Method, system, device and medium for detecting the surface shape of vehicle windshield

Abstract

The present invention relates to a surface detection method, system, device and medium for a vehicle windshield, and relates to the technical field of windshield detection, wherein the method comprises: collecting two initial RMS images of the same windshield surface; filtering and feature enhancing each initial RMS image to obtain two RMS images; performing color space conversion on the two RMS images; judging whether the two collected initial RMS images have external environmental interference based on the two RMS images after color space conversion, if not, obtaining the surface detection of the windshield according to the RMS images; if so, correcting the two RMS images with external environmental interference, and obtaining the surface detection result of the windshield according to the corrected RMS images. The present invention can reduce the influence of external environmental factors on the windshield surface detection process, improve the accuracy and reliability of the windshield surface detection result, and thus improve the yield rate of the windshield.

Description

Technical Field

[0001] The present invention relates to the technical field of windshield detection, and in particular to a surface detection method, system, equipment and medium for vehicle windshields. Background Art

[0002] As head-up display (HUD) systems become more and more common in automobiles, industry requirements and consumer expectations for the visual quality and performance of HUD systems are also increasing. However, the current mainstream HUD technologies all use the car windshield as the display carrier, and the shape of the windshield has a significant impact on the visual quality of the HUD projection.

[0003] The production of automobile windshields (front windshields) involves cutting, edging, cleaning, high-temperature continuous bending, etc. Each of these processing steps has a great impact on the surface shape of the windshield, causing the curvature to be too large or too small, which will affect the final HUD imaging effect. The curvature is often measured in time during the processing, and appropriate process adjustments are made to ensure that the curvature of the windshield meets the required standards.

[0004] Therefore, it is necessary to effectively detect the surface shape of the windshield, but the quality of the windshield surface shape is difficult to measure directly, which makes it particularly difficult to adjust the processing technology, resulting in the final processing of defective or even scrapped products.

[0005] In the prior art, the PV value and RMS value of the windshield measured by the existing structured light surface shape measurement device may be affected by various factors, resulting in inaccurate measurement results, including environmental conditions (such as air disturbance, environmental vibration, etc.), the accuracy of the measurement equipment, the accuracy of the measurement method, etc. It is necessary to design a new windshield surface shape measurement method to improve the accuracy and reliability of windshield surface shape detection. Summary of the invention

[0006] Therefore, the technical problem to be solved by the present invention is to overcome the problem in the prior art that the windshield surface shape detection is affected by various factors, resulting in inaccurate windshield surface shape detection measurement results.

[0007] In order to solve the above technical problems, the present invention provides a surface shape detection method for a vehicle windshield, comprising:

[0008] Step S1: collecting two initial RMS images of the same windshield surface;

[0009] Step S2: filtering and feature enhancement are performed on each initial RMS image to obtain two RMS images;

[0010] Step S3: performing color space conversion on the two RMS images to obtain two RMS images after color space conversion;

[0011] Step S4: Based on the two RMS images after color space conversion, determine whether the two initial RMS images collected have external environmental interference. If there is no external environmental interference, obtain the surface shape detection result of the windshield according to the RMS image; if there is external environmental interference, correct the two RMS images with external environmental interference to obtain corrected RMS images, and obtain the surface shape detection result of the windshield according to the corrected RMS images.

[0012] In one embodiment of the present invention, the method of performing color space conversion on the two RMS images in step S3 includes:

[0013] Convert the RMS image from RGB color space to HSV color space;

[0014] Under the HSV color space condition, the hue of each pixel in the RMS image is counted, and the pixels with hues between 100° and 140° are set to green, the pixels with hues between 180° and 240° are set to blue, and the pixels with hues between 0° and 50° are set to orange-red. The formula is:

[0015] ;

[0016] in, is green, is blue, Orange-red. For color tone.

[0017] In one embodiment of the present invention, the method for judging whether the two collected initial RMS images are interfered by the external environment based on the two RMS images after the color space conversion in step S4 includes:

[0018] For the two RMS images after color space conversion, the pixel areas where green, blue, and orange-red pixels are located are divided by counting the hues of the pixels in the two RMS images;

[0019] Get the edge pixel coordinates of the pixel areas where the green, blue, and orange-red pixels are located in the two RMS images;

[0020] According to the edge pixel coordinates of the pixel areas where the green, blue and orange-red pixels are located, the color edge curves of the respective colors are obtained;

[0021] Compare whether the two color edge curves corresponding to the same color pixels and the same pixel areas in the two RMS images are consistent. If the two color edge curves are inconsistent, it indicates that the initial RMS image obtained is interfered by the external environment; if the two color edge curves are consistent, it indicates that the initial RMS image obtained is not interfered by the external environment.

[0022] In one embodiment of the present invention, the method for correcting the two RMS images with external environmental interference in step S4 includes:

[0023] Determine whether there are the same coordinate points between the two color edge curves. If there are the same coordinate points, it indicates that there is an intersection, and the two color edge curves are corrected by the first optimization method; if there are no same coordinate points, it indicates that there is no intersection, and the two color edge curves are corrected by the second optimization method.

