A drill blade surface defect detection method

Abstract

The application relates to the technical field of machine vision detection, and particularly discloses a drill blade surface defect detection method. The method comprises the following steps: collecting a drill blade surface image, extracting a blade surface contour and establishing a unified geometric reference system comprising a direction reference, a position reference and a geometric center; performing posture normalization on the blade surface based on the unified geometric reference system, and performing blade surface integrity detection; establishing a theoretical cutting edge model on the basis of the blade surface integrity, limiting an actual cutting edge detection range according to the theoretical cutting edge model, and obtaining cutting edge defect parameters by analyzing deviation data between the theoretical cutting edge model and an actual cutting edge contour; and simultaneously, completing groove angle, blade width, blade surface diameter and blade type parameter detection under the unified geometric reference system, and outputting detection results. The application improves the consistency of multi-parameter measurement by using a unified detection reference, and improves the stability of micro cutting edge defect detection by using a theoretical cutting edge model.

Description

Technical Field

[0001] This invention relates to the field of machine vision inspection technology, and in particular to a method for detecting defects on the cutting surface of a drill bit. Background Technology

[0002] Micro-industrial drills with dimensions of 0.2–2 mm are important tools for micro-scale hole machining and are widely used in high-end manufacturing fields such as PCB printed circuit boards, semiconductor packaging, precision electronic components, medical devices, aerospace, and precision instruments. Due to the small size and high machining accuracy requirements of micro-drills, the quality of their cutting edge directly affects the machining accuracy and performance of the product. Therefore, defect detection and dimensional inspection of the drill bit surface are usually required during the production process to ensure product quality.

[0003] Currently, the main method for inspecting the cutting surface of micro drill bits is machine vision inspection. This method involves using an industrial camera to capture images of the cutting surface and combining them with image processing algorithms to detect parameters such as cutting edge defects, flute angle, cutting width, and cutting surface diameter.

[0004] Regarding the aforementioned technologies, the inventors believe that the following defects exist: Existing machine vision inspection solutions typically perform positioning and measurement separately for different inspection items, lacking a unified geometric measurement benchmark between the various inspection items. This makes it easy for changes in the blade's posture, positioning errors, and contour extraction errors to affect the consistency of the inspection results. At the same time, existing blade defect detection methods mostly rely directly on the analysis of actual contours, lacking a theoretical model as a detection benchmark. This makes them susceptible to image noise and local edge fluctuations, resulting in insufficient stability and accuracy in the detection of minute defects. Summary of the Invention

[0005] To address the issues of inconsistent testing standards and insufficient stability in the detection of cutting edge defects in existing drill bit face inspection processes, this application provides a method for detecting drill bit face defects.

[0006] The present invention provides a method for detecting defects on the cutting surface of a drill bit, which adopts the following technical solution:

[0007] A method for detecting defects on the cutting surface of a drill bit includes the following steps:

[0008] S1. Acquire an image of the drill bit's cutting surface, extract the boundary of the cutting surface image, and establish a unified geometric reference system for the cutting surface. The unified geometric reference system includes the direction reference, position reference, and geometric center of the cutting surface.

[0009] S2. The pose of the blade image is normalized according to the unified geometric reference system, and a standard detection area corresponding to the blade is established. The blade integrity is detected in the standard detection area.

[0010] S3. When the blade integrity detection result meets the preset conditions, a theoretical cutting edge model is established based on the standard detection area, and the detection range of the actual cutting edge is limited based on the theoretical cutting edge model. The actual cutting edge contour is extracted within the detection range.

[0011] S4. Calculate the deviation data between the theoretical cutting edge model and the actual cutting edge profile, and determine the degree of deviation of the actual cutting edge relative to the theoretical cutting edge model based on the deviation data, and obtain the cutting edge defect parameters based on the degree of deviation.

[0012] S5. Using the unified geometric reference system as a common testing benchmark, measure the groove angle, blade width, blade diameter and blade shape parameters separately.

[0013] S6. Output the drill bit inspection results based on the cutting edge defect parameters and various measurement parameters.

[0014] Preferably, the unified geometric reference system obtains the direction reference by fitting the edge of the cutting edge, obtains the position reference by the enclosing region corresponding to the cutting edge contour, and establishes the geometric center based on the intersection characteristics of the cutting edges.

