Scale line measurement method based on 3D vision

Abstract

The invention belongs to the technical field of image data processing, and particularly relates to a scribed line measuring method based on 3D vision. The method comprises the steps that a 3D camera collects data and refills abnormal values in an output depth map into normal values; drawing a straight line on the depth map along the scribed line direction, and calculating a line width compensation angle and a rectangular inclination angle of the measurement area; a flat regular rectangle is drawn on the depth map and used for roughly positioning edge points of the scribed lines, and at least two edge points are detected on each scribed line to be used for follow-up fine positioning of the edge points; combining the plurality of coarse positioning edge points on the same scribed line into one fine positioning edge point according to the set maximum point spacing; according to the fine positioning edge points and the rectangular measurement area, obtaining a 3D profile contour of the scribed line; and calculating the depth and width of the scribed line according to each 3D profile contour of the scribed line slot. The invention provides a scribed line measuring method based on 3D vision. The scribed line width and depth can be accurately and efficiently measured.

Description

Technical Field

[0001] The present invention belongs to the technical field of image data processing, and particularly relates to a scribing measurement method based on 3D vision. Background Art

[0002] In a lithium battery electrode laser scribing device, it is necessary to etch the coating in the electrode area of the battery cell. According to the requirements of the battery cell process, the device continuously feeds the electrode through unwinding and rewinding, and after rectification and positioning, quickly burns, vaporizes, expands, and peels off the active material in the designated areas on the front and back sides of the agreed coating area to remove the dressing within the agreed range and depth, and make it meet the etching quality and battery cell performance requirements.

[0003] The width and depth of the etched wire grooves directly affect the yield of the product. Therefore, it is necessary to detect the wire width and wire depth. Since both the wire width and wire depth of the etched line are in the micron level, how to accurately and efficiently measure and meet the real-time requirements of the industrial scenario has become a difficulty in the measurement. The scribing measurement method based on 3D vision provided by the present invention effectively solves the above difficulties. Summary of the Invention

[0004] The problem to be solved by the present invention is to provide a method for accurately and efficiently measuring the wire width and wire depth of an etched line.

[0005] The present invention discloses a scribing measurement method based on 3D vision, and its specific technical solution is as follows: A scribing measurement method based on 3D vision includes the following steps: Step S1: The 3D camera collects data and refills the outliers in the output depth map with normal values; Step S2: Draw a straight line along the scribing direction on the depth map, and calculate the wire width compensation angle and the inclination angle of the rectangular measurement area; Step S3: Draw a flat regular rectangle on the depth map to roughly locate the edge points of the scribing. At least two roughly located edge points should be detected on each scribing line for subsequent precise positioning of edge points; Step S4: According to the set maximum point spacing, merge multiple roughly located edge points on the same scribing line into 1 precisely located edge point; Step S5: Set a rectangular measurement area according to the positions of the precisely located edge points, and obtain the 3D profile of the scribing line within the rectangular measurement area; Step S6: Calculate the depth and width of the scribing line according to each 3D profile of the scribing line groove.

[0006] Further, the specific steps of Step S1 include: Step S1.1: The 3D camera collects data to obtain a depth map; Step S1.2: Refill the outliers in the depth map with normal values for more friendly visualization; Among them, when filling the outliers, the automatic filling mode or the manual filling mode is adopted. When adopting the automatic filling, the average depth value of the valid value area in the depth map will be calculated first , and this average value is filled into all outlier areas. The calculation of this average value is as follows: ; Among them, R is the image area for which the average value needs to be calculated, F is the pixel area of area R, p is the pixel point on area R, is the gray value of point p.

[0007] Furthermore, the specific steps of step S2 include: After drawing a straight line on the depth map, two points on this straight line can be obtained. Using these two points, the line width compensation angle and the inclination angle of the measurement area rectangle can be calculated. The calculation methods are as follows: ; ; Among them is the line width compensation angle, is the inclination angle of the measurement area rectangle, and respectively represent the pixel precisions in the horizontal and vertical directions of the image, , respectively represent the pixel coordinates of the starting point and the ending point when drawing the straight line.

[0008] Furthermore, the specific steps of step S3 include: Step S3.1: Draw a positive rectangular ROI for finding rough positioning edge points on the depth map to determine the measurement area; Step S3.2: Set appropriate smoothing coefficients and minimum edge thresholds so that at least two rough positioning edge points can be roughly positioned on each engraved line, which are the transitions from bright to dark and from dark to bright respectively, for subsequent use in accurately positioning edge points.