[0024] In one embodiment of the present invention, the first optimization method is specifically:

[0025] If the number of intersections is one, it is defined as ;

[0026] If the intersection If the two color edge curves on both sides are separated, then get the color at the intersection. The starting coordinates of the two color edge curves on one side and , and obtain the The end point coordinates of the two color edge curves on the other side and ,in, and are the starting point and end point of the first color edge curve, and are the starting point and end point of the second color edge curve; Calculate the new starting point coordinates ,pass Calculate the new end point coordinates ; Set the new starting point coordinates With this intersection Connect the curves to get the first curve segment and set the new end point coordinates With this intersection Connect the curves to obtain a second curve segment, and replace the original two color edge curves with a curve formed by reconnecting the first curve segment and the second curve segment;

[0027] If the intersection If only one of the two color edge curves on the two sides is separated, then get the color edge curve at the intersection. The starting coordinates of the two color edge curves on one side and ,pass Calculate the new starting point coordinates , the new starting point coordinates With this intersection Connect the curves to obtain curve segments, and replace the original two corresponding color edge curve segments with the curve segments.

[0028] In one embodiment of the present invention, the first optimization method is specifically:

[0029] If the number of intersections is two, it is defined as and , then count the two intersection points and The total number of pixels in the closed area formed by the two color edge curves;

[0030] If the total number of pixels is less than the preset number, the two intersection points are taken and Any color edge curve segment between them is used as the intersection point of the two and The color edge curve segment between ;

[0031] If the total number of pixels is greater than or equal to the preset number, calculate the two intersection points and The maximum Euclidean distance between the two color edge curves, take the midpoint of the line segment where the maximum Euclidean distance is located , the two intersection points and Respectively with the midpoint Connect to form a curve segment and use it as the two intersection points and The color edge curve segment between them replaces the two original corresponding color edge curve segments;

[0032] Finally, determine the two intersection points and Are the two color edge curves other than the two color edge curves separated? If not, the two intersection points and The two color edge curves other than are the same curve and do not need to be optimized; if they are separated, the two intersection points and The other two color edge curves perform steps corresponding to the situation that only the two color edge curves on one side of the intersection are in a separated state.

[0033] In one embodiment of the present invention, the first optimization method is specifically:

[0034] If the number of intersections is three, it is defined as , , , and the intersection Located in and between;

[0035] With three intersections , , Construct a circumscribed circle as the reference, and place the three intersection points , , The circumscribed circle segment is taken as the curve segment and used as the three intersection points , , The color edge curve segment is used to replace the original two corresponding color edge curve segments;

[0036] Finally, determine the two intersection points and Are the two color edge curves other than the two color edge curves separated? If not, the two intersection points and The two color edge curves other than are the same curve and do not need to be optimized; if they are separated, the two intersection points and The other two color edge curves perform steps corresponding to the situation that only the two color edge curves on one side of the intersection are in a separated state.

[0037] In one embodiment of the present invention, the method for correcting two color edge curves by the second optimization method includes:

[0038] Construct a curve to correct the two color edge curves. This curve is located between the two color edge curves, and count the number of pixels between this curve and the first color edge curve. , count the number of pixels between this curve and the second color edge curve ,and and satisfy: .

[0039] In one embodiment of the present invention, in step S1, a sinusoidal periodic fringe image of the same windshield surface is collected by a structured light surface measurement device, a plurality of sinusoidal periodic fringe images are phase-deconstructed to obtain an actual three-dimensional point cloud image of the windshield, the actual three-dimensional point cloud image and the theoretical three-dimensional point cloud image are subtracted to obtain a pseudo-color image with a height error, and two pseudo-color images with a height error are obtained as initial RMS images, wherein:

[0040] The structured light surface measurement device comprises:

[0041] A first windshield detection module includes a first image generator and a first CCD camera connected to a computer, wherein the first image generator is used to display first sinusoidal periodic fringes, first incident light formed by the first sinusoidal periodic fringes reaches one side of the windshield to form first reflected light, and the first CCD camera is used to collect the first reflected light and feed it back to the computer to obtain a surface detection result of one side of the windshield;

[0042] The second windshield detection module includes a second image generator and a second CCD camera connected to the computer, the second image generator is used to display second sinusoidal periodic stripes, the second incident light formed by the second sinusoidal periodic stripes reaches the other side of the windshield to form second reflected light, the second CCD camera is used to collect the second reflected light and feed it back to the computer to obtain the surface detection result of the other side of the windshield, wherein the first reflected light collected by the first CCD camera and the second reflected light collected by the second CCD camera are both sinusoidal periodic stripe images.

[0043] In order to solve the above technical problems, the present invention provides a surface shape detection system for a vehicle windshield, comprising:

[0044] Image extraction module: used to collect two initial RMS images of the same windshield surface;

[0045] Image enhancement module: used to filter and enhance the features of each initial RMS image to obtain two RMS images;

[0046] Color space conversion module: used to perform color space conversion on two RMS images to obtain two RMS images after color space conversion;

[0047] Judgment and correction module: used to judge whether the two initial RMS images collected have external environmental interference based on the two RMS images after color space conversion. If there is no external environmental interference, the surface detection result of the windshield is obtained according to the RMS image; if there is external environmental interference, the two RMS images with external environmental interference are corrected to obtain the corrected RMS images, and the surface detection result of the windshield is obtained according to the corrected RMS images.

[0048] To solve the above technical problems, the present invention provides a detection device, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor implements the steps of the above-mentioned surface shape detection method for a vehicle windshield when executing the computer program.

[0049] In order to solve the above technical problems, the present invention provides a computer-readable storage medium having a computer program stored thereon. When the computer program is executed by a processor, the steps of the above-mentioned method for detecting the surface shape of a vehicle windshield are implemented.

