A visual measurement method for the punching precision of the front end of a tape measure

Abstract

The application provides a visual measurement method for precision of a tape measure front end punching, comprising: step S1, obtaining a gray image of the tape measure front end to perform gradient calculation and generate a vertical edge response graph through threshold segmentation; step S2, performing multi-scale morphological processing on the vertical edge response graph to obtain an enhanced structure graph, performing low-frequency smoothing processing, threshold segmentation and morphological closing operation on the gray image, and combining the enhanced structure graph to generate a candidate target area mask; step S3, extracting a connected region in the candidate target area mask and screening to obtain a punching area; step S4, performing local gray analysis in a neighborhood of the punching area to extract a scale line candidate area and screening and sorting to obtain an effective scale line sequence; and step S5, calculating a deviation between a punching position and a theoretical reference value according to the punching area position and the effective scale line sequence. The beneficial effect is that the application can automatically and stably extract the punching area and the scale line of the tape measure front end with high precision, and quantitatively evaluate the punching deviation.

Description

Technical Field

[0001] This invention relates to the technical field of industrial visual inspection, and more specifically, to a visual measurement method for the punching accuracy of the front end of a measuring tape. Background Technology

[0002] In the production process of measuring tapes, the standard measuring tape size measurement diagram is as follows: Figure 1 As shown, if the punching position is too far forward or too far back, it will cause a mismatch between the zero position of the ruler hook and the starting point of the scale, resulting in system measurement errors. Therefore, rapid and accurate detection of the punching accuracy at the front end of the measuring tape is a key step in quality control during the production process.

[0003] Currently, the industry commonly uses manual visual inspection or offline sampling with the aid of tools such as magnifying glasses and projectors. However, manual inspection methods have significant shortcomings: on the one hand, the inspection speed is slow and the labor intensity is high, making it difficult to achieve full inspection on high-speed production lines; on the other hand, different inspectors have different subjective judgment standards for minor deviations, which easily leads to missed detections and misjudgments. In addition, some companies have tried to use general machine vision systems for inspection, but the surface of measuring tapes usually has characteristics such as metallic reflection, complex printed scale textures, and high background noise. Conventional grayscale threshold segmentation or edge detection algorithms are difficult to reliably extract punching boundaries and fine scale line structures. In particular, they are prone to misidentifying surface scratches, stains, or non-perpendicular textures as target areas, resulting in poor punching positioning accuracy or even failure to identify effective areas. At the same time, most existing methods fail to jointly analyze the punching position and scale line sequence, lack the ability to identify structural defects such as missing scales and abnormal spacing, and are also difficult to adapt to the online real-time inspection needs of different models of measuring tapes.

[0004] Therefore, developing a visual measurement method that can automatically, stably, and accurately extract the punching area and scale lines at the front end of a measuring tape, and quantitatively evaluate the punching deviation, has significant engineering application value. Summary of the Invention

[0005] The technical problem to be solved by the present invention is how to automatically, stably and with high precision extract the punching area and scale lines at the front end of a measuring tape, and quantitatively evaluate the punching deviation. In order to overcome the defects of the above-mentioned prior art (or related art), the present invention provides a visual measurement method for the punching accuracy of the front end of a measuring tape.

[0006] This invention provides a visual measurement method for the punching accuracy of the front end of a measuring tape, comprising the following steps:

[0007] Step S1: Obtain the grayscale image of the front end of the measuring tape, perform gradient calculation to extract the structural change information in the vertical direction, and generate a vertical edge response map through threshold segmentation.

[0008] Step S2: performing multi-scale morphological processing on the vertical edge response map to obtain an enhanced structure map, performing low-frequency smoothing processing, threshold segmentation and morphological closing operation on the grayscale image, and then generating a candidate target region mask in combination with the enhanced structure map;

[0009] Step S3: extracting connected regions from the candidate target region mask, screening out candidate punching regions whose area meets a preset condition, and determining a unique effective punching region through region expansion and boundary compensation;

[0010] Step S4: performing local grayscale analysis on the grayscale image in the neighborhood of the punching region, extracting candidate scale line regions through a dynamic threshold, and screening and sorting the scale lines in the candidate scale line regions based on length, position, spacing and width consistency to obtain an effective scale line sequence;

[0011] Step S5: calculating the deviation between the punching position and a theoretical reference value according to the position of the punching region and the effective scale line sequence, judging that the product is qualified if the deviation is within a preset allowable range, and judging that there is a punching error defect if the deviation is not within the preset allowable range.

