A method for measuring the surface topography of a grinding wheel

Abstract

This invention discloses a three-dimensional measurement method for the surface morphology of grinding wheels, relating to the field of grinding inspection technology. The method involves fixing the grinding wheel to be tested on a rotatable grinding wheel clamping table, and acquiring a sequence of circumferential, multi-field, and multi-focal surface images of the grinding wheel's circumferential surface using an industrial camera that can move up and down and left and right. After image preprocessing, the pixel focus height is determined based on the sharpness response, and the grinding wheel's circumferential curved surface reference is subtracted to generate a three-dimensional height field of the grinding wheel surface. Then, by combining the three-dimensional height, grayscale texture, and topological connectivity, the abrasive grains, binder, and pore regions are separated, and the abrasive grain exposure, exposure volume, effective abrasive grain spacing, pore connectivity, and texture features are extracted. Finally, the results are compared with benchmark topographic data to output the grinding wheel surface topography evaluation results and suggestions for dressing or replacement. This method can be used for grinding wheel selection, dressing quality verification, and accurate judgment of grinding conditions.

Description

Technical Field

[0001] This invention relates to the field of grinding inspection technology, specifically a three-dimensional measurement method for the surface morphology of a grinding wheel. Background Technology

[0002] In precision grinding, the grinding wheel is the specific tool directly involved in material removal. The surface of all parts used in precision grinding comprises three parts: abrasive grains, bonding agent, and pores. The abrasive grains perform the cutting action, the bonding agent provides retention, and the pores handle chip removal and heat dissipation. Abrasive grains affect workpiece surface quality, grinding force, grinding heat, the timing of wheel dressing, and the stability of the grinding process. Therefore, it is necessary to inspect the surface morphology of the grinding wheel in aspects such as wheel selection, dressing quality verification, grinding condition diagnosis, and process maintenance. Currently, in industrial settings, contact measurement, imprint observation, photoelectric detection, line array camera inspection, microscopic observation, or profilometry are commonly used to obtain information about the grinding wheel surface to determine abrasive grain distribution, abrasive grain shedding, wheel wear, and dressing status.

[0003] Chinese patent application CN108426537A discloses a method and system for rapid full-field inspection of grinding wheels in situ based on a linear array camera, indicating that the technical field involved is the inspection of two-dimensional surface morphology of grinding wheels. The paper proposes that a linear array camera can acquire two-dimensional surface morphology information of the grinding wheel across the entire field. The method includes a three-axis translation mechanism, a linear array camera, a light source, and an offline inspection platform. The linear array camera is mounted on the three-axis translation mechanism and can move up, down, left, right, and back and forth relative to the grinding wheel. The light source illuminates the area of ​​the grinding wheel to be inspected. The offline inspection platform is equipped with a lifting platform and an electrically controlled rotary table. The grinding wheel can be mounted on the electrically controlled rotary table using a centering shaft and a nut. During operation, the three-axis translation mechanism is controlled to align the linear array camera with the area of ​​the grinding wheel to be inspected, ensuring that the direction of the camera's photosensitive element is parallel to the central axis of the grinding wheel. Then, the camera's imaging speed is coupled with the grinding wheel's rotation speed to obtain a sequence of images. The abrasive grains in the sequence of images are then segmented from the background to obtain two-dimensional information about the abrasive grains, allowing for the determination of whether abrasive grains have detached and their coordinates.

[0004] The paper also discloses that operations such as grayscale nonlinear transformation, linear sharpening filtering, clustering, threshold binarization, morphological processing, and feature extraction can be performed on sequential images to obtain two-dimensional parameters such as the area, diameter, spacing, density distribution, and coordinates of a single abrasive grain.

[0005] The aforementioned existing technologies can perform two-dimensional morphology detection across the entire surface of the grinding wheel and can be used for in-situ detection, but they have certain limitations when evaluating the actual working topography of the grinding wheel. The two-dimensional morphology detection results of the grinding wheel mainly come from two-dimensional images and their segmentation results. Parameters such as abrasive grain area, grain size, coordinates, and spacing are essentially projection information in the image plane. The grinding wheel surface is a three-dimensional rough surface with circumferential curvature. Furthermore, the cutting ability of the abrasive grains also depends on their exposure height relative to the binder surface, the volume of local protrusions, and the depth of pores. During optical imaging, the edges of the abrasive grains on the grinding wheel surface are prone to high-brightness reflection, while dark areas are easily generated at the bottom of the pores, and the binder surface may produce low-contrast textures. When these phenomena are compressed into a two-dimensional grayscale image, high-brightness reflections, shadow boundaries, and the actual abrasive grain contours are easily mixed. Moreover, because the curvature of the grinding wheel's circumferential surface causes different circumferential positions to exhibit scale and viewing angle changes on the imaging plane, relying solely on two-dimensional image stitching and grayscale segmentation inevitably leads to problems such as abrasive grain adhesion, misjudgment of pores, or insufficient evaluation of local morphology.

[0006] The aforementioned problems affect the judgment of grinding wheel dressing quality, the identification of abrasive grain shedding, and the evaluation of pore blockage, thus impacting grinding condition maintenance. Therefore, the core technical problem that urgently needs to be solved is: how to obtain detection data that can represent the true topographical state of the grinding wheel surface under the conditions of coexisting grinding wheel circumferential curvature, abrasive grain reflection, pore dark areas, and low-contrast texture of the bond agent. Summary of the Invention

[0007] (a) Technical problems to be solved

[0008] To address the shortcomings of existing technologies, this invention provides a three-dimensional measurement method for the surface topography of grinding wheels. The method involves acquiring a sequence of images of the circumferential partitions, multiple fields of view, and multiple focal planes of the grinding wheel's circumferential surface using an industrial camera capable of vertical and horizontal movement. After image preprocessing, the pixel focus height is determined based on the sharpness response, and the grinding wheel