected crack area, and determine the inhibitory results of the pixel points in the edge detection inhibition process; a second detection module, using the inhibitory results, performs a second edge detection on the suspected crack area, and determines the crack edge area of ​​the ultrasonic equipment data line.

[0015] The beneficial effects of the present invention are as follows: A quality inspection method for an ultrasonic device data line in the present invention includes: performing a first edge detection on the image data to be detected to determine a suspected crack area; obtaining pixel points in the suspected crack area to determine the inhibitory results of the pixel points in the edge detection suppression process; using the inhibitory results, performing a second edge detection on the suspected crack area to determine the crack edge area of ​​the ultrasonic device data line; that is, the present application processes the image data to be detected through the first edge detection to determine the suspected crack area; then, analyzing the pixel points in the suspected crack area to determine the inhibitory results of the pixel points in the suspected crack area in the edge detection suppression process to ensure that key crack information is not missed; finally, using these inhibitory results to perform a second edge detection to accurately distinguish between the reflective area and the crack edge area, identify the crack edge area of ​​the ultrasonic device data line, and thereby improve the accuracy of the detection to ensure the safety and reliability of the equipment. BRIEF DESCRIPTION OF THE DRAWINGS

[0016] In order to more clearly illustrate the technical solutions and advantages in the embodiments of the present invention or the prior art, the drawings required for use in the embodiments or the prior art descriptions are briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0017] Figure 1It is a structural schematic diagram of the quality inspection method of the ultrasonic equipment data line provided by the present invention;

[0018] Figure 2 It is a structural schematic diagram of the quality inspection device for the data line of the ultrasonic equipment provided by the present invention. DETAILED DESCRIPTION

[0019] In order to further explain the technical means and effects adopted by the present invention to achieve the predetermined invention purpose, the following is a detailed description of the quality inspection device and method of the data line of an ultrasonic device proposed by the present invention, its specific implementation, structure, characteristics and effects, in combination with the accompanying drawings and preferred embodiments. In the following description, different "one embodiment" or "another embodiment" does not necessarily refer to the same embodiment. In addition, specific features, structures or characteristics in one or more embodiments may be combined in any suitable form.

[0020] Unless defined otherwise, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention belongs.

[0021] The specific scheme of the quality inspection device and method for the data line of an ultrasonic device provided by the present invention is described in detail below with reference to the accompanying drawings.

[0022] See also Figure 1 , which shows a structural schematic diagram of the quality inspection method for the ultrasound equipment data line provided by the present invention.

[0023] like Figure 1 As shown, the quality inspection method of the ultrasonic equipment data line includes the following steps:

[0024] S11, performing a first edge detection on the image data to be detected to determine a suspected crack area.

[0025] Among them, edge detection algorithms, such as Canny or Sobel, can effectively extract edge information in the image. Cracks usually appear as slender and irregular edges. The image to be detected includes the data cable plug, the data cable body, and the area where the data cable casing has fallen off.

[0026] The suspected crack area is generally composed of a reflective area and a crack area. The reflective area is the light reflection formed by the light directly hitting a certain part of the data cable shell. Since the surface of the data cable is curved, only a small part in the middle of the curve is perpendicular to the incident light. Therefore, this part of the light will be reflected and captured by the camera, so the observed reflective area appears as a relatively centered linear or strip-shaped area. In addition, the cracks in the suspected crack area are mostly initial cracks in the data cable, usually in the shape of strips or lines, and the distribution direction is often consistent with the reflective area. In addition, compared with the completely detached area, the suspected crack area often appears in a strip shape on the initial edge detection image, while the completely detached area exposes copper wires, which are clearly distributed throughout the detached area. Therefore, the suspected crack area can be effectively located through the characteristics of the reflective area and the crack area.