[0012] Compared with the prior art, the visual measuring method for the front end punching precision of a tape measure of the present invention has the following advantages:

[0013] In the present invention, the vertical edge response map is generated in step S1, the candidate target region mask is generated in step S2, the punching region is screened in step S3, the effective scale line sequence is generated in step S4, and the deviation is calculated in step S5, so that automatic, stable and high-precision extraction of the punching region and scale lines at the front end of the tape measure is realized, which can replace manual visual inspection and reduce subjective errors; meanwhile, through the combination of gradient calculation and multi-scale morphology, the extraction capability of vertical structures is enhanced, and the stability and repeatability of visual measurement are improved; finally, the qualification or punching error defect is determined through quantitative deviation comparison, which is suitable for online quality inspection scenarios.

[0014] In a possible implementation, the multi-scale morphological processing in step S2 comprises sequentially performing a longitudinal closing operation, a longitudinal opening operation and a transverse closing operation, so as to enhance continuous structures in the vertical direction and suppress interference in the horizontal direction.

[0015] Compared with the prior art, adoption of the above technical solution can specifically enhance the continuous structures in the vertical direction in the grayscale image of the tape measure, and effectively suppress noise interference in the horizontal direction, which improves the accuracy of structure extraction and the anti-interference capability.

[0016] In a possible implementation, the generation process of the candidate target region mask in step S2 comprises:

[0017] The initial entity region is obtained by thresholding the grayscale image after low-frequency smoothing, and morphological closing operation is performed on the initial entity region to fill the internal holes, thereby generating the candidate target region mask.

[0018] Compared with existing technologies, the above technical solution can reduce the interference of fine-grained noise on threshold segmentation through low-frequency smoothing. Threshold segmentation can initially separate the target from the background, and morphological closing operation can fill the small holes in the target area, thereby obtaining a more complete and continuous candidate target area mask and reducing the risk of missed detection and over-segmentation.

[0019] In one possible implementation, the screening process for the candidate punching regions in step S3 includes:

[0020] Calculate the pixel area of ​​each connected region, remove connected regions with pixel areas smaller than a first preset threshold, and retain one or more connected regions with the largest pixel areas as candidate punching regions.

[0021] Compared with existing technologies, the above technical solution can eliminate excessively small connected regions by comparing area thresholds and retain one or more connected regions with the largest area. It can quickly and efficiently lock the most likely punching region from a complex background, reduce the search range of subsequent processing, and improve computational efficiency and positioning accuracy.

[0022] In one possible implementation, the region expansion and boundary compensation in step S3 includes:

[0023] Based on the candidate punching region, a preset pixel range is expanded outward, and the boundary contraction caused by morphological operations is compensated to restore the original structural continuity within the candidate punching region.

[0024] Compared with existing technologies, the above-mentioned technical solution can restore the boundary shrinkage or incomplete region caused by morphological operations, making the final determined punching area closer to the real structure and improving the accuracy and robustness of punching position positioning.

[0025] In one possible implementation, the extraction process of the candidate scale line region in step S4 includes:

[0026] Within the neighborhood of the punching region, local grayscale analysis is performed on the grayscale image to obtain the local grayscale change rate. A candidate tick mark binary image is obtained through dynamic threshold segmentation, and the connected region in the candidate tick mark binary image is extracted as the tick mark candidate region.

[0027] Compared with existing technologies, the above-mentioned technical solution can adapt to the illumination changes of different tick mark candidate regions through dynamic threshold segmentation, thereby improving the adaptability of tick mark segmentation; at the same time, combined with connected component extraction, it can accurately obtain the potential tick mark positions, providing high-quality candidate data for subsequent screening.

[0028] In one possible implementation, the process of filtering and sorting the scale lines in step S4 includes:

[0029] Short scale line candidate regions located in the top interference region are removed from the scale line candidate region. The scale line candidate regions whose length and width meet the preset range are retained as valid scale lines. Then, the valid scale lines are sorted according to their position coordinates, and the valid scale lines with abnormal spacing or width are removed to obtain the valid scale line sequence.

[0030] Compared with existing technologies, the above technical solution can effectively eliminate edge interference, false scale lines or abnormal structures, and the remaining effective scale line sequence has good consistency and sortability, which can significantly improve the accuracy and reliability of scale recognition.