[0050] The above technical solution of the present invention has the following advantages compared with the prior art:

[0051] The surface shape detection method for vehicle windshields of the present invention fully considers that the detection of windshields by structured light surface shape measurement devices may be affected by factors such as environmental conditions, the precision of measurement equipment, and the accuracy of measurement methods, which may lead to inaccurate surface shape measurement results. The present invention starts with the RMS image and first converts the RMS image from the RGB color space to the HSV color space to enhance the perceived intensity and vividness of the color, which is beneficial to the subsequent color edge curve analysis between different color regions, thereby improving the accuracy and reliability of windshield surface shape detection and improving the windshield yield rate;

[0052] The present invention analyzes two color edge curves corresponding to the same color pixels and the same pixel area in two RMS images, constructs a new curve (or curve segment) to replace the curve (or curve segment) in the original two RMS images, obtains a corrected RMS image, and obtains a face detection result based on the corrected RMS image, thereby effectively eliminating the influence of external environmental factors;

[0053] The present invention also designs a structured light surface measurement device, which can not only perform non-contact surface detection on the concave and convex surfaces of the windshield, but also can accommodate any type of windshield, effectively saving detection costs and facilitating large-scale promotion. BRIEF DESCRIPTION OF THE DRAWINGS

[0054] In order to make the contents of the present invention more clearly understood, the present invention is further described in detail below based on specific embodiments of the present invention in conjunction with the accompanying drawings.

[0055] Figure 1 is a flow chart of the method of the present invention;

[0056] Figure 2 is a schematic diagram of an initial RMS image in an embodiment of the present invention;

[0057] FIG3 (a) is a schematic diagram showing a state in which two color edge curves have an intersection and both sides of the intersection are separated in an embodiment of the present invention;

[0058] FIG3 (b) is a schematic diagram showing a state in which two color edge curves have an intersection and one side of the intersection is in a separation state in an embodiment of the present invention;

[0059] FIG4 (a) is a schematic diagram showing a state in which two color edge curves have two intersection points and are in a separated state except for the two intersection points according to an embodiment of the present invention;

[0060] FIG4( b ) is a schematic diagram showing a state in which two color edge curves have two intersection points and are in a non-separated state except for the two intersection points according to an embodiment of the present invention;

[0061] FIG5 (a) is a schematic diagram showing a state in which two color edge curves have three intersections and are in a separated state except for the three intersections in an embodiment of the present invention;

[0062] FIG5( b ) is a schematic diagram of a state in which two color edge curves have three intersections and all the other three intersections except the intersections located on both sides are in a non-separated state in an embodiment of the present invention;

[0063] Figure 6 is a schematic diagram of two color edge curves having no intersection in an embodiment of the present invention;

[0064] Figure 7 is a schematic diagram of the structure of a structured light surface measurement device in an embodiment of the present invention;

[0065] Figure 8 is a side view of a structured light surface type measuring device in an embodiment of the present invention;

[0066] Fig. 9 is a schematic diagram of the first windshield detection module or the second windshield detection module performing surface detection on a windshield in an embodiment of the present invention;

[0067] Fig.10 1 is a schematic diagram of the structures of a first windshield detection module and a second windshield detection module for a large-caliber windshield in an embodiment of the present invention;

[0068] Fig.11 is a schematic diagram of the structure of a motion module in an embodiment of the present invention;

[0069] Fig.12 Schematic diagram of the structure of the inspection tool platform in the embodiment of the present invention. DETAILED DESCRIPTION

[0070] The present invention is further described below in conjunction with the accompanying drawings and specific embodiments so that those skilled in the art can better understand the present invention and implement it, but the embodiments are not intended to limit the present invention. Embodiment 1

[0071] Reference Figure 1 As shown, the present invention relates to a surface shape detection method for a vehicle windshield, comprising:

[0072] Step S1: collecting two initial RMS images of the same windshield surface by a structured light surface measurement device;

[0073] Step S2: filtering and feature enhancement are performed on each initial RMS image to obtain two RMS images;

[0074] Step S3: performing color space conversion on the two RMS images to obtain two RMS images after color space conversion;

[0075] Step S4: Based on the two RMS images after color space conversion, determine whether the two initial RMS images collected have external environmental interference. If there is no external environmental interference, obtain the surface shape detection result of the windshield according to the RMS image; if there is external environmental interference, correct the two RMS images with external environmental interference to obtain corrected RMS images, and obtain the surface shape detection result of the windshield according to the corrected RMS images.

[0076] It should be noted that the surface detection of the windshield mainly detects the PV value and the RMS value, wherein the PV value is also called the peak-to-valley value, which is used to describe the maximum error value of the free-form surface shape after production compared to the designed surface shape. The PV value represents the maximum deviation between the actual surface shape and the designed surface shape on the entire free-form surface. The RMS value of this embodiment is reflected by the RMS image. The RMS value refers to the root mean square of the error value between the actual surface shape and the designed surface shape at each sampling pixel point on the free-form surface. The smaller the RMS value, the closer the surface shape is to the designed surface shape, and the better the imaging quality is generally. This embodiment only optimizes the RMS value obtained by the surface detection of the windshield. The RMS value can highlight the average degree of bending or wrinkling of the windshield as a whole, which is more intuitive and effective for windshield surface detection.