[0009] Furthermore, the specific steps of step S5 include: Step S5.1: Set the half-width and half-height of the rectangular measurement area, set the height of the small measurement rectangle on each engraved line, the number of profiles to be measured on each engraved line, and the Gaussian smoothing coefficient of the 3D profile contour; Step S5.2: Determine the area where the 3D profile contour to be obtained is located according to the accurately positioned edge points and the measurement area rectangle inclination angle parameters that have been calculated, and use the corresponding interpolation method on the depth map to obtain the 3D profile contour of the engraved line; among them, the obtained one-dimensional 3D profile contour needs to be preprocessed by Gaussian smoothing. The calculation method is as follows: ; where is the Gaussian smoothing coefficient, is the abscissa, is the Gaussian filter kernel coefficient.

[0010] Further, the step S6 specifically includes: Step S6.1: Set an appropriate offset and the search range for the minimum value of the wire groove; Step S6.2: Measure the 3D profile of each engraved line according to the calculated line width compensation angle to obtain the line width and line depth of each 3D profile; Among them, when calculating the line depth, the lowest point of the profile needs to be obtained, and the first derivative of the smoothed 3D profile needs to be calculated. For discrete one-dimensional profile points, the calculation method is as follows: ; where , represent the pixel values of the next and previous pixels at the position respectively, and represent the current position, the next position, and the previous position respectively.

[0011] The present invention collects data through a 3D camera and refills the outliers in the depth map output by the 3D camera with normal values; draws a straight line along the engraved line direction on the filled depth map, calculates the line width compensation angle and the inclination angle of the rectangular measurement area; then draws a regular rectangle with a flat shape for roughly positioning the edge points of the engraved line, and at least two edge points should be detected on each engraved line for subsequent precise positioning of the edge points; according to the set maximum point spacing, merge multiple roughly positioned edge points on the same engraved line into 1 precisely positioned edge point; according to the position of the precisely positioned edge point and the set rectangular measurement area, obtain the 3D profile of the engraved line; calculate the depth and width of the engraved line according to each 3D profile of the engraved line groove. Through the above steps, the present invention can efficiently and accurately detect the line width and line depth of the etched line, and has practical application value in industrial detection scenarios that require real-time performance and accuracy. BRIEF DESCRIPTION OF THE DRAWINGS

[0012] Figure 1 is a schematic flow chart of the method of the present invention; Figure 2 is a schematic diagram of the depth map collected by the 3D camera and the redrawn depth map in the embodiment of the present invention; Figure 3 is a schematic diagram of the straight line drawn when using the angle calculation tool in the embodiment of the present invention; Figure 4 is a schematic diagram of the regular rectangle ROI drawn when roughly positioning the edge points in the embodiment of the present invention; Figure 5 For Figure 4 The enlarged schematic diagram at point C in Figure 6 The schematic diagram of the edge point positions calculated by the fine-positioning edge points in the embodiment of the present invention; Figure 7 The schematic diagram of the rectangular measurement area centered on the fine-positioning edge points in the embodiment of the present invention; Figure 8 The schematic diagram of the 3D profile obtained in the embodiment of the present invention; Figure 9 The schematic diagram of the points and lines used to calculate the line width and line depth in the embodiment of the present invention. Detailed implementation manners

[0013] The present invention will be further clarified below in conjunction with the accompanying drawings and detailed implementation manners.

[0014] Figure 1 The flowchart of the method of the present invention includes the following steps: Step S1: The 3D camera collects data and repopulates the outliers in the output depth map with normal values; Step S1.1: The 3D camera collects data to obtain a depth map; Step S1.2: Repopulate the outliers in the depth map with normal values for more friendly visualization.

[0015] Among them, when filling the outliers, two modes of automatic filling or manual filling can be adopted. When automatic filling is adopted, the average depth value of the valid value area in the depth map will be calculated first , and this average value is filled into all outlier areas. The calculation of this average value is as follows: ; Among them, R is the image area for which the average value needs to be calculated, F is the pixel area of area R, p is the pixel point on area R, is the gray value of point p.