[0031] In one possible implementation, step S4 further includes:

[0032] When there are missing scales in the effective scale line sequence, linear interpolation is performed based on the average spacing between adjacent scale lines to fill in the missing scale positions.

[0033] Compared with existing technologies, by adopting the above technical solution, even if some scale lines cannot be extracted due to local contamination, obstruction or insufficient contrast, the complete effective scale line sequence can still be completed by interpolation, ensuring the continuity and reliability of subsequent error assessment and reducing the misjudgment rate.

[0034] In one possible implementation, the conditions for determining a punching error defect in step S5 include the deviation exceeding the preset allowable range, the failure to detect the effective scale line sequence, the number of scales in the effective scale line sequence being lower than a second preset threshold, the presence of abnormal spacing in the effective scale line sequence, and the inability to identify at least one of the punching areas.

[0035] Compared with existing technologies, the above-mentioned technical solution can construct a multi-dimensional and multi-level defect judgment logic, which can not only detect punching position deviation, but also identify structural problems such as missing scale and abnormal spacing, thereby improving the comprehensiveness and robustness of visual measurement methods.

[0036] In one possible implementation, the theoretical benchmark value in step S5 is determined based on the standard punching position in the tape measure design drawing or by statistically analyzing multiple qualified sample images to obtain the standard punching position.

[0037] Compared with existing technologies, the above-mentioned technical solution can provide a flexible and standardized source of theoretical benchmark values, which can not only adapt to the design requirements of different models of measuring tapes, but also adapt to actual production fluctuations through statistical learning, thereby improving the rationality and adaptability of deviation assessment. Attached Figure Description

[0038] Figure 1 This is a schematic diagram of the conventional dimensional measurement diagram of the present invention;

[0039] Figure 2 This is a flowchart of the steps of the present invention;

[0040] Figure 3 This is a schematic diagram of the vertical edge response map of the present invention;

[0041] Figure 4 This is a schematic diagram of the candidate target region mask of the present invention;

[0042] Figure 5 This is a schematic diagram of the punching area of ​​the present invention. Detailed Implementation

[0043] First, those skilled in the art should understand that these embodiments are merely illustrative of the technical principles of the present invention and are not intended to limit the scope of protection of the present invention. Those skilled in the art can make adjustments as needed to adapt to specific application scenarios.

[0044] The present invention will now be described in further detail with reference to the accompanying drawings and specific embodiments.

[0045] See Figure 2 This invention discloses a visual measurement method for the punching accuracy of the front end of a measuring tape, comprising the following steps:

[0046] Step S1: Obtain the grayscale image of the front end of the measuring tape, perform gradient calculation to extract the structural change information in the vertical direction, and generate a vertical edge response map through threshold segmentation.

[0047] Step S2: Perform multi-scale morphological processing on the vertical edge response map to obtain an enhanced structure map, and perform low-frequency smoothing, threshold segmentation and morphological closing operation on the grayscale image. Then, combine the enhanced structure map to generate a candidate target region mask.

[0048] Step S3: Extract connected regions from the candidate target region mask, filter out candidate punching regions whose areas meet the preset conditions, and determine the unique effective punching region through region expansion and boundary compensation.

[0049] Step S4: Perform local grayscale analysis on the grayscale image in the neighborhood of the punching area, extract candidate regions of the scale line through dynamic thresholding, and filter and sort the scale lines in the candidate regions of the scale line based on the consistency of length, position, spacing and width to obtain an effective scale line sequence.

[0050] Step S5: Calculate the deviation between the punching position and the theoretical reference value based on the position of the punching area and the effective scale line sequence. If the deviation is within the preset allowable range, it is judged as qualified; if it is not within the preset allowable range, it is judged as a punching error defect.

[0051] In this embodiment of the invention, in step S1, an industrial camera (e.g., a CMOS camera with a resolution of at least 1280×1024 pixels) is first used in conjunction with a ring light source or a bar light source to acquire the original image of the front end area of ​​the measuring tape. The original image is then converted into a grayscale image. Subsequently, the Sobel operator (with a kernel size of 3×3 or 5×5) is used to calculate the gradient magnitude of each pixel in the grayscale image in the vertical direction to highlight vertical edge structures, such as punched boundaries and vertical segments of scale lines. The gradient magnitude map is then segmented using a fixed threshold (e.g., a threshold range of 30~50, with the specific value dynamically adjusted according to the image grayscale) or Otsu adaptive threshold segmentation to generate a binarized vertical edge response map, such as... Figure 3 As shown in the vertical edge response map, high-brightness pixels represent candidate edge locations where the vertical direction changes drastically.