[0077] The following is a detailed introduction to this embodiment:

[0078] Furthermore, the method for performing color space conversion on the two RMS images in step S3 includes:

[0079] Convert the RMS image from RGB color space to HSV color space, which enhances the perceived intensity and vividness of the color and facilitates subsequent pixel analysis between different color regions;

[0080] Under the HSV color space condition, the hue of each pixel in the RMS image is counted, and the pixels with hues between 100° and 140° are set to green, the pixels with hues between 180° and 240° are set to blue, and the pixels with hues between 0° and 50° are set to orange-red. The formula is:

[0081] ;

[0082] in, is green, is blue, Orange-red. The hue range of this embodiment is both distinctive and can make the color recognition range wider and more accurate, which is convenient for the subsequent division of pixel areas. The RMS image of this embodiment is essentially a pseudo-color image with height error. The orange-red area indicates that the windshield height is higher than the theoretical value, the green area indicates that the windshield height is close to the theoretical value, and the blue area indicates that the windshield height is lower than the theoretical value.

[0083] Furthermore, the method for judging whether the two collected initial RMS images have external environmental interference based on the two RMS images after color space conversion in step S4 includes:

[0084] For the two RMS images after color space conversion, the pixel areas corresponding to green, blue, and orange-red pixels are divided by counting the hues of the pixels in the two RMS images. Figure 2 , Figure 2 The number 1 represents the pixel area where the orange-red pixel is located, the number 2 represents the pixel area where the blue pixel is located, and the number 3 represents the pixel area where the green pixel is located.

[0085] Obtain edge pixel coordinates of the pixel regions where the green, blue, and orange-red pixels are located in the two RMS images (equivalent to the outermost edge of each pixel region); obtain color edge curves of the respective colors according to the edge pixel coordinates of the pixel regions where the green, blue, and orange-red pixels are located; compare whether the two color edge curves corresponding to the same color pixels and the same pixel regions in the two RMS images are consistent. If the two color edge curves are inconsistent, it indicates that the initial RMS image obtained is interfered by the external environment; if the two color edge curves are consistent, it indicates that the initial RMS image obtained is not interfered by the external environment.

[0086] Furthermore, in step S4, the two RMS images with external environmental interference are corrected, and the correction method includes:

[0087] Determine whether there are the same coordinate points between the two corresponding color edge curves. If there are the same coordinate points, it indicates that there is an intersection, and the two color edge curves are corrected by a first optimization method, and the first optimization method includes a first case, a second case, and a third case; if there are no same coordinate points, it indicates that there is no intersection, and the two color edge curves are corrected by a second optimization method.

[0088] (I) The first situation of the first optimization method is as follows:

[0089] If the number of intersections is one, it is defined as ;

[0090] (1) If the intersection The two color edge curves on both sides are in a separated state (see Figure 3 (a) for details, the intersection point in Figure 3 (a) is A and the two color edge curves on both sides of the intersection point A are in a separated state), then obtain the color edge curve at the intersection point The starting coordinates of the two color edge curves on one side and , and obtain the The end coordinates of the two color edge curves on the other side and ,in, and are the starting point and end point of the first color edge curve, and are the starting point and end point of the second color edge curve; Calculate the new starting point coordinates ,pass Calculate the new end point coordinates ; Set the new starting point coordinates With this intersection Connect the curves to get the first curve segment and set the new end point coordinates With this intersection Connect the curves to obtain a second curve segment, and replace the original two color edge curves with a curve formed by reconnecting the first curve segment and the second curve segment;

[0091] (2) If the intersection If only one of the two color edge curves on the two sides is separated (see Figure 3 (b), the intersection point in Figure 3 (b) is A), then obtain the color edge curve at the intersection point. The starting coordinates of the two color edge curves on one side and ,pass Calculate the new starting point coordinates , the new starting point coordinates With this intersection Connect the curves to obtain curve segments, and replace the original two corresponding color edge curve segments with the curve segments.

[0092] (II) The second situation of the first optimization method is as follows:

[0093] If the number of intersections is two (see Figure 4(a), where the intersections are A and B), it is defined as and , count two intersection points and The total number of pixels in the closed area formed by the two color edge curves;

[0094] (1) If the total number of pixels is less than the preset number, take the two intersection points and Any color edge curve segment between them is used as the intersection point of the two and The color edge curve segment between the two intersection points and The area formed between them is so small that the error can be ignored.

[0095] (2) If the total number of pixels is greater than or equal to the preset number, calculate the two intersection points and The maximum Euclidean distance between the two color edge curves (the Euclidean distance requires the slopes of the two curves to be as consistent as possible, so as to obtain the true Euclidean distance between the two color edge curves), and take the midpoint of the line segment where the maximum Euclidean distance is located , the two intersection points and Respectively with the midpoint Connect to form a curve segment and use it as the two intersection points and The color edge curve segment between them replaces the two original corresponding color edge curve segments;

[0096] (3) Finally, determine the two intersection points and Are the two color edge curves other than the two color edge curves separated? If not, the two intersection points and If the two color edge curves other than the intersection point A and B are the same curve (see Figure 4 (b)), no optimization is required; if they are separated (see Figure 4 (a), the two color edge curves on both sides of the intersection point A and B are separated), then the two intersection points and The two color edge curves outside the intersection execute the steps corresponding to the situation in the first case where only the two color edge curves on one side of the intersection are in a separated state. In short, the color edge curve between the intersection points A and B does not need to be considered because it has been optimized. It is only necessary to consider whether the two color edge curves outside the intersection points A and B are in a separated state.