[0027] Specifically, first, the original color image of the data line to be detected is obtained through a photographic device or sensor, and converted into a grayscale image for subsequent computer processing. Then, Gaussian blur is used to remove image noise to reduce the impact of pseudo edges on the detection results. Next, edge detection algorithms such as Canny or Sobel are used to extract edge information in the image, because cracks usually appear as slender and irregular edges. The suspected crack area is generally distributed between the two parallel edges of the data line, often in the form of strips or lines, and its distribution position is consistent with the direction of the reflective area. In the image of the initial edge detection, the suspected crack area usually shows a strip shape, while the completely detached area in contrast appears as exposed copper wire, and the distribution is more significant. Finally, using these features, the suspected crack area can be effectively located. That is, by combining the features of the suspected crack and the completely detached area, the accuracy and reliability of crack detection are improved.

[0028] To further locate the suspected crack area, the edge pixel sequence that meets the conditions is extracted from the data line image. Due to the shedding of the data line shell, the edge pixel points are widely distributed in the width range of the data line and are mainly concentrated in the middle area of ​​the data line; while the suspected crack area is distributed in the central area of ​​the data line in a strip shape. Through the characteristic differences in the edge pixel distribution of these two areas, the suspected crack area can be effectively identified and distinguished.

[0029] S12. Obtain pixel points in the suspected crack area, and determine the suppression results of the pixel points in the edge detection suppression process.

[0030] The pixels in the suspected crack region refer to the pixels that are identified as containing crack features in the image. In these regions, the grayscale gradient of each pixel is calculated to evaluate its inhibitory properties in the edge detection suppression process.

[0031] Specifically, the pixels in the suspected crack area are obtained, and the inhibition results of these pixels in the edge detection inhibition process are determined. First, the suspected crack area is analyzed to identify pixels with slender and irregular edge features. Subsequently, protective values ​​are assigned to these pixels according to the characteristics of the crack edge to ensure that key pixels are not misjudged or suppressed. Next, the grayscale gradient of each pixel, that is, the brightness change rate, is combined to calculate the size of the inhibition it receives during the edge detection process. Finally, the inhibition result of each pixel in the inhibition process is determined based on the grayscale gradient and crack characteristics, so as to accurately distinguish the crack edge from other areas. That is, in the determined suspected crack area, relevant pixels are extracted, and the inhibition results of these pixels in the edge detection inhibition process are analyzed. This process includes assigning protective values ​​to the crack edge and evaluating the degree of inhibition of each pixel in combination with the grayscale gradient to ensure that important crack information is not mis-suppressed.

[0032] S13. Using the suppression result, perform a second edge detection on the suspected crack area to determine the crack edge area of ​​the ultrasonic equipment data line.

[0033] Among them, by utilizing the inhibitory results, adjusting the parameters of the edge detection algorithm, and performing a second edge detection on the suspected crack area, the crack edge area in the data line of the ultrasonic equipment can be more accurately identified and confirmed, thereby improving the accuracy and reliability of the detection.

[0034] In this embodiment, the image data to be detected is processed through the first edge detection to determine the suspected crack area; then, the pixels in the suspected crack area are analyzed to determine the inhibition results of the pixels in the suspected crack area during the edge detection inhibition process to ensure that key crack information is not missed; finally, these inhibition results are used to perform a second edge detection to accurately identify the crack edge area of ​​the ultrasonic equipment data line to ensure the safety and reliability of the equipment.

[0035] The data line has a curved surface appearance. The light perpendicular to the data line surface is only approximately perpendicular to the data line surface in the middle area of ​​the curved surface. Then the area will form reflected light, and then form a reflective area. Therefore, the reflective area is often distributed in the central area of ​​the data line in the shape of lines or strips.