[0052] In this embodiment of the invention, in step S2, the following three morphological operations are sequentially performed on the vertical edge response map:

[0053] Vertical closure operation: Using a vertical structuring element (e.g., a rectangular kernel with a height of 21 pixels and a width of 3 pixels), it first expands and then erodes to connect broken scale lines or punched boundaries in the vertical direction.

[0054] Vertical opening operation: Using the same or slightly smaller vertical structural elements (e.g., 15 pixels high and 3 pixels wide), erosion is performed first and then dilation is performed to remove isolated noise points or short interference segments in the vertical direction.

[0055] Horizontal closure operation: Uses horizontal structural elements (such as a rectangular kernel with a height of 3 pixels and a width of 21 pixels) to connect horizontal breaks in the same tick line caused by uneven lighting, while keeping the vertical structure unaffected.

[0056] The dimensions of the above structural elements can be scaled according to the camera resolution and the actual width of the measuring tape (typically 5-30 pixels in height and 2-5 pixels in width). After processing, an enhanced structural image is obtained, in which continuous vertical structures are significantly enhanced and horizontal noise is suppressed.

[0057] In this embodiment of the invention, in step S2, the original grayscale image undergoes low-frequency smoothing processing, for example, using Gaussian filtering (kernel size 5×5, standard deviation σ=1.0) or mean filtering (kernel size 7×7) to remove high-frequency noise. The smoothed grayscale image is then segmented using a fixed threshold (e.g., threshold range 80~120) or the Otsu's method to obtain initial entity regions. Since these initial entity regions may contain holes (e.g., scale gaps or reflective points), morphological closing operations (using circular or square structuring elements, size selectable 11×11 pixels) are further performed to fill the holes and generate a complete candidate target region mask, such as... Figure 4 As shown, the candidate target area mask covers the measuring tape entity, the punched end face, and part of the adjacent background.

[0058] In this embodiment of the invention, in step S3, a binary image connected component labeling algorithm (such as the Two-Pass algorithm or the seed filling method) is used to extract all connected components in the candidate target region mask. The pixel area (unit: pixels) of each connected component is calculated, and small regions with pixel areas smaller than a first preset threshold (e.g., 500 pixels, which is calculated based on the image resolution and the actual size of the front end of the measuring tape, typically ranging from 200 to 2000 pixels) are removed. These are usually noise or tiny reflective points. Among the remaining connected components, one or more with the largest pixel area are selected as candidate punching regions. If there are multiple candidate punching regions with similar pixel areas (e.g., more than 80% of the largest area), they can be further filtered according to their position (e.g., near the left or right side of the image) or shape (aspect ratio).

[0059] In this embodiment of the invention, the specific process of region expansion and boundary compensation in step S3 is as follows: Based on the candidate punching region, expand outward by a preset pixel range (for example, expand by 20 pixels to the left and right, and by 10 pixels to the up and down directions). Then, perform a morphological dilation operation on the expanded region (using a circular structuring element with a radius of 5 pixels) to compensate for the shrinkage of the actual punching boundary caused by filtering or opening operations. The final region obtained is the unique and effective punching region. Figure 5 As shown, its right boundary (or a boundary specified according to the design) will be used for subsequent punching position positioning.

[0060] In this embodiment of the invention, in step S4, the right boundary (or left boundary, depending on the tape measure design) of the punching area is first determined. Within the neighborhood of the right boundary (e.g., 50-300 pixels to the right, covering the height range of the punching area upwards / downwards), a scale line search area is defined. Then, local grayscale analysis is performed on the grayscale image within the scale line search area: a sliding window (e.g., window size 21×21 pixels) is used to calculate the mean and standard deviation of grayscale within the window. A dynamic threshold is set as mean grayscale ± k × standard deviation (k is usually 2-3). Pixels below the low threshold are identified as candidate points for the scale line, generating a binary map of candidate scale lines. Subsequently, the connected regions in the binary map of the scale line are extracted as candidate regions for the scale line.

[0061] In this embodiment of the invention, in step S4, the candidate regions for the scale lines are filtered and sorted as follows:

[0062] Remove short tick marks located within the top interference area. The top interference area is defined as the area within 10 to 30 pixels above the top of the punching area. If it contains candidate areas whose length is less than a preset length threshold (e.g., 15 pixels), they are discarded.