[0097] (III) The third situation of the first optimization method is as follows:

[0098] If the number of intersections is three (see Figure 5(a), where the intersections are A, B, and C), it is defined as , , , and the intersection Located in and between;

[0099] (1) Three intersection points , , Construct a circumscribed circle as the reference, and place the three intersection points , , The circumscribed circle segment is taken as the curve segment and used as the three intersection points , , The color edge curve segment is used to replace the original two corresponding color edge curve segments;

[0100] (2) Finally, determine the two intersection points and The two color edge curves other than the intersection points A and C in Figure 5 (a) are separated (i.e., whether the two color edge curves other than the intersection points A and C in Figure 5 (a) are separated). If they are not separated (the two color edge curves other than the intersection points A and C in Figure 5 (b) are not separated), then the two intersection points and The two color edge curves other than are the same curve and do not need to be optimized; if they are separated, the two intersection points and The other two color edge curves perform the steps corresponding to the first situation in which only the two color edge curves on one side of the intersection are in the separated state.

[0101] It should be noted that if there is an intersection and the above-mentioned first, second and third conditions are not met, it means that the two color edge curves have too many intersections (greater than or equal to 4). The two color edge curves can be fitted by linear difference, or return to step S1 to re-acquire the initial RMS image.

[0102] For further information, see Figure 6 If the two color edge curves do not have the same coordinate point, it indicates that there is no intersection point. The two color edge curves are corrected by the second optimization method. The correction method includes:

[0103] Construct a curve to correct the two color edge curves. This curve is located between the two color edge curves, and count the number of pixels between this curve and the first color edge curve. , count the number of pixels between this curve and the second color edge curve ,and and satisfy: It is not difficult to find that the curve needs to pass through the two original color edge curves and ensure that the pixels on both sides are as equal as possible, so as to minimize the error.

[0104] In summary, the first optimization method and the second optimization method can correct the two RMS images with external environmental interference and obtain the corrected RMS images.

[0105] Furthermore, in step S4, the surface shape of the windshield is detected based on the corrected RMS image. Specifically, any corrected RMS image is converted from the HSV color space to the RGB color space, and then input into the computer to obtain the corresponding RMS value. The smaller the RMS value, the more it can reflect the average degree of bending or wrinkling of the windshield as a whole. The smaller the RMS value, the closer the surface shape of the windshield is to the ideal surface shape.

[0106] Furthermore, the method of filtering and feature enhancing each initial RMS image in step S2 includes:

[0107] Filter each initial RMS image by a Gaussian filter or a median filter to remove noise;

[0108] The feature enhancement of each filtered initial RMS image is performed through the Laplace filter to enhance the contrast and edge of the image.

[0109] It should be noted that in step S1, a plurality of sinusoidal periodic fringe images on the surface of the windshield 6 are collected by a structured light surface measurement device, and the actual three-dimensional point cloud image of the windshield 6 is obtained by performing phase resolution on the plurality of sinusoidal periodic fringe images by a computer, and a pseudo-color image with a high degree of error is obtained by subtracting the actual three-dimensional point cloud image from the theoretical three-dimensional point cloud image, i.e., the initial RMS image mentioned above in this embodiment. In this embodiment, a total of two initial RMS images are obtained. The reason for obtaining two initial RMS images is that when collecting sinusoidal periodic fringe images, they will be affected by, for example, environmental conditions (such as air disturbance, environmental vibration, etc.), the accuracy of the measuring equipment, the accuracy of the measuring method, etc., and the sinusoidal periodic fringe images will be affected, which will cause errors in the initial RMS images.

[0110] The structured light surface measurement device of this embodiment minimizes the influence of the environment (such as air disturbance, environmental vibration, etc.) on the windshield 6 during the process of collecting sinusoidal periodic fringe images, and reduces the influence from the source, so that the two initial RMS images are as close as possible, that is, the error is as small as possible.

[0111] It is worth mentioning that see Figure 7 and Figure 8The structured light surface measurement device of this embodiment includes a first windshield detection module 1, an internal bracket 2, a motion module 3, a gauge platform 4, an angle adjustment telescopic module 5, a windshield 6, a windshield gauge 7, a second windshield detection module 8, and an external bracket 9. The second windshield detection module 8 is arranged on the top of the external bracket 9, the first windshield detection module 1 is arranged inside the internal bracket 2, the internal bracket 2 is also used to carry the gauge platform 4, the gauge platform 4 is specifically arranged on the top of the internal bracket 2, the gauge platform 4 is used to place the windshield 6, and the internal bracket 2 is arranged inside the external bracket 9.

[0112] See also Fig. 9 The first windshield detection module 1 (used to detect the convex surface of the windshield 6) of this embodiment includes a first image generator 101 and a first CCD camera 102, both of which are connected to a computer. The first image generator 101 and the first CCD camera 102 can be integrated into one. Similarly, the second windshield detection module 8 (used to detect the concave surface of the windshield 6) includes a second image generator 801 and a second CCD camera 802, both of which are connected to a computer. The second image generator 801 and the second CCD camera 802 can be integrated into one.

[0113] Furthermore, the first image generator 101 or the second image generator 801 of the present embodiment uses a high-precision LCD display screen to display sinusoidal periodic fringes, and irradiates the displayed sinusoidal periodic fringes onto the surface to be tested of the windshield 6 (i.e. Fig. 9 The incident light in the windshield is then captured by the first CCD camera 102 or the second CCD camera 802 after being reflected by the mirror surface to be measured on the windshield 6 (i.e. Fig. 9 The method further comprises the following steps: first, the computer 102 and the second CCD camera 802 are used to detect the reflected light in the windshield 6. Specifically, dozens of sinusoidal periodic fringe images need to be taken and transmitted to the computer. Finally, the computer dephases the deformed fringe (i.e., sinusoidal periodic fringe image) taken by the first CCD camera 102 or the second CCD camera 802 to obtain an actual three-dimensional point cloud image. The actual three-dimensional point cloud image is compared with the theoretical three-dimensional point cloud image, and the surface shape detection result of the test area of ​​the windshield 6 (which may include a PV value and an RMS value) is obtained according to the comparison result.