[0016] Step S2: Draw a straight line along the scribed line direction on the depth map, and calculate the line width compensation angle and the rectangular inclination angle of the measurement area; After drawing a straight line on the depth map, two points on this straight line can be obtained, and the line width compensation angle and the rectangular inclination angle of the measurement area can be calculated using these two points. The calculation method is as follows: ; ; Among them is the line width compensation angle, is the rectangular inclination angle of the measurement area, and respectively represent the pixel precision in the horizontal and vertical directions of the image and respectively represent the pixel coordinates of the starting point and the ending point when drawing a straight line

[0017] Step S3: Draw a flat and regular rectangle on the depth map to roughly locate the edge points of the engraved line. At least two roughly located edge points should be detected on each engraved line for subsequent precise location of edge points Step S3.1: Draw a regular rectangle ROI for finding roughly located edge points on the depth map to determine the measurement area Step S3.2: Set appropriate smoothing coefficients and minimum edge thresholds so that at least two roughly located edge points are roughly located on each engraved line, which are the transitions from bright to dark and from dark to bright respectively, for subsequent precise location of edge points

[0018] Step S4: According to the set maximum point spacing, merge multiple roughly located edge points on the same engraved line into 1 precisely located edge point

[0019] Step S5: Set a rectangular measurement area according to the position of the precisely located edge points, and obtain the 3D profile of the engraved line within the rectangular measurement area Step S5.1: Set the half-width and half-height of the rectangular measurement area, set the height of the small measurement rectangle on each engraved line, the number of profiles to be measured on each engraved line, and the Gaussian smoothing coefficient of the 3D profile Step S5.2: Determine the area where the 3D profile to be obtained is located according to the calculated precisely located edge points and the rectangular inclination parameter of the measurement area, and use the corresponding interpolation method on the depth map to obtain the 3D profile of the engraved line. Among them, the obtained one-dimensional 3D profile needs to be preprocessed by Gaussian smoothing, and the calculation method is as follows ; where is the Gaussian smoothing coefficient is the abscissa is the Gaussian filter kernel coefficient

[0020] Step S6: Calculate the depth and width of the engraved line according to each 3D profile of the engraved line groove Step S6.1: Set appropriate offsets and the search range for the minimum value of the wire groove Step S6.2: Measure each 3D profile of the engraved line according to the calculated line width compensation angle to obtain the line width and line depth of each 3D profile

[0021] Among them, when calculating the line depth, the lowest point of the profile contour needs to be obtained, and the first derivative of the smoothed 3D profile contour needs to be calculated. For discrete one-dimensional profile points, the calculation method is as follows: ; where 、 represent the positions of the pixel values of the next and previous pixels respectively, represent the current position, the next position and the previous position respectively.

[0022] The following takes an embodiment as an example to elaborate the specific calculation process of the above six steps in detail: Control the industrial camera through software to collect the depth map. Since the scales of the depth and width of the measured wire groove are small, the industrial 3D camera used in this embodiment is the 3D intelligent line confocal sensor of LMI, model Gocator 5504. The repeatability accuracy of this camera in the Z direction is 0.05um, which can meet the high-precision measurement requirements of the wire groove.

[0023] Since the pixel values of the invalid points in the depth map often deviate greatly from the average depth value of the effective area of the image, the wire groove captured in the image looks unclear, and even the whole image looks like the same gray value so that the wire groove cannot be seen. As Figure 2 shown in (a) of , the depth values of a small number of invalid points in this depth map are 0, which is much lower than the average depth value of other effective areas. For convenient and reasonable display for debugging, the abnormal values in this depth map can be filled with normal values. The effect after filling is as Figure 2 shown in (b) of . It can be seen that the wire groove in the depth map can be seen more clearly at this time.

[0024] When filling the invalid points, two modes of automatic filling or manual filling can be adopted. When adopting automatic filling, the average depth value of the valid value area in the depth map will be calculated first , and this average value will be filled into all abnormal value areas. The calculation of this average value is as follows: ; where R is the image area for which the average value needs to be calculated, F is the pixel area of area R, p is the pixel point on area R, is the gray value of point p.

[0025] When adopting manual filling, the abnormal value area will be filled with the manually set value.

[0026] After the depth map filling is completed, the subsequent calculations are based on this filled depth map, as Figure 2 shown in (b) of .

[0027] Draw a straight line approximately parallel to the scribed line direction on the filled depth map, as Figure 3 shown. After drawing the straight line, the starting point A and the ending point B on this straight line can be obtained. The coordinates of point A , and the coordinates of point B . Using points A and B, the line width compensation angle and the inclination angle of the measurement area rectangle can be calculated. The calculation method is as follows: ; ; where represents the line width compensation angle, represents the inclination angle of the measurement area rectangle, which is the angle between the drawn straight line and the horizontal direction in the image, and represent the pixel precisions in the horizontal and vertical directions of the image respectively, , represent the pixel coordinates of the starting point and the ending point respectively when drawing the straight line.

[0028] After calculating the line width compensation angle and the inclination angle of the measurement area rectangle, draw a positive rectangular ROI for finding edge points on the depth map to determine the measurement area, as Figure 4 shown. This positive rectangular ROI has the characteristics of a small height and a large width. By performing one-dimensional measurement within the measurement area it determines, a set of one-dimensional edge points can be obtained. These points are the roughly located scribed line edge points used to roughly determine the position of the scribed line.