[0036] Data line cracks are usually divided into transverse cracks and longitudinal cracks. Transverse cracks are mainly formed due to excessive bending of the data line, which is usually perpendicular to the extension direction of the data line and is easier to distinguish; while longitudinal cracks are mostly caused by the friction between the multiple copper wires inside the data line and the outer shell, which is not easy to accurately detect and completely extract by naked eyes or existing algorithms. The reflective area is distributed in the center of the data line and has a higher grayscale value in the grayscale image due to reflection, while the crack area is randomly distributed and has a darker grayscale value due to the tortuous absorption of light by the crack area. The grayscale value size and position area are used to make a preliminary distinction between the two. However, the existing edge detection algorithm is not sufficient for the extraction of longitudinal cracks.

[0037] In some embodiments, S11 includes:

[0038] S111 . Acquire image data to be detected, and perform a first edge detection on the image data to determine an edge detection image.

[0039] The image data to be detected can be obtained by a photographic device or sensor, usually a color image. A variety of algorithms can be used for the first edge detection, such as the Sobel operator or the Canny edge detection. The edge detection image is a binary image generated by the edge detection algorithm, in which the edge position is marked as white and the rest of the area is black.

[0040] The suspected crack area of ​​the data line is distributed in the data line area, that is, in the central area between two parallel edges in the initial edge detection image. Usually, the target pixel points extracted from the data line shell peeling area in the middle area of ​​the data line edge pixel sequence are relatively wide in the data line width range and are almost completely distributed in the middle area of ​​the data line. The suspected crack area is distributed in the center of the data line in a strip shape.

[0041] Specifically, the suspected crack area is located through the following steps to distinguish the suspected crack area in the data line area.

[0042] First, a data line region in an edge detection image and a first edge pixel point sequence corresponding to the data line region are obtained.

[0043] The data line area refers to a specific part of the image containing the data line, which is usually represented by two parallel edges, corresponding to the left and right or upper and lower boundaries of the data line. In the edge detection image, the outline of the data line area is usually a set of relatively parallel and continuous edge lines, and the spacing (i.e., the data line width) remains relatively consistent across the entire data line. The first edge pixel point sequence refers to an ordered set of edge pixel points of one of the two parallel edges in the data line area.

[0044] Secondly, a first proportion of the target pixel point in the sliding window on the data line width is obtained, and a second proportion of the target pixel point in the sliding window is obtained; wherein the sliding window is established along the extension direction of the data line in the data line area.

[0045] The sliding window is a small rectangular area used for local analysis. The sliding window moves in the image along the specified direction, with a fixed step size each time. The data covered by the sliding window each time is analyzed as an independent area, thereby achieving a step-by-step, local scan of the entire data line area.

[0046] Next, the first ratio and the second ratio are used to determine the positional relationship of the suspected crack area in the edge detection image.

[0047] The following is a possibility formula for determining whether the area of ​​the sliding window is a suspected crack area based on the first proportion and the second proportion, and then determining the position relationship of the suspected crack area in the edge detection image:

[0048]

[0049] In the formula, Indicates the first proportion of the target pixel in the sliding window on the data line width, Indicates the second proportion of the target pixel in the sliding window. It represents the product of the first proportion and the second proportion. The larger the value is, the greater the possibility that the area is completely detached from the data cable housing. On the contrary, The smaller the value, the The larger the value is, the greater the possibility that the area is completely detached from the data cable housing, that is, the smaller the possibility that the area is a suspected crack area. c indicates whether the area of ​​the sliding window is a suspected crack area. The threshold is set to an empirical value of 0.8. If it is greater than 0.8, the area in the sliding window is considered to be a suspected crack area.

[0050] S112, determining a suspected crack area from the edge detection image.

[0051] Specifically, all sliding windows of suspected crack areas are extracted and spliced ​​to determine the suspected crack areas.

[0052] In this embodiment, the image data to be detected is obtained, and the first edge detection is performed on the image data to determine the edge detection image, and the suspected crack area is located based on the edge detection image to determine the suspected crack area.

[0053] The existing edge detection algorithm uses the non-maximum suppression method to extract the edge of the image region, that is, only retaining the pixel point with the largest gradient value in its gradient direction, and suppressing other gradient pixels. This method only judges whether the pixel point is the edge of the region from the perspective of gradient, and cannot completely distinguish different edge regions with similar grayscale value differences.