[0063] Retain candidate areas whose length and width meet the preset range. The length range can be set to 20~200 pixels and the width range to 2~10 pixels (depending on the image resolution and the actual scale size).

[0064] The retained valid scale lines are sorted from smallest to largest according to their vertical position coordinates (Y-axis direction);

[0065] Remove scale lines with abnormal spacing or width. Calculate the vertical spacing between adjacent scale lines. If the vertical spacing exceeds 1.5 times or is less than 0.5 times the average spacing, remove the corresponding scale line. If the width of a scale line exceeds 2 times or is less than 0.5 times the average width, also remove it.

[0066] After the above screening, a valid scale line sequence is obtained. If the number of scales in the valid scale line sequence is still greater than or equal to the minimum requirement (e.g., at least 3), then proceed to the next step; otherwise, trigger the missing compensation mechanism.

[0067] In this embodiment of the invention, in step S4, when there are missing scales in the effective scale line sequence, linear interpolation is performed based on the average spacing between adjacent scale lines to fill in the missing scale position. For example, if there are two consecutive effective scale line positions x_i and x_{i+2}, and x_{i+1} is missing in the middle, then the position to be filled in is (x_i + x_{i+2}) / 2. The interpolation is repeated until all missing positions are filled in, generating a complete effective scale line sequence.

[0068] In this embodiment of the invention, in step S5, the position of the punching area can be defined as the pixel coordinates of the right boundary (or left boundary, consistent with the theoretical reference definition) of the unique effective punching area obtained in step S3. The position of the effective scale line sequence can be selected as the left edge or center coordinates of the first complete scale line in the sequence. The punching position is compared with the theoretical reference value, which is defined as the standard relative offset of the first effective scale line (starting from the punching boundary). This theoretical reference value is stored in the system in advance. The deviation can be calculated by pixel deviation or converted to the actual physical size (millimeters). The preset allowable range is ±0.5 mm (the specific value can be set according to the product accuracy requirements, such as ±0.2 mm to ±1.0 mm). If the deviation is within the preset allowable range, the tape measure is determined to be a qualified product, and a pass signal is output; otherwise, it is determined to be a punching error defect, and a fail signal is output, and an alarm can be triggered optionally.

[0069] In this embodiment of the invention, if any of the following situations occur in step S5, it is also determined to be a punching error defect:

[0070] The deviation exceeds the preset allowable range;

[0071] No valid tick mark sequence was detected (i.e., the number of valid tick marks is 0).

[0072] The number of scales in the valid scale line sequence is less than the second preset threshold (e.g., 3 scales).

[0073] There is a significant spacing anomaly in the effective scale line sequence (even after interpolation, an equally spaced sequence cannot be formed).

[0074] Unable to identify punching regions (e.g., candidate punching regions are empty or too small).

[0075] In this embodiment of the invention, in step S5, the theoretical reference value can be determined based on the standard punching position in the tape measure design drawing. For example, if the design drawing specifies that the theoretical distance between the punching boundary and the first complete scale line is 2.5mm, then the theoretical reference value is this 2.5mm. Alternatively, in the actual production line, multiple (e.g., 30-50) qualified tape measure sample images verified by high-precision measuring equipment are collected, and the actual offset between the punching position and the first scale line of each qualified tape measure sample image is calculated. The average value is then used as the standard punching position. Both methods support pre-calibration and storage in the detection system.

[0076] In this embodiment of the invention, as alternative processing methods, in step S2, the low-frequency smoothing process can use bilateral filtering instead of Gaussian filtering to better preserve edge information; in step S4, the dynamic threshold segmentation can also use the Sauvola algorithm or the Niblack algorithm to improve the segmentation stability under uneven local contrast conditions; the entire detection process can be implemented on an industrial control computer in C++ or Python, and the processing time for a single frame image is recommended to be controlled within 50ms to meet the requirements of online real-time detection.

[0077] In the description of this invention, the terms "one embodiment," "some embodiments," "in this embodiment," "specific example," or "some examples," etc., refer to specific features, mechanisms, materials, or characteristics described in connection with that embodiment or example, which are included in at least one embodiment or example of the invention. In this specification, the illustrative expressions of the above terms do not necessarily refer to the same embodiment or example. Furthermore, the specific features, mechanisms, materials, or characteristics described may be combined in any suitable manner in one or more embodiments or examples. Moreover, without contradiction, those skilled in the art can combine and integrate the different embodiments or examples described in this specification, as well as the features of different embodiments or examples.