[0029] When performing one-dimensional measurement of the rough edge points, the following two parameters need to be set: Smoothing coefficient: The Gaussian smoothing coefficient Sigma when determining rough edge points through one-dimensional measurement.

[0030] Minimum edge threshold: Used to screen edge points, and points higher than this threshold will be selected as roughly located edge points.

[0031] By setting appropriate smoothing coefficients and minimum edge thresholds, at least one pair of edge points can be roughly located on each scribed line, which are the transitions from bright to dark and from dark to bright respectively. As Figure 4 shown, three edge points are measured for each of the two scribed lines. Figure 5 is the enlarged view at C in Figure 4 . Figure 5 shows the enlarged details of the three roughly located edge points measured for the right scribed line among the two scribed lines , and .

[0032] After obtaining the roughly located edge points, these points need to be provided to the subsequent fine positioning tool in order to obtain the fine positioned edge points. In the fine positioning tool, an appropriate maximum point spacing needs to be set. This parameter is the distance threshold for judging whether different edge points belong to the same engraved line, and is used to merge multiple roughly located edge points on the same engraved line into 1 fine positioned edge point. If the spacing between adjacent roughly located edge points is greater than the set maximum point spacing, they are regarded as edge points on different engraved lines; otherwise, they are regarded as edge points on the same engraved line. As Figure 6 shown, there are 2 fine positioned edge points on 2 engraved lines. It can be seen from the figure that the 3 roughly located edge points extracted from each engraved line have been merged into 1. For the N edge point coordinates , , …,[[]]END]] extracted roughly for each engraved line, the following calculation method is used to merge them into 1 fine positioned edge point :[[]]END]] ;[[]]END]] ;[[]]END]] where respectively represent the row and column coordinates of the fine positioned edge point , , respectively represent the row and column coordinates of the i-th roughly located edge point.[[]]END]]

[0033] After completing the above calculations, enter the tool for obtaining the profile contour. This tool needs to set the following parameters:[[]]END]] Half-width of the detection area: The half-width of the rectangular measurement area.[[]]END]]

[0034] Half-height of the detection area: The half-height of the rectangular measurement area.[[]]END]]

[0035] Height of the measurement rectangle: The height of the small measurement rectangle on each engraved line.[[]]END]]

[0036] Smoothing coefficient: The Gaussian smoothing coefficient of the 3D profile contour.[[]]END]]

[0037] Number of single-line profile contours: The number of profiles to be measured on each engraved line.[[]]END]]

[0038] After completing the setting of these 5 parameters, according to the fine positioned edge points calculated, the inclination angle of the measurement area rectangle, the 3D profile contour of the engraved line can be obtained on the depth map by using the corresponding interpolation method. As Figure 7 shown, the two large rectangles are the rectangular measurement areas for obtaining the profile contour. The center point of the rectangle is the fine positioned edge point obtained in the previous step, and the inclination angle of the rectangle is the inclination angle of the measurement area rectangle calculated in the previous step.[[]]END]], 10 profile lines are calculated in each rectangular area in the figure. As Figure 8 shown, it is a 3D profile contour obtained. Among them, the obtained one-dimensional 3D profile contour needs to be preprocessed by Gaussian smoothing, and the calculation method is as follows: ; Among them is the Gaussian smoothing coefficient, is the abscissa, and its value range is determined by decided, is the Gaussian filter kernel coefficient.

[0039] After obtaining the smoothed 3D profile contour of the engraved line, the final line width and line depth measurement can be carried out. The parameters that need to be set are as follows: Offset: After the 3D profile line contour is fitted into a straight line, the offset when the straight line is translated in the y direction. The distance between the two intersection points 、 of the translated straight line and the profile line contour is the line width, and the y-direction distance between the straight line before translation and the lowest point in the wire groove is the line depth. As Figure 9 shown, the distance between the left and right two intersection points 、 of the translated straight line and the 3D profile contour at the edge of the wire groove is the line width, and the y-direction distance from the lowest point to the straight line before translation is the line depth, which is the y-direction distance between the points and in the image. Calculating the line width using the offset straight line can effectively improve the stability of the line width calculation.

[0040] Search range of wire groove minimum value: Used to expand the search range of the minimum value points in the wire groove. Usually, this parameter can be set to 5%.