[0054] In addition, data line cracks are divided into transverse cracks and longitudinal cracks. Transverse cracks are often formed due to excessive bending. The cracks are usually perpendicular to the extension direction of the data line and are relatively easy to distinguish. Longitudinal cracks are mostly formed due to the continuous friction between the multiple copper wires distributed inside the data line and the data line shell. This type of crack is not easy to distinguish with the naked eye, and it is not easy to accurately and completely extract it in the edge detection image of the existing algorithm.

[0055] In order to improve the detection accuracy, the present application can locate the suspected crack area based on the grayscale image and use the following step S12 to distinguish the crack area from the reflective area to prevent the reflective area from interfering with the crack detection result.

[0056] The reflective area is distributed in the center of the data line and has a higher grayscale value in the grayscale image due to the reflection, while the crack area is randomly distributed and has a darker grayscale value due to the crack area bending and absorbing light. Step S12 makes a preliminary distinction between the two using the grayscale value size and the position area.

[0057] In some embodiments, S12, obtaining pixel points in the suspected crack area, and determining the suppression result of the pixel points in the edge detection suppression process; includes:

[0058] S121. Obtain edge features of the suspected crack region, and determine the suspected crack edge region in the suspected crack region using the edge features.

[0059] The edge features of the suspected crack area include a first edge feature, a second edge feature and a third edge feature.

[0060] Obtain edge features of the suspected crack region, and use the edge features to determine the suspected crack edge region in the suspected crack region. The specific steps are as follows:

[0061] First, the first edge feature, the second edge feature, and the third edge feature corresponding to the suspected crack area are obtained. Specifically, the target pixel point and the second edge pixel point sequence are obtained, and the first distance from each target pixel point to the second edge pixel point sequence is calculated. This distance is the first edge feature. Using this distance, the first suspected crack area can be determined.

[0062] Specifically, first, based on the suspected crack area extracted above, it is located in the grayscale image, and all the suspected crack areas located in the grayscale image are extracted. The grayscale value median value that distinguishes the bright area and the crack area is set to d, and all pixels with grayscale values ​​greater than d are extracted. Next, based on the data line width parameter e, the reflective area and the crack area are distinguished by using the reflective area located in the center of the data line. Specifically, since the appearance of the data line presents a curved surface, the light perpendicular to the surface of the data line is only approximately perpendicular to the surface of the data line in the middle area of ​​the curved surface, then the area will form a reflected light, and then form a reflective area, so that the reflective area is often distributed in the central area of ​​the data line in a linear or strip shape. Therefore, for each pixel with a grayscale value greater than d, the distance D to the edges on both sides of the data line is calculated, and the distance difference F between the two is determined. If the distance from a certain pixel to the edges on both sides is close, it is considered to be located in the center of the data line. To ensure accuracy, an empirical threshold of 0.05 mm was used to screen out all pixel areas with F>0.05 mm, mark them as the first suspected crack area, and conduct further detection based on this.

[0063] In addition, the crack edge area extracted by edge detection is usually composed of two zigzag line segments, which often appear to be intersecting or about to intersect. The pixel directions of these zigzag line segments vary greatly, which is manifested as a significant change in the angle between the vectors composed of edge pixels. At the same time, the distance between the two line segments gradually increases or decreases along the crack direction. This change trend indicates the existence of a crack area. In particular, when the start-end distance of the two line segments changes greatly, it is more likely to be a crack area. The following is to locate the crack area by analyzing the relationship between the change in the angle of the edge pixel sequence of the suspected crack edge area and the change in its slope.

[0064] A plurality of first pixel point sequences in the first suspected crack region are obtained, and the vector angles between adjacent first pixel point sequences are calculated, and then the sum of the changes of all vector angles is determined, and the sum of the changes is the second edge feature. Based on this feature, the second suspected crack region is further determined. Based on this, the second suspected crack region is determined.