[0015] Preferably, the pose normalization of the tool face image based on the unified geometric reference system includes the following steps:

[0016] Calculate the attitude angle of the cutting face based on the stated direction reference;

[0017] The blade image is rotated and corrected according to the attitude angle to transform the blade to a preset detection attitude.

[0018] Preferably, the blade integrity detection includes:

[0019] Extract the continuous contour of the cutting edge;

[0020] The presence of a defect region exceeding a preset threshold is determined based on the continuity of the continuous contour and the area of ​​the missing region.

[0021] If a defective area is found, the subsequent testing will be terminated.

[0022] Preferably, the theoretical cutting edge model is established based on the fitting of the cutting edge of the tool face;

[0023] A local detection region is established based on the theoretical cutting edge model, so that the actual cutting edge contour is limited to the region near the theoretical cutting edge.

[0024] The local detection area covers the theoretical cutting edge and the actual cutting edge within a preset range.

[0025] Preferably, the actual cutting edge contour is extracted within the local detection area;

[0026] Calculate the deviation between the actual cutting edge profile and the corresponding sampling point positions of the theoretical cutting edge model;

[0027] The cutting edge deviation data is constructed based on several of the aforementioned deviation values.

[0028] Preferably, the cutting edge defect parameters are obtained based on the cutting edge deviation data, and the cutting edge defect parameters include at least the maximum deviation value.

[0029] Preferably, the groove angle is obtained by establishing a measurement reference line through the geometric center in the unified geometric reference system and at least two feature points on the blade profile, and based on the included angle between the measurement reference lines.

[0030] Preferably, the blade width is obtained based on the minimum enclosing area of ​​the main blade surface;

[0031] The diameter of the cutting edge is obtained based on the smallest enclosing circle of the cutting edge;

[0032] The blade shape is obtained based on the directional relationship between the main blade face and the secondary blade face.

[0033] Preferably, the unified geometric reference system is used throughout the entire process of tool face integrity detection, cutting edge defect detection, flute angle detection, tool width detection, tool face diameter detection, and tool shape detection, and all detection parameters share the unified geometric reference system as a common measurement benchmark.

[0034] The beneficial effects of this invention are as follows: By establishing a unified geometric reference system, this invention achieves the normalization of the tool face posture and the sharing of the same measurement benchmark among various detection items, reducing the impact of posture changes and positioning errors on the detection results; by establishing a theoretical cutting edge model to limit the actual cutting edge detection range, and combining the deviation data between the actual cutting edge contour and the theoretical cutting edge model for defect analysis, the stability and consistency of cutting edge defect detection are improved; at the same time, by completing the measurement of parameters such as flute angle, tool width, tool face diameter and tool shape under a unified geometric reference system, it is beneficial to reduce the errors caused by repeated positioning and improve the automation level and detection reliability of drill bit cutting edge detection. Attached Figure Description

[0035] Figure 1 This is the main flowchart of the drill bit cutting surface detection method in the embodiments of this application;

[0036] Figure 2 This is a schematic diagram of the steps of the drill bit cutting surface detection method in the embodiments of this application;

[0037] Figure 3 This is a system structure block diagram of the drill bit cutting surface detection system in the embodiments of this application;

[0038] Figure 4 This is a schematic diagram of the unified geometric reference system in the embodiments of this application;

[0039] Figure 5 This is a schematic diagram showing the deviation between the theoretical cutting edge model and the actual cutting edge in the embodiments of this application. Detailed Implementation

[0040] The following will combine Figures 1-5 The present invention will be further illustrated by the embodiments.

[0041] It should be understood that this embodiment is only used to explain the present invention and is not intended to limit the scope of protection of the present invention. Those skilled in the art can make various modifications or equivalent substitutions to the present invention without departing from the concept of the present invention, and all such modifications or substitutions should fall within the scope of protection of the present invention.

[0042] This embodiment discloses a method for detecting defects on the cutting surface of drill bits, which is mainly applicable to the quality inspection of the cutting surface of micro industrial drill bits with a specification of 0.2 to 2 mm. The objects to be inspected include PCB drill bits, precision carbide drill bits, and other micro drill bits with similar cutting surface structures, which will not be described in detail here.