[0114] The reason why the windshield detection modules are provided on both sides (i.e., the concave surface and the convex surface) of the windshield 6 in this embodiment is to be able to perform surface shape detection on both the concave surface and the convex surface of the windshield 6. The first windshield detection module 1 and the second windshield detection module 8 of this embodiment are both built with a laser connected to a computer, and the laser is used to locate the spatial position coordinates of the windshield 6 to be tested, so as to facilitate the subsequent calculation of the surface shape of the windshield 6. In actual use, the first windshield detection module 1 and the second windshield detection module 8 are not used at the same time. If the convex surface shape of the windshield 6 needs to be measured, the first windshield detection module 1 is turned on, and if the concave surface shape of the windshield 6 needs to be measured, the second windshield detection module 8 is turned on.

[0115] Furthermore, considering the characteristics that the concave surface of the windshield 6 is easy to focus light and the convex surface is easy to scatter light, the display area of ​​the first image generator 101 of this embodiment is larger than the display area of ​​the second image generator 801 to facilitate actual measurement.

[0116] See also Fig. 9 , Fig. 9 The first windshield detection module 1 includes only a first CCD camera 102, Fig. 9 The second windshield inspection module 8 includes only one second CCD camera 802. Setting one CCD camera can effectively measure the central area of ​​the windshield 6 to be inspected. However, considering that in addition to measuring the surface shape of the central area of ​​the windshield 6, it is also necessary to measure the surface shape of the surrounding areas of the windshield 6, so it is necessary to configure multiple CCD cameras. For details, please refer to Fig.10 The first windshield detection module 1 is equipped with a plurality of first CCD cameras 102, and the second windshield detection module 8 is also equipped with a plurality of second CCD cameras 802. The purpose of setting a plurality of CCD cameras is to meet the measurement of a large-diameter windshield 6, or to meet the measurement of a lateral range of the windshield 6 (i.e., the left and right sides of the windshield 6). In short, the number of CCD cameras can be specifically set according to actual measurement requirements.

[0117] The internal bracket 2 is slidably disposed in the external bracket 9 , and the internal bracket 2 is disposed on the motion module 3 . The motion module 3 drives the internal bracket 2 to slide to the outside of the external bracket 9 , thereby facilitating the staff to place the windshield 6 on the inspection fixture platform 4 of the internal bracket 2 .

[0118] See also Fig.11The motion module 3 of this embodiment is a servo electric slide, which is used to slide the internal bracket 2 to the outside of the structured light surface measurement device. The internal bracket 2 is moved on the motion module 3 (i.e., the servo electric slide), which facilitates the staff to place the windshield 6. Since the servo electric slide is a prior art, it is not described in detail in this embodiment.

[0119] The angle adjustment telescopic module 5 of this embodiment is a telescopic rod with adjustable length. One end of the angle adjustment telescopic module 5 is fixedly connected to the internal bracket 2, and the other end is connected to one end of the inspection tool platform 4. The other end of the inspection tool platform 4 is also connected to the internal bracket 2 through a rotating component (such as a rotating shaft). Specifically, the angle adjustment telescopic module 5 is extended to drive the rotating component to rotate, thereby adjusting the placement angle of the windshield 6. It is not difficult to find that the angle adjustment telescopic module 5 can effectively facilitate the staff to set the measurement angle.

[0120] See also Fig.12 The inspection fixture platform 4 of this embodiment is hollowed out. The purpose of setting the hollowed-out shape is to allow the first windshield detection module 1 and the second windshield detection module 8 located on both sides of the windshield 6 to detect the surface shape of the windshield 6. The non-hollowed-out surroundings of the inspection fixture platform 4 of this embodiment include a plurality of support frames 401, which support the windshield 6 to be tested, and the position of the support frame 401 is adjustable to accommodate windshields 6 of different sizes. In this embodiment, the support frame 401 can be adjusted to open (open outward) or tighten (tighten inward) around by manual means (specifically, by the bolts provided on the support frame 401 to drive the gears to rotate). In other embodiments, the position of the support frame 401 can also be adjusted electrically.

[0121] The external bracket 9 of this embodiment is not only used to fix the second windshield detection module 8, but also used to place the internal bracket 2 loaded with the windshield 6 inside the external bracket 9.

[0122] In summary, the structured light surface shape measurement device of the present embodiment can not only perform non-contact surface shape detection on both the concave and convex surfaces of the windshield 6 , but also has a fast detection speed (detection is completed within 40 seconds), and can also accommodate any type of windshield 6 , effectively saving detection costs and being easy to promote on a large scale. Embodiment 2

[0123] This embodiment provides a surface shape detection system for a vehicle windshield, comprising:

[0124] Image extraction module: used to collect two initial RMS images of the same windshield surface;

[0125] Image enhancement module: used to filter and enhance the features of each initial RMS image to obtain two RMS images;

[0126] Color space conversion module: used to perform color space conversion on two RMS images to obtain two RMS images after color space conversion;

[0127] Judgment and correction module: used to judge whether the two initial RMS images collected have external environmental interference based on the two RMS images after color space conversion. If there is no external environmental interference, the surface detection result of the windshield is obtained according to the RMS image; if there is external environmental interference, the two RMS images with external environmental interference are corrected to obtain the corrected RMS images, and the surface detection result of the windshield is obtained according to the corrected RMS images. Embodiment 3

[0128] This embodiment provides a detection device, including a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the computer program, the steps of the surface shape detection method for a vehicle windshield described in the first embodiment are implemented. Embodiment 4

[0129] This embodiment provides a computer-readable storage medium on which a computer program is stored. When the computer program is executed by a processor, the steps of the surface shape detection method for a vehicle windshield described in the first embodiment are implemented.