[0065] The greater the sum of the changes in all vector angles is, the more tortuous the first pixel sequence is, and the greater the possibility that the region becomes a crack edge region is.

[0066] Specifically, a first pixel point sequence of a first suspected crack area is obtained, based on any extracted first pixel point sequence To analyze, get the first pixel sequence The direction vector between two adjacent pixels in the image is: the magnitude of the direction vector between two adjacent pixels is the absolute value of the difference between the grayscale values ​​of the two pixels, and the direction of the direction vector is the direction from the previous pixel to the next pixel in the two adjacent pixels. Then calculate the angle between the two adjacent direction vectors ; Then, the formula is used to express the possibility that the edge pixel sequence is the crack edge area:

[0067]

[0068] In this formula, Represents the first pixel sequence The angle between the direction vector of the i-th pixel and the direction vector of the i+1-th pixel, n represents the first pixel sequence The number of angles between all two adjacent direction vectors in , is the probability that the edge pixel sequence is the crack edge area. The larger the value, the greater the possibility that the area is the crack edge area. Used to reflect the total change of all vector angles in the entire sequence.

[0069] in, The threshold value can be set to 0.8 as an empirical value. The area where the value is greater than 0.8 is the second suspected crack area.

[0070] Finally, a plurality of second pixel point sequences in the second suspected crack region are obtained, and the suspected crack edge region in the suspected crack region is identified through distance change values ​​between each pair, wherein the distance change value is the third edge feature.

[0071] Specifically, first, extract multiple second pixel point sequences in the second suspected crack area, and pair these sequences in pairs. For each pair of paired second edge pixel point sequences, calculate multiple distance values ​​at equal intervals along their extension direction, and obtain the difference between the maximum and minimum values ​​of these distance values, which is recorded as the distance change h. This distance change h is used as the third edge feature to further determine the possibility of cracks. The larger the h value, the greater the possibility that the area where the two second edge pixel point sequences are located is the crack edge area. Represents the normalized value of the second edge pixel sequence. The larger the value is, the greater the possibility that the area where the two second edge pixel sequences are located is the crack edge area.

[0072] S122, obtaining protective data and assignment data of pixel points in the suspected crack edge area, and then using the protective data and assignment data to determine the inhibitory data corresponding to the suspected crack edge area.

[0073] The protective data is used to indicate the importance of certain pixels in edge detection, while the assignment data is used to weight these pixels.

[0074] Specifically, obtaining protective data and value assignment data of pixel points in the suspected crack edge area specifically includes the following steps:

[0075] First, the sum of the changes in all vector angles of the first pixel point sequence is obtained, and the distance change values ​​between each pair of the second pixel point sequences are obtained.

[0076] The detailed steps of obtaining the sum of changes in all vector angles of the first pixel point sequence and obtaining the distance change values ​​between each pair of the second pixel point sequences have been described in detail above and will not be repeated here.

[0077] Next, the protective data of the pixel points in the suspected crack edge area are determined using the total change amount and the distance change value.

[0078] Among them, the above has described in detail how to use the total change amount and the distance change value to determine the suspected crack edge area, which will not be repeated here.

[0079] In addition, the greater the possibility that a pixel point in the suspected crack area is the crack edge area, the greater the protection of the point when it is suppressed by the edge detection algorithm. When judging whether to suppress the edge detection of the point, it is necessary to comprehensively consider the grayscale value gradient and the protective size to obtain the inhibition size of the point, and then use the edge detection algorithm to extract the crack edge area in the suspected crack area.

[0080] Specifically, the formula for determining the protective size of the pixel points in the suspected crack area based on the total change amount and the distance change value is as follows:

[0081]

[0082] in, Represents the normalized value of the sum of changes in all vector angles of the first pixel sequence, represents the normalized value of the maximum and minimum distance difference between the edge pixel sequence where the second pixel is located and the second edge pixel sequence that matches it, along the extension direction of these two edge pixel sequences, It represents the product of the normalized differences of the four features, and is used to indicate the protectiveness of the pixel in the suspected crack edge area.