[0043] Reference Figure 1 and Figure 2 This embodiment provides a method for detecting defects on the cutting edge of a drill bit. The overall detection process includes multiple steps such as image acquisition, image preprocessing, cutting edge positioning, cutting edge integrity detection, cutting edge and size detection, and detection result output, as detailed below:

[0044] S1. Acquire images of the drill bit's face, extract the boundaries of the face images, and establish a unified geometric reference system for the face. The unified geometric reference system includes the face's orientation reference, position reference, and geometric center.

[0045] S2. Normalize the pose of the tool face image based on a unified geometric reference system, establish a standard detection area corresponding to the tool face, and perform tool face integrity detection on the standard detection area.

[0046] S3. When the blade integrity test result meets the preset conditions, a theoretical cutting edge model is established based on the standard test area, and the actual cutting edge test range is limited based on the theoretical cutting edge model. The actual cutting edge contour is extracted within the test range.

[0047] S4. Calculate the deviation data between the theoretical cutting edge model and the actual cutting edge profile, and determine the degree of deviation of the actual cutting edge relative to the theoretical cutting edge model based on the deviation data. Obtain the cutting edge defect parameters based on the degree of deviation.

[0048] S5. Using a unified geometric reference system as a common testing standard, measure the groove angle, blade width, blade diameter, and blade shape parameters separately.

[0049] S6. Output drill bit inspection results based on cutting edge defect parameters and various measurement parameters.

[0050] Specifically, firstly, an industrial camera is used to acquire images of the drill bit's cutting surface, and the acquired raw images are preprocessed to improve the clarity of the cutting surface contour and the edge of the cutting edge, providing stable image data for subsequent image analysis.

[0051] Subsequently, the drill bit to be detected is extracted from the preprocessed image, and it is determined whether the drill bit can be correctly extracted from the current image. If the drill bit cannot be correctly extracted, it indicates that there is an abnormality in the placement of the drill bit, an abnormality in the imaging, or severe damage to the drill bit. The system directly outputs an abnormality signal and sends an NG result to the PLC control system, which then transports the corresponding product to the manual processing station.

[0052] On the other hand, after successfully extracting the blade surface image, the integrity of the blade surface is checked. If there are large-area defects, broken contours, or other abnormalities that affect subsequent inspections, the subsequent inspections are stopped, and an NG result is sent to the PLC control system for manual re-inspection or processing. When the blade surface integrity meets the inspection requirements, the blade surface inspection stage is entered, and the degree of edge defects, groove angle, blade width, blade surface diameter, and blade shape parameters are measured.

[0053] After completing all inspection items, the system integrates all inspection parameters to obtain the final inspection result. When the inspection result meets the preset requirements, the system outputs the corresponding inspection data and result for the drill bit; when any inspection parameter exceeds the allowable range, an NG signal is sent to the PLC control system, which then sorts the non-conforming products and initiates the subsequent manual processing flow.

[0054] In this embodiment, the first step in the entire inspection process is to locate and screen the blade surface of the object to be inspected. Only blade surfaces that meet the inspection conditions are subjected to subsequent edge and size inspections to avoid abnormal images from entering the subsequent inspection process. At the same time, a unified geometric reference system is used as the common measurement benchmark for all inspection items, so that edge defect inspection, flute angle inspection, blade width inspection, blade surface diameter inspection and blade shape inspection share the same coordinate system, thereby improving the consistency and inspection stability among multiple inspection items.

[0055] Reference Figure 2 and Figure 3 The drill bit cutting surface detection system used in this embodiment mainly includes an image acquisition module, an image processing module, and a detection result output module.

[0056] The image acquisition module includes an industrial camera, an industrial lens, and a light source assembly, used to acquire high-definition images of the drill bit's cutting surface. To ensure good imaging quality of the cutting edge and contour, an industrial lens with the appropriate magnification can be selected according to the drill bit specifications, and it can be used in conjunction with a coaxial light source, a ring light source, or other lighting methods that meet the inspection requirements to improve the contrast between the cutting surface contour and the background.

[0057] Furthermore, the image processing module is used to complete the tool face image analysis. It mainly includes a unified geometric reference system establishment module, a tool face integrity detection module, a theoretical cutting edge model establishment module, a cutting edge contour extraction module, a deviation data analysis module, a flute angle detection module, a tool width detection module, a tool face diameter detection module, and a tool shape detection module.