[0130] Those skilled in the art will appreciate that the embodiments of the present application may be provided as methods, systems, or computer program products. Therefore, the present application may adopt the form of a complete hardware embodiment, a complete software embodiment, or an embodiment combining software and hardware. Moreover, the present application may adopt the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program codes. The schemes in the embodiments of the present application may be implemented in various computer languages, for example, object-oriented programming language Java and literal scripting language JavaScript, etc.

[0131] The present application is described with reference to the flowcharts and / or block diagrams of the methods, devices (systems), and computer program products according to the embodiments of the present application. It should be understood that each process and / or box in the flowchart and / or block diagram, as well as the combination of the processes and / or boxes in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing device to generate a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the processes in the flowchart and / or block diagram. Figure 1 A process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.

[0132] These computer program instructions may also be stored in a computer-readable memory capable of directing a computer or other programmable data processing device to operate in a specific manner, so that the instructions stored in the computer-readable memory produce an article of manufacture comprising an instruction device, which implements the process Figure 1 A process or multiple processes and / or boxes Figure 1 A function specified in one or more boxes.

[0133] These computer program instructions can also be loaded onto a computer or other programmable data processing device so that a series of operating steps are executed on the computer or other programmable device to produce a computer-implemented process, thereby providing instructions for implementing the process. Figure 1 A process or multiple processes and / or boxes Figure 1 The steps for the functions specified in one or more boxes.

[0134] Although the preferred embodiments of the present application have been described, those skilled in the art may make other changes and modifications to these embodiments once they have learned the basic creative concept. Therefore, the appended claims are intended to be interpreted as including the preferred embodiments and all changes and modifications falling within the scope of the present application.

[0135] Obviously, the above embodiments are merely examples for clear explanation and are not intended to limit the implementation methods. For those skilled in the art, other different forms of changes or modifications can be made based on the above description. It is not necessary and impossible to list all the implementation methods here. The obvious changes or modifications derived from these are still within the protection scope of the invention.

Claims

1. A method for detecting the surface shape of a vehicle windshield, characterized in that: include: Step S1: obtaining two initial RMS images of the same windshield surface; Step S2: filtering and feature enhancement are performed on each initial RMS image to obtain two RMS images; Step S3: performing color space conversion on the two RMS images to obtain two RMS images after color space conversion; Step S4: judging whether the two initial RMS images collected have external environmental interference based on the two RMS images after color space conversion; if there is no external environmental interference, obtaining the surface shape detection result of the windshield according to the RMS images; if there is external environmental interference, correcting the two RMS images with external environmental interference to obtain corrected RMS images, and obtaining the surface shape detection result of the windshield according to the corrected RMS images; The method for judging whether the two collected initial RMS images are interfered by the external environment based on the two RMS images after the color space conversion in step S4 includes: For the two RMS images after color space conversion, the pixel areas corresponding to green, blue, and orange-red pixels are divided by counting the hues of the pixels in the two RMS images; Get the edge pixel coordinates of the pixel areas where the green, blue, and orange-red pixels are located in the two RMS images; According to the edge pixel coordinates of the pixel areas where the green, blue and orange-red pixels are located, the color edge curves of the respective colors are obtained; Compare whether the two color edge curves corresponding to the same color pixels and the same pixel areas in the two RMS images are consistent. If the two color edge curves are inconsistent, it indicates that the initial RMS image obtained is interfered by the external environment; if the two color edge curves are consistent, it indicates that the initial RMS image obtained is not interfered by the external environment.

2. The surface shape detection method for a vehicle windshield according to claim 1, characterized in that: The method for performing color space conversion on the two RMS images in step S3 includes: Convert the RMS image from RGB color space to HSV color space; Under the HSV color space condition, the hue of each pixel in the RMS image is counted, and the pixels with hues between 100° and 140° are set to green, the pixels with hues between 180° and 240° are set to blue, and the pixels with hues between 0° and 50° are set to orange-red. The formula is: ； in, is green, is blue, Orange-red. For color tone.

3. The surface shape detection method for a vehicle windshield according to claim 1, characterized in that: The method for correcting the two RMS images with external environmental interference in step S4 includes: Determine whether there are the same coordinate points between the two color edge curves. If there are the same coordinate points, it indicates that there is an intersection, and the two color edge curves are corrected by the first optimization method; if there are no same coordinate points, it indicates that there is no intersection, and the two color edge curves are corrected by the second optimization method.