[0083] Next, a first number of target pixel points is obtained, and a gradient value ranking of the target pixel points is obtained.

[0084] Specifically, when performing edge detection on the suspected crack edge area, the grayscale change assignment size P is obtained for N pixels along the direction of the grayscale gradient change of the pixel points according to their grayscale change degree, that is, the grayscale gradient changes are sorted according to size and assigned from large to small.

[0085]

[0086] In the formula, N represents the first number of target pixel points that meet the gradient condition in the gradient change of the target pixel point, and M is the order of the gradient values ​​of these target pixel points from large to small. Represents the difference between the values ​​assigned to different gradient target pixels. Represents the difference between the value assigned to the Mth target pixel and 1. Indicates the value of the target pixel. The larger the value, the greater the gradient change of the target pixel.

[0087] Finally, the first quantity and the gradient value sorting are used to determine the value assignment data of the pixel points.

[0088] Specifically, by combining a pixel point with protection and gradient change, we can get the size of the inhibition when performing edge detection on it. The formula is as follows:

[0089]

[0090] In this formula, Indicates the protective size of the pixel points in the suspected crack area. It indicates the value assigned to the pixel in the area combined with the gray value gradient change. Indicates the suppression level of a pixel in the suspected crack edge area during edge detection. The greater the possibility that the pixel is the crack edge area, the greater its suppression. It indicates the suppression degree of the pixels in this area during edge detection. The smaller the value, the greater the suppression degree.

[0091] In this embodiment, three edge features of the suspected crack area, namely the first, second and third edge features, are firstly extracted in the first step, thereby effectively determining the suspected crack edge area. Subsequently, in the second step, the protective data and assignment data of the pixel points in the suspected crack edge area are obtained to further enhance the recognition and processing of important pixels, so as to effectively assign weights to the pixel points, thereby achieving more accurate crack recognition in the edge detection process.

[0092] The following embodiments further locate the pixels of the suspected crack edge area based on the feature differences between the crack area and the reflective area in the initial edge detection image, as a basis for improving the edge detection algorithm. A new consistency judgment standard is proposed based on the original edge detection algorithm's suppression method based on the size of the grayscale gradient change of the pixel point. This step will perform edge detection on the data line grayscale image again based on the new standard.

[0093] S13. Using the suppression result, a second edge detection is performed on the suspected crack area to determine the crack edge area of ​​the ultrasonic equipment data line. The specific steps are as follows:

[0094] First, the suppression results of all pixels in the suspected crack area are obtained.

[0095] Secondly, the inhibition result is compared with the inhibition threshold, and the area corresponding to the pixel points smaller than the inhibition threshold is used as the crack edge area of ​​the ultrasonic device data line, thereby determining the edge detection image of the ultrasonic device data line.

[0096] Specifically, by using the suppression results to perform a second edge detection on the suspected crack area, the crack edge identification process is further refined. First, the suppression results of all pixels in the suspected crack area are extracted. Then, the suppression results of these pixels are compared with the set empirical threshold (0.7). If the r value of the suppression result of a certain pixel is greater than 0.7, it is considered that it does not belong to the crack edge and is suppressed; if the r value is less than 0.7, it is marked as a crack edge area. If the r value is equal to 0.7, it is not directly marked as a crack edge area. The r value of the surrounding pixels can be considered. If the r value of most of the surrounding pixels is less than 0.7, the pixel can be marked as a crack edge, otherwise it is marked as not belonging to the crack edge.

[0097] In this embodiment, the edge detection image generated according to the new evaluation criteria can clearly present the crack edge area of ​​the data line of the ultrasonic device, thereby achieving more accurate crack identification and edge detection.