[0058] Correspondingly, the test result output module is used to summarize the test parameters obtained by each test module and output the drill bit test results according to the preset test standards. When the test parameters meet the preset requirements, the current drill bit can be determined as a qualified product; when any test parameter exceeds the allowable range, the current drill bit is determined as a non-qualified product, and the corresponding non-qualification type and test data are output.

[0059] Furthermore, the aforementioned image processing module can be deployed in an industrial control computer, a vision controller, or other computing devices with image processing capabilities. In this embodiment, the image processing process can be implemented using HALCON (a vision algorithm library), or the same image processing process can be completed using OpenCV (an open-source computer vision library) or other image processing platforms; this invention does not limit this.

[0060] On the other hand, since the placement and rotation of the drill bit when it enters the inspection station are usually somewhat random, there may be positional offsets, different rotation angles, and imaging errors between different drill bits. If the cutting edge inspection and dimensional measurement are performed directly based on the acquired original images, each inspection item often requires establishing a separate inspection benchmark, which not only increases the number of repeated positioning steps but also easily leads to inconsistent measurement results between different inspection items due to the accumulation of positioning errors. Therefore, this embodiment first establishes a unified geometric reference system for the cutting edge.

[0061] Reference Figure 2 and Figure 4 Specifically, after the industrial camera acquires an image of the drill bit's face, edge extraction is first performed on the image to obtain the overall contour information of the face. A directional reference is obtained by fitting straight lines to the face edges; a bounding region is established based on the overall contour of the face, and this bounding region is used as the position reference. Simultaneously, the geometric center of the face is determined based on the intersection of the main face and the secondary face. For example, a contour intersection calculation algorithm (such as `intersection_contours_xld` in HALCON) can be used to obtain the intersection point between the main face contour and the secondary face contour, and this intersection point can be used as the geometric center of the face; alternatively, correction can be performed using the center of a circular positioning area to improve the positioning accuracy of the geometric center. Correspondingly, the directional reference, the position reference, and the geometric center together constitute a unified geometric reference system for the face.

[0062] After establishing a unified geometric reference system, the attitude angle of the current cutting face is calculated based on the direction reference, and the cutting face image is rotated and corrected according to the attitude angle, so that the drill cutting face acquired under different attitudes is converted to a unified detection attitude, thereby realizing the normalization of the cutting face attitude.

[0063] After attitude normalization, a standard detection area corresponding to the blade face is established based on the position reference, and blade face integrity is detected within the standard detection area. Specifically, the continuity of the blade face contour within the standard detection area is analyzed to determine whether there are obvious breaks, missing parts, or large-area defects in the blade face contour. For example, the edges_sub_pix (sub-pixel edge extraction algorithm) in HALCON can be used to extract the overall contour of the blade face, and the presence of large-area defects in the blade corner area can be detected based on the extracted overall contour.

[0064] When there are defective areas on the cutting edge exceeding the preset threshold, the current drill bit cutting edge integrity test is deemed unqualified, and subsequent edge defect detection and dimensional detection are terminated; when the cutting edge integrity meets the detection requirements, the theoretical edge model establishment step is entered.

[0065] Before proceeding to the theoretical cutting edge model establishment step, the actual cutting edge region can be further extracted based on the aforementioned circular positioning region. For example, the actual cutting edge region can be extracted within the circular positioning region using binary_threshold (image binarization segmentation algorithm), and the actual cutting edge region can be used as the basis for subsequent cutting edge defect detection, flute angle detection, tool width detection, and tool surface diameter detection.

[0066] In this embodiment, the cutting edge extraction can be performed using a sub-pixel edge extraction method to obtain a high-precision edge contour; the orientation reference can be obtained by fitting a straight line to the cutting edge; the enclosing region corresponding to the cutting edge contour can be calculated based on the overall cutting edge contour; the attitude angle can be calculated based on the orientation reference and combined with image rotation to complete attitude normalization.

[0067] For example, the fit_line_contour_xld (contour line fitting algorithm) in HALCON can be used to fit the edge of the cutting edge, the smallest_circle (minimum bounding circle calculation algorithm) can be used to obtain the bounding region corresponding to the cutting edge contour, the angle_lx (line direction angle calculation algorithm) can be used to calculate the cutting edge attitude angle, and the cutting edge attitude can be normalized by combining affine transformation. Of course, other image processing algorithms that can achieve the same function can also be used, and this invention does not limit them.