4. The surface shape detection method for a vehicle windshield according to claim 3, characterized in that: The first optimization method is specifically: If the number of intersections is one, it is defined as ; If the intersection If the two color edge curves on both sides are separated, then get the color at the intersection. The starting coordinates of the two color edge curves on one side and , and obtain the The end point coordinates of the two color edge curves on the other side and ,in, and are the starting point and end point of the first color edge curve, and are the starting point and end point of the second color edge curve; Calculate the new starting point coordinates ,pass Calculate the new end point coordinates ; Set the new starting point coordinates With this intersection Connect the curves to get the first curve segment and set the new end point coordinates With this intersection Connect the curves to obtain a second curve segment, and replace the original two color edge curves with a curve formed by reconnecting the first curve segment and the second curve segment; If the intersection If only one of the two color edge curves on the two sides is separated, then get the color edge curve at the intersection. The starting coordinates of the two color edge curves on one side and ,pass Calculate the new starting point coordinates , the new starting point coordinates With this intersection Connect the curves to obtain curve segments, and replace the two original corresponding color edge curve segments with the curve segments.

5. The surface shape detection method for a vehicle windshield according to claim 4, characterized in that: The first optimization method is specifically: If the number of intersections is two, it is defined as and , then count the two intersection points and The total number of pixels in the closed area formed by the two color edge curves; If the total number of pixels is less than the preset number, the two intersection points are taken and Any color edge curve segment between them is used as the intersection point of the two and The color edge curve segment between ; If the total number of pixels is greater than or equal to the preset number, calculate the two intersection points and The maximum Euclidean distance between the two color edge curves, take the midpoint of the line segment where the maximum Euclidean distance is located , the two intersection points and Respectively with the midpoint Connect to form a curve segment and use it as the two intersection points and The color edge curve segment between them replaces the two original corresponding color edge curve segments; Finally, determine the two intersection points and Are the two color edge curves other than the two color edge curves separated? If not, the two intersection points and The two color edge curves other than are the same curve and do not need to be optimized; if they are separated, the two intersection points and The other two color edge curves perform steps corresponding to the situation that only the two color edge curves on one side of the intersection are in a separated state.

6. The surface shape detection method for a vehicle windshield according to claim 4, characterized in that: The first optimization method is specifically: If the number of intersections is three, it is defined as , , , and the intersection Located in and between; With three intersections , , Construct a circumscribed circle as the reference, and place the three intersection points , , The circumscribed circle segment is taken as the curve segment and used as the three intersection points , , The color edge curve segment is used to replace the original two corresponding color edge curve segments; Finally, determine the two intersection points and Are the two color edge curves other than the two color edge curves separated? If not, the two intersection points and The two color edge curves other than are the same curve and do not need to be optimized; if they are separated, the two intersection points and The other two color edge curves perform steps corresponding to the situation that only the two color edge curves on one side of the intersection are in a separated state.

7. The surface shape detection method for a vehicle windshield according to claim 3, characterized in that: The method for correcting two color edge curves by the second optimization method includes: Construct a curve to correct the two color edge curves. This curve is located between the two color edge curves, and count the number of pixels between this curve and the first color edge curve. , count the number of pixels between this curve and the second color edge curve ,and and satisfy: .

8. The surface shape detection method for a vehicle windshield according to claim 1, characterized in that: In step S1, a sinusoidal periodic fringe image of the same windshield surface is collected by a structured light surface measurement device, a plurality of sinusoidal periodic fringe images are phase-deconstructed to obtain an actual three-dimensional point cloud image of the windshield, the actual three-dimensional point cloud image and the theoretical three-dimensional point cloud image are subtracted to obtain a pseudo-color image with a high error, and two pseudo-color images with a high error are obtained as initial RMS images, wherein: The structured light surface measurement device comprises: A first windshield detection module (1) comprises a first image generator (101) and a first CCD camera (102) connected to a computer, wherein the first image generator (101) is used to display first sinusoidal periodic fringes, first incident light formed by the first sinusoidal periodic fringes reaches one side of the windshield (6) to form first reflected light, and the first CCD camera (102) is used to collect the first reflected light and feed it back to the computer to obtain a surface shape detection result of one side of the windshield (6); The second windshield detection module (8) comprises a second image generator (801) and a second CCD camera (802) connected to a computer, wherein the second image generator (801) is used to display second sinusoidal periodic stripes, and second incident light formed by the second sinusoidal periodic stripes reaches the other side of the windshield (6) to form second reflected light, and the second CCD camera (802) is used to collect the second reflected light and feed it back to the computer to obtain a surface detection result of the other side of the windshield (6), wherein the first reflected light collected by the first CCD camera (102) and the second reflected light collected by the second CCD camera (802) are both sinusoidal periodic stripe images.

9. A surface shape detection system for a vehicle windshield, using the surface shape detection method for a vehicle windshield as claimed in claim 1, characterized in that: include: Image extraction module: used to collect two initial RMS images of the same windshield surface; Image enhancement module: used to filter and enhance the features of each initial RMS image to obtain two RMS images; Color space conversion module: used to perform color space conversion on two RMS images to obtain two RMS images after color space conversion; Judgment and correction module: used to judge whether the two initial RMS images collected have external environmental interference based on the two RMS images after color space conversion. If there is no external environmental interference, the surface detection result of the windshield is obtained according to the RMS image; if there is external environmental interference, the two RMS images with external environmental interference are corrected to obtain the corrected RMS images, and the surface detection result of the windshield is obtained according to the corrected RMS images.

10. A detection device, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that: When the processor executes the computer program, the steps of the surface shape detection method for a vehicle windshield as claimed in any one of claims 1 to 8 are implemented.

11. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the steps of the surface shape detection method for a vehicle windshield as claimed in any one of claims 1 to 8 are implemented.

Images