[0098] See also Figure 2 , Figure 2 It is a structural schematic diagram of the quality inspection device for the data line of the ultrasonic device provided by the present invention. The quality inspection device can execute the steps in the quality inspection method for the data line of the ultrasonic device described above. The quality inspection device 300 includes: a first detection module 310, which is used to perform a first edge detection on the image data to be detected and determine the suspected crack area; a determination module 320, which is used to obtain the pixel points in the suspected crack area and determine the inhibitory results of the pixel points in the edge detection inhibition process; a second detection module 330, which uses the inhibitory results to perform a second edge detection on the suspected crack area and determine the crack edge area of ​​the data line of the ultrasonic device.

[0099] It should be noted that the sequence of the above embodiments of the present invention is only for description and does not represent the advantages and disadvantages of the embodiments. The processes depicted in the accompanying drawings do not necessarily require the specific order or continuous order shown to achieve the desired results. In some embodiments, multitasking and parallel processing are also possible or may be advantageous.

[0100] The various embodiments in this specification are described in a progressive manner, and the same or similar parts between the various embodiments can be referenced to each other, and each embodiment focuses on the differences from other embodiments.

Claims

1. A quality inspection method for an ultrasonic device data line, characterized in that: The quality inspection method of the ultrasonic equipment data line comprises: Perform the first edge detection on the image data to be detected to determine the suspected crack area; Acquire pixel points in the suspected crack area, and determine the suppression results of the pixel points in the edge detection suppression process; Using the suppression result, performing a second edge detection on the suspected crack area to determine the crack edge area of ​​the ultrasonic device data line; The obtaining of pixel points in the suspected crack area and determining the suppression result of the pixel points in the edge detection suppression process includes: Acquire edge features of the suspected crack region, and determine a suspected crack edge region in the suspected crack region using the edge features; Acquire protective data and assignment data of pixel points in the suspected crack edge region, and then determine the inhibition data corresponding to the suspected crack edge region using the protective data and the assignment data; The first edge detection is performed on the image data to be detected to determine the suspected crack area, including: Acquire image data to be detected, and perform a first edge detection on the image data to determine an edge detection image; Determining the suspected crack area from the edge detection image; Also includes: Acquire a data line region in the edge detection image and a first edge pixel point sequence corresponding to the data line region; Obtaining a first proportion of a target pixel point in a sliding window on a data line width, and obtaining a second proportion of a target pixel point in the sliding window in the sliding window; wherein the sliding window is established along an extension direction of the data line in the data line area; Determine the positional relationship of the suspected crack area in the edge detection image by using the first proportion and the second proportion; The obtaining edge features of the suspected crack region and determining the suspected crack edge region in the suspected crack region by using the edge features includes: Acquire a first edge feature, a second edge feature, and a third edge feature corresponding to the suspected crack region, wherein the edge features include the first edge feature, the second edge feature, and the third edge feature; Determine the suspected crack edge region in the suspected crack region by using the first edge feature, the second edge feature and the third edge feature; Acquire a target pixel point and a second edge pixel point sequence in the suspected crack area, and determine a first distance from each of the target pixel points to the second edge pixel point sequence, wherein the first distance is the first edge feature; Determine a first suspected crack region using the first distance; Acquire multiple first pixel point sequences of the first suspected crack area, and acquire vector angles corresponding to adjacent first pixel point sequences, and then determine the sum of changes of all the vector angles, wherein the sum of changes is the second edge feature; Determine a second suspected crack region using the sum of the changes; Acquire multiple second pixel point sequences of the second suspected crack area, and determine the suspected crack edge area in the suspected crack area according to the distance change value between the second pixel point sequences in pairs, wherein the distance change value is a third edge feature; The step of obtaining protective data and value assignment data of pixel points in the suspected crack edge region includes: Obtaining the sum of changes in all the vector angles of the first pixel point sequence, and obtaining distance change values ​​between any two of the second pixel point sequences; Determine the protective data of the pixel points in the suspected crack edge area by using the total amount of changes and the distance change value; Acquire a first number of the target pixel points, and acquire a gradient value ranking of the target pixel points; The first number and the gradient value sorting are used to determine the value assignment data of the pixel point.