[0041] After setting the above two parameters, calculate the compensation angle according to the calculated line width, and then the 3D profile contour of each engraved line can be measured to obtain the line width and line depth of each 3D profile contour, that is, the calculation of the width and depth of the engraved line is completed. Among them, when calculating the line depth, the lowest point of the profile contour needs to be obtained, and the first derivative of the smoothed 3D profile contour needs to be calculated. For discrete one-dimensional profile points, the calculation method is as follows: ; Among them 、 represent the pixel values of the pixel at the position one pixel after and one pixel before respectively, represent the current position, the position one pixel after and the position one pixel before respectively.

[0042] The description of the above embodiments only represents one implementation means of the present technology. The implementation means combined or derived from the methods of the present technology are still within the protection scope of the present invention.

Claims

1. A line engraving measurement method based on 3D vision, characterized in that, Including the following steps: Step S1: The 3D camera collects data and repopulates the outliers in the output depth map with normal values; Step S2: Draw a straight line on the depth map along the direction of the engraved line, and calculate the line width compensation angle and the inclination angle of the rectangular measurement area; Step S3: Draw a regular rectangle with a flat shape on the depth map to roughly locate the edge points of the engraved line. At least two roughly located edge points should be detected on each engraved line for subsequent precise positioning of edge points; Step S4: According to the set maximum point spacing, merge multiple roughly located edge points on the same engraved line into 1 precisely located edge point; Step S5: Set a rectangular measurement area based on the positions of the precisely located edge points, and obtain the 3D profile of the engraved line within the rectangular measurement area; Step S6: Calculate the depth and width of the engraved line based on the 3D profile of each groove of the engraved line; 2. The line measurement method based on 3D vision according to claim 1, wherein The specific steps of Step S1 include: Step S1.1: The 3D camera collects data to obtain a depth map; Step S1.2: Repopulate the outliers in the depth map with normal values for more friendly visualization; Among them, when filling abnormal values, an automatic filling mode or a manual filling mode is adopted. When automatic filling is used, the average depth value of the valid value area in the depth map will be calculated first , and this average value is filled into all abnormal value areas. The calculation of this average value is as follows: ； Among them, R is the image region for which the average value needs to be calculated, F is the pixel area of region R, and p is the pixel point on region R. is the gray value of point p.

3. The line measurement method based on 3D vision according to claim 1, characterized in that The specific steps of Step S2 include: After drawing a straight line on the depth map, two points on this line can be obtained, and the line width compensation angle and the inclination angle of the rectangular measurement area can be calculated using these two points. The calculation method is as follows: ； ； wherein is the line width compensation angle, is the inclination angle of the rectangle of the measurement area, and respectively represent the pixel precisions in the horizontal and vertical directions of the image, 、 respectively represent the pixel coordinates of the starting point and the ending point when drawing a straight line.

4. The line measurement method based on 3D vision according to claim 1, characterized in that The specific steps of Step S3 include: Step S3.1: Draw a regular rectangle ROI for finding roughly located edge points on the depth map to determine the measurement area; Step S3.2: Set appropriate smoothing coefficients and minimum edge thresholds so that at least two roughly located edge points are roughly located on each engraved line, which are the transitions from bright to dark and from dark to bright respectively, for subsequent precise positioning of edge points; 5. The line engraving measurement method based on 3D vision according to claim 1, wherein The specific steps of Step S5 include: Step S5.1: Set the half-width and half-height of the rectangular measurement area, the height of the small measurement rectangle on each engraved line, the number of profiles to be measured on each engraved line, and the Gaussian smoothing coefficient of the 3D profile; Step S5.2: Determine the area where the 3D profile to be obtained is located based on the already calculated precisely located edge points and the inclination angle parameters of the measurement area rectangle, and use the corresponding interpolation method on the depth map to obtain the 3D profile of the engraved line; among them, the obtained one-dimensional 3D profile needs to be preprocessed by Gaussian smoothing. The calculation method is as follows: ； wherein is the Gaussian smoothing coefficient, is the abscissa, is the Gaussian filter kernel coefficient.

6. The line measurement method based on 3D vision according to claim 1, wherein The specific steps of Step S6 include: Step S6.1: Set appropriate offset and the search range for the minimum value of the groove; Step S6.2: Measure each 3D profile of the engraved line according to the already calculated line width compensation angle to obtain the line width and line depth of each 3D profile; Among them, when calculating the line depth, the lowest point of the profile needs to be obtained, and the first derivative of the smoothed 3D profile needs to be calculated. For discrete one-dimensional profile points, the calculation method is as follows: ； Among them and represent the pixel values of the subsequent and previous pixels at position respectively, and represent the current position, the subsequent position, and the previous position respectively.

Images