[0078] The above description is merely a specific embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the technical scope disclosed in the present invention should be included within the scope of protection of the present invention. Therefore, the scope of protection of the present invention should be determined by the scope of the claims.

Claims

1. A visual measurement method for the punching accuracy of the front end of a measuring tape, characterized in that, Includes the following steps: Step S1: Obtain the grayscale image of the front end of the measuring tape, perform gradient calculation to extract the structural change information in the vertical direction, and generate a vertical edge response map through threshold segmentation. Step S2: Perform multi-scale morphological processing on the vertical edge response map to obtain an enhanced structure map, and perform low-frequency smoothing, threshold segmentation and morphological closing operation on the grayscale image. Then, combine the enhanced structure map to generate a candidate target region mask. Step S3: Extract connected regions from the candidate target region mask, filter out candidate punching regions whose areas meet preset conditions, and determine the unique effective punching region through region expansion and boundary compensation. Step S4: Perform local grayscale analysis on the grayscale image in the neighborhood of the punching area, extract candidate regions for scale lines through dynamic thresholding, and filter and sort the scale lines in the candidate regions based on the consistency of length, position, spacing and width to obtain an effective scale line sequence. Step S5: Calculate the deviation between the punching position and the theoretical reference value based on the position of the punching area and the effective scale line sequence. If the deviation is within the preset allowable range, it is determined to be qualified; if it is not within the preset allowable range, it is determined to be a punching error defect.

2. The visual measurement method according to claim 1, characterized in that, The multi-scale morphological processing in step S2 includes sequentially performing longitudinal closure, longitudinal opening, and lateral closure operations to enhance the continuous structure in the vertical direction and suppress interference in the horizontal direction.

3. The visual measurement method according to claim 1, characterized in that, The process of generating the candidate target region mask in step S2 includes: The initial entity region is obtained by thresholding the grayscale image after low-frequency smoothing, and morphological closing operation is performed on the initial entity region to fill the internal holes, thereby generating the candidate target region mask.

4. The visual measurement method according to claim 1, characterized in that, The screening process for the candidate punching region in step S3 includes: Calculate the pixel area of ​​each connected region, remove connected regions with pixel areas smaller than a first preset threshold, and retain one or more connected regions with the largest pixel areas as candidate punching regions.

5. The visual measurement method according to claim 1, characterized in that, The region expansion and boundary compensation in step S3 include: Based on the candidate punching region, a preset pixel range is expanded outward, and the boundary contraction caused by morphological operations is compensated to restore the original structural continuity within the candidate punching region.

6. The visual measurement method according to claim 1, characterized in that, The extraction process of the candidate region for the scale line in step S4 includes: Within the neighborhood of the punching region, local grayscale analysis is performed on the grayscale image to obtain the local grayscale change rate. A candidate tick mark binary image is obtained through dynamic threshold segmentation, and the connected region in the candidate tick mark binary image is extracted as the tick mark candidate region.

7. The visual measurement method according to claim 1, characterized in that, The process of filtering and sorting the scale lines in step S4 includes: Short scale line candidate regions located in the top interference region are removed from the scale line candidate region. The scale line candidate regions whose length and width meet the preset range are retained as valid scale lines. Then, the valid scale lines are sorted according to their position coordinates, and the valid scale lines with abnormal spacing or width are removed to obtain the valid scale line sequence.

8. The visual measurement method according to claim 7, characterized in that, Step S4 also includes: When there are missing scales in the effective scale line sequence, linear interpolation is performed based on the average spacing between adjacent scale lines to fill in the missing scale positions.

9. The visual measurement method according to claim 1, characterized in that, The conditions for determining a punching error defect in step S5 include the deviation exceeding the preset allowable range, the failure to detect the effective scale line sequence, the number of scales in the effective scale line sequence being lower than the second preset threshold, the presence of abnormal spacing in the effective scale line sequence, and the inability to identify at least one of the punching areas.

10. The visual measurement method according to claim 1, characterized in that, The theoretical benchmark value in step S5 is determined based on the standard punching position in the tape measure design drawing or by statistically analyzing multiple qualified sample images to obtain the standard punching position.

Images