2. The quality inspection method for ultrasonic equipment data line according to claim 1, characterized in that: The using the suppression result to perform a second edge detection on the suspected crack area to determine the crack edge area of ​​the ultrasonic device data line includes: Obtaining suppression results for all pixels in the suspected crack area; The suppression result is compared with the suppression threshold to screen the crack edge area of ​​the ultrasonic equipment data line.

3. The quality inspection method for ultrasonic equipment data line according to claim 2, characterized in that: The method of comparing the inhibitory result with the inhibitory threshold to screen the crack edge area of ​​the ultrasonic device data line includes: taking the area corresponding to the pixel points smaller than the inhibitory threshold as the crack edge area of ​​the ultrasonic device data line.

4. A quality inspection device for an ultrasonic equipment data line, characterized in that: include: A first detection module is used to perform a first edge detection on the image data to be detected to determine a suspected crack area; A determination module, used for acquiring pixel points in the suspected crack area and determining the inhibition result of the pixel points in the edge detection inhibition process; A second detection module, using the suppression result, performs a second edge detection on the suspected crack area to determine the crack edge area of ​​the ultrasonic device data line; The obtaining of pixel points in the suspected crack area and determining the suppression result of the pixel points in the edge detection suppression process includes: Acquire edge features of the suspected crack region, and determine a suspected crack edge region in the suspected crack region using the edge features; Acquire protective data and assignment data of pixel points in the suspected crack edge region, and then determine the inhibition data corresponding to the suspected crack edge region using the protective data and the assignment data; The first edge detection is performed on the image data to be detected to determine the suspected crack area, including: Acquire image data to be detected, and perform a first edge detection on the image data to determine an edge detection image; Determining the suspected crack area from the edge detection image; Also includes: Acquire a data line region in the edge detection image and a first edge pixel point sequence corresponding to the data line region; Obtaining a first proportion of a target pixel point in a sliding window on a data line width, and obtaining a second proportion of a target pixel point in the sliding window in the sliding window; wherein the sliding window is established along an extension direction of the data line in the data line area; Determine the positional relationship of the suspected crack area in the edge detection image by using the first proportion and the second proportion; The obtaining edge features of the suspected crack region and determining the suspected crack edge region in the suspected crack region by using the edge features includes: Acquire a first edge feature, a second edge feature, and a third edge feature corresponding to the suspected crack region, wherein the edge features include the first edge feature, the second edge feature, and the third edge feature; Determine the suspected crack edge region in the suspected crack region by using the first edge feature, the second edge feature and the third edge feature; Acquire a target pixel point and a second edge pixel point sequence in the suspected crack area, and determine a first distance from each of the target pixel points to the second edge pixel point sequence, wherein the first distance is the first edge feature; Determine a first suspected crack region using the first distance; Acquire multiple first pixel point sequences of the first suspected crack area, and acquire vector angles corresponding to adjacent first pixel point sequences, and then determine the sum of changes of all the vector angles, wherein the sum of changes is the second edge feature; Determine a second suspected crack region using the sum of the changes; Acquire multiple second pixel point sequences of the second suspected crack area, and determine the suspected crack edge area in the suspected crack area according to the distance change value between the second pixel point sequences in pairs, wherein the distance change value is a third edge feature; The step of obtaining protective data and value assignment data of pixel points in the suspected crack edge region includes: Obtaining the sum of changes in all the vector angles of the first pixel point sequence, and obtaining distance change values ​​between any two of the second pixel point sequences; Determine the protective data of the pixel points in the suspected crack edge area by using the total amount of changes and the distance change value; Acquire a first number of the target pixel points, and acquire a gradient value ranking of the target pixel points; The first number and the gradient value sorting are used to determine the value assignment data of the pixel point.

Images