[0068] Therefore, after establishing a unified geometric reference system, the subsequent establishment of theoretical cutting edge models, extraction of actual cutting edge contours, analysis of cutting edge defects, and detection of flute angle, tool width, tool face diameter, and tool shape are all completed based on the unified geometric reference system. This allows each detection item to share the same measurement benchmark, avoids the cumulative error caused by repeatedly establishing reference coordinates for different detection items, and provides a unified spatial reference for the subsequent detection of cutting edge defects under the constraints of the theoretical cutting edge model.

[0069] After the cutter face integrity test is completed and the cutter face integrity is qualified, this embodiment begins to perform defect detection on the drill bit cutting edge.

[0070] Since the drill bit face includes not only the cutting edge but also the edge, groove boundary, and other geometric structures, directly extracting the actual cutting edge contour from the entire cutting edge area is easily affected by the background area, cutting edge texture, and local noise, leading to unstable cutting edge contour extraction results. Therefore, this embodiment first establishes a theoretical cutting edge model and then limits the detection range of the actual cutting edge based on the theoretical cutting edge model. In this embodiment, the theoretical cutting edge model can be pre-established according to the drill bit model or generated in real time based on the current geometric features of the cutting edge.

[0071] Specifically, the cutting edges of the main cutting surface and the secondary cutting surface are extracted based on the geometric features of the cutting surface, and the cutting edges of the main cutting surface and the secondary cutting surface are fitted with straight lines to obtain the theoretical cutting edge models of the corresponding cutting edges. In one embodiment, the cutting edge straight lines of the main cutting surface and the secondary cutting surface can be obtained separately, and the two cutting edge straight lines are used as the theoretical cutting edge models of the corresponding cutting edges.

[0072] The theoretical cutting edge model describes the theoretical position and extension direction of the drill bit's cutting edge in a defect-free state, and serves as a reference for subsequent actual cutting edge inspection. After establishing the theoretical cutting edge model, a local inspection area corresponding to the cutting edge is established with the theoretical cutting edge model as the center.

[0073] Specifically, the minimum bounding region of the corresponding cutting edge can be calculated based on the theoretical cutting edge model. In one embodiment, the minimum bounding region is the minimum circumscribed rectangle corresponding to the theoretical cutting edge model, and the minimum circumscribed rectangle is extended outward by a preset number of pixels to form a local detection region covering the theoretical cutting edge and the actual cutting edge fluctuation range.

[0074] The local detection area covers the theoretical cutting edge and the actual cutting edge position that may exist within the preset range, so that the subsequent actual cutting edge contour is extracted only within the local detection area, instead of performing edge search on the entire cutting edge area, thereby reducing the influence of the background area and other cutting edge structures on the cutting edge contour extraction result.

[0075] Subsequently, the actual cutting edge contour is extracted in the local detection area using a sub-pixel edge extraction method to obtain continuous contour data corresponding to the actual cutting edge.

[0076] In this embodiment, the theoretical cutting edge model is used to limit the detection range of the actual cutting edge, so that the actual contour is always established around the theoretical cutting edge, avoiding the entire tool surface contour from directly participating in the cutting edge detection, improving the stability of the actual cutting edge contour extraction, and providing a reliable basis for subsequent deviation data analysis.

[0077] After extracting the actual cutting edge contour, this embodiment further calculates the deviation data between the actual cutting edge contour and the theoretical cutting edge model to obtain the cutting edge defect parameters.

[0078] Reference Figure 5 Specifically, under a unified geometric reference system, a one-to-one correspondence is established between the theoretical cutting edge model and the actual cutting edge profile, and the theoretical cutting edge model is continuously sampled according to a preset sampling interval.

[0079] Subsequently, the distance between the actual cutting edge profile and the theoretical cutting edge model at each sampling point is calculated, where each distance value represents the deviation of the actual cutting edge from the theoretical cutting edge model at the corresponding position.

[0080] The deviation values ​​corresponding to all sampling points together constitute the cutting edge deviation data. For example, the cutting edge deviation data can be represented as a difference array formed between the theoretical cutting edge model and the actual cutting edge profile, where each element in the difference array represents the deviation of the actual cutting edge from the theoretical cutting edge model at the corresponding sampling position.

[0081] Under normal conditions, the actual cutting edge profile basically coincides with the theoretical cutting edge model, and the deviation value of each sampling point remains within the allowable range. When there are notches, chipping, or wear in the local area of ​​the cutting edge, the actual cutting edge profile in the corresponding area will be significantly offset from the theoretical cutting edge model, causing the deviation value of the corresponding sampling point to increase significantly.

[0082] Therefore, this embodiment analyzes the actual cutting edge defects based on all deviation data, and determines the degree of deviation of the actual cutting edge from the theoretical cutting edge model based on the deviation data.

[0083] Furthermore, the maximum deviation value among all deviation values ​​can be used as the cutting edge defect parameter. In one embodiment, the maximum value in the difference array can be used as the cutting edge defect parameter; when there is a notch in the actual cutting edge, the deviation value at the corresponding position in the difference array increases significantly, so the maximum value in the difference array can characterize the size of the notch. Therefore, when the maximum deviation value exceeds a preset threshold, it is determined that there is a cutting edge defect at the corresponding position; otherwise, the cutting edge inspection is deemed qualified.

[0084] For example, the above measurement scheme is illustrated with the following example:

[0085] First, the straight lines of the two cutting edges can be fitted in the cutting edge area using the get_metrology_object_result_contour (caliper measurement algorithm). Then, the minimum bounding rectangle of the straight line is calculated using geometric calculation. The rectangle is then expanded outward by a few pixels. Finally, the contour of the cutting edge is calculated using the sub-pixel precision algorithm edges_sub_pixel. After the contour is calculated, the correspondence between the contour and the theoretical cutting edge model is established, and the distance between corresponding points is calculated to form a difference array. For example, the difference array between the contour and the straight line is calculated using the geometric relationship calculation algorithm intersection_lines. Then, the maximum value is found in the array. This maximum value is the notch of the corresponding cutting edge, i.e., the maximum deviation value. Furthermore, the location of the cutting edge defect can be determined based on the sampling point position corresponding to the maximum deviation value.

[0086] On the other hand, in addition to the maximum deviation value mentioned above, this embodiment can also calculate the average deviation value, continuous deviation length, deviation area width or other parameters that can characterize the degree of cutting edge defects based on the deviation data, and select the corresponding parameters as the basis for judging cutting edge defects according to the detection requirements. This invention does not limit this.

[0087] After completing the edge defect detection, this embodiment further detects the groove angle of the drill bit face. Since the cutter face has completed the establishment of a unified geometric reference system and attitude normalization, each drill bit maintains the same detection attitude, and the unified geometric reference system can be directly used as the measurement benchmark for the groove angle.

[0088] Specifically, at least two feature points are extracted from the tool face profile to characterize the groove angle, and a corresponding measurement baseline is established in conjunction with the geometric center of the tool face. The feature points can be determined based on the geometric features of the tool face profile, such as the highest point, the rightmost point, or other boundary features that reflect the groove angle.

[0089] For example, the profile of the main cutting face can be converted into an array of point coordinates. The highest point of the main cutting face is determined by the minimum value of the longitudinal coordinate, and the rightmost point of the main cutting face is determined by the maximum value of the transverse coordinate. Then, two measurement reference lines are established by connecting the geometric center of the cutting face with the highest point and the geometric center of the cutting face with the rightmost point, respectively. The angle between the two measurement reference lines is calculated using angle_ll (angle between two straight lines) as the groove angle.

[0090] Furthermore, in the above process, since the unified geometric reference system has already completed the normalization of the tool face posture, the minimum enclosing region is always established in a unified direction, ensuring a consistent tool width measurement benchmark among different drill bits, thereby avoiding tool width measurement errors caused by different drill bit rotation angles. The minimum enclosing region can be calculated based on the tool face profile, and the tool width dimension can be obtained based on the width corresponding to the minimum enclosing region.

[0091] In one implementation, the main cutting face area can be divided into half-size regions along the center of the cutting face, and a minimum bounding rectangle can be established based on this half-size region. For example, the minimum bounding rectangle of the half-size region can be obtained using `smallest_rectangle1` in HALCON. Then, the distance between the top and bottom edges of the minimum bounding rectangle is calculated; this distance is the cutting width of the main cutting face. Alternatively, other algorithms that can obtain the minimum bounding area can be used to measure the cutting width; this invention does not limit this method. After obtaining the cutting width dimension, it can be determined whether the current drill bit width meets the inspection requirements based on the product's preset tolerance range.

[0092] After completing the tool width detection, this embodiment further completes the tool face diameter detection. Specifically, the overall contour of the tool face is extracted under a unified geometric reference system, and a corresponding minimum bounding circle is established based on the overall contour of the tool face. The center of the minimum bounding circle serves as the positioning center of the tool face, and its diameter serves as the detection diameter of the drill bit face.

[0093] Since the cutter face diameter is always measured based on the smallest bounding circle established under a unified geometric reference system, detection deviations caused by changes in drill bit posture or positioning errors can be avoided, thus improving the stability of cutter face diameter detection. In this embodiment, a smallest bounding circle calculation algorithm can be used to calculate the smallest bounding circle corresponding to the cutter face profile. For example, the smallest_circle_xld algorithm in HALCON can be used to obtain the smallest bounding circle corresponding to the cutter face, and then the cutter face diameter is calculated by multiplying the output smallest circumscribed circle radius by 2.

[0094] After measuring the cutting face diameter, it can be determined whether the current drill bit meets the product processing requirements based on the product dimensional tolerance range. Furthermore, under a unified geometric reference system, the direction information of the main cutting face and the secondary cutting face are obtained separately, and a cutting tool shape determination model is established based on the direction relationship between the main and secondary cutting faces. In this embodiment, the cutting tool shape of the current drill bit can be comprehensively determined based on the included angle, relative position relationship, and contour distribution characteristics of the main and secondary cutting faces.

[0095] For example, when the directional relationship between the main cutting face and the secondary cutting face meets the preset standard, the current drill bit shape is determined to meet the requirements; when the directional relationship between the main cutting face and the secondary cutting face exceeds the preset range, the cutting shape is determined to be abnormal. Since the cutting shape detection is based on a unified geometric reference system, it can ensure that different drill bits use the same judgment criteria, thus improving the consistency of the cutting shape detection results.

[0096] In this embodiment, a contour direction fitting algorithm can be used to obtain the direction information of the main cutting face and the secondary cutting face, and the cutting face determination can be completed by combining it with a preset cutting face model. This invention does not limit this. For example, the smallest_rectangle2_xld can be used to calculate the cutting edge direction angles corresponding to the main cutting face and the secondary cutting face, and the current drill bit is determined to be a large-end cutter or a small-end cutter based on the angular relationship between the two direction angles. Large-end cutters and small-end cutters correspond to different product specifications of cutting face classification. After completing the cutting edge defect detection, flute angle detection, cutting width detection, cutting face diameter detection, and cutting face detection, this embodiment summarizes and analyzes the detection parameters obtained by each detection module and outputs the final detection results.

[0097] Specifically, the cutting edge defect parameters, flute angle, tool width, tool face diameter, and tool shape parameters are compared with the corresponding product inspection standards. When all inspection parameters are within the preset allowable range, the current drill bit is deemed to have passed inspection, and the corresponding inspection data is output. When any inspection parameter exceeds the preset allowable range, the current drill bit is deemed to have failed inspection, and the corresponding abnormal parameter is recorded. In this embodiment, the inspection results can be displayed in real time on a display terminal, or uploaded to a host computer, MES system, or quality management system for data storage and statistical analysis.

[0098] Simultaneously, the test results can be sent to the PLC control system, which then controls the sorting mechanism to automatically separate qualified and unqualified products, thus automating the drill bit inspection process. Furthermore, to facilitate product quality traceability, this embodiment can establish a correspondence between test data and product numbers, storing the test results for each drill bit's corresponding cutting edge defect parameters, flute angle, tool width, tool face diameter, and tool type, forming a complete test record. This provides data support for subsequent product quality analysis and process optimization.

[0099] The implementation principle of the drill bit cutting edge defect detection method in this embodiment is as follows: First, the drill bit cutting edge image is acquired, and a unified geometric reference system for the cutting edge is established, which serves as the common measurement benchmark for each detection item. Then, a theoretical cutting edge model is established, and the detection range of the actual cutting edge is limited based on the theoretical cutting edge model. The cutting edge defect parameters are obtained by analyzing the deviation data between the theoretical cutting edge model and the actual cutting edge contour. Finally, the parameters such as the cutting edge groove angle, cutting edge width, cutting edge diameter, and cutting edge shape are detected under the unified geometric reference system, and the drill bit detection results are output by integrating all detection parameters.

[0100] The above are all preferred embodiments of the present invention and are not intended to limit the scope of protection of this application. Any equivalent substitutions or modifications made to the method steps, algorithm implementation or data processing methods of this application under the technical concept of this application shall be covered within the scope of protection of this application.

Claims

1. A method for detecting defects on the cutting surface of a drill bit, characterized in that, Includes the following steps: S1. Acquire an image of the drill bit's cutting surface, extract the boundary of the cutting surface image, and establish a unified geometric reference system for the cutting surface. The unified geometric reference system includes the direction reference, position reference, and geometric center of the cutting surface. S2. The pose of the blade image is normalized according to the unified geometric reference system, and a standard detection area corresponding to the blade is established. The blade integrity is detected in the standard detection area. S3. When the blade integrity detection result meets the preset conditions, a theoretical cutting edge model is established based on the standard detection area, and the detection range of the actual cutting edge is limited based on the theoretical cutting edge model. The actual cutting edge contour is extracted within the detection range. S4. Calculate the deviation data between the theoretical cutting edge model and the actual cutting edge profile, and determine the degree of deviation of the actual cutting edge relative to the theoretical cutting edge model based on the deviation data, and obtain the cutting edge defect parameters based on the degree of deviation. S5. Using the unified geometric reference system as a common testing benchmark, measure the groove angle, blade width, blade diameter and blade shape parameters separately. S6. Output the drill bit inspection results based on the cutting edge defect parameters and various measurement parameters.

2. The method for detecting defects on the cutting surface of a drill bit according to claim 1, characterized in that, The unified geometric reference system obtains the directional reference by fitting the edge of the cutting edge, obtains the position reference by the enclosing region corresponding to the cutting edge contour, and establishes the geometric center based on the intersection characteristics of the cutting edges.

3. The method for detecting defects on the cutting surface of a drill bit according to claim 2, characterized in that, The pose normalization of the blade image based on the unified geometric reference system includes the following steps: Calculate the attitude angle of the cutting face based on the stated direction reference; The blade image is rotated and corrected according to the attitude angle to transform the blade to a preset detection attitude.

4. The method for detecting defects on the cutting surface of a drill bit according to claim 1, characterized in that, The blade integrity detection includes: Extract the continuous contour of the cutting edge; The presence of a defect region exceeding a preset threshold is determined based on the continuity of the continuous contour and the area of ​​the missing region. If a defective area is found, the subsequent testing will be terminated.

5. The method for detecting defects on the cutting surface of a drill bit according to claim 1, characterized in that: The theoretical cutting edge model is established based on the fitting of the cutting edge of the cutting surface; A local detection region is established based on the theoretical cutting edge model, so that the actual cutting edge contour is limited to the region near the theoretical cutting edge. The local detection area covers the theoretical cutting edge and the actual cutting edge within a preset range.

6. The method for detecting defects on the cutting surface of a drill bit according to claim 5, characterized in that: Extract the actual cutting edge contour within the local detection area; Calculate the deviation between the actual cutting edge profile and the corresponding sampling point positions of the theoretical cutting edge model; The cutting edge deviation data is constructed based on several of the aforementioned deviation values.

7. The method for detecting defects on the cutting surface of a drill bit according to claim 6, characterized in that, The cutting edge defect parameters are obtained based on the cutting edge deviation data, and the cutting edge defect parameters include at least the maximum deviation value.

8. The method for detecting defects on the cutting surface of a drill bit according to claim 1, characterized in that, The groove angle is obtained by establishing a measurement baseline through the geometric center in the unified geometric reference system and at least two feature points on the blade profile, and based on the included angle between the measurement baselines.

9. The method for detecting defects on the cutting surface of a drill bit according to claim 1, characterized in that: The blade width is obtained based on the minimum enclosed area of ​​the main blade face; The diameter of the cutting edge is obtained based on the smallest enclosing circle of the cutting edge; The blade shape is obtained based on the directional relationship between the main blade face and the secondary blade face.

10. The method for detecting defects on the cutting surface of a drill bit according to claim 1, characterized in that, The unified geometric reference system is used throughout the entire process of tool face integrity detection, cutting edge defect detection, flute angle detection, tool width detection, tool face diameter detection, and tool shape detection. All detection parameters share the unified geometric reference system as a common measurement benchmark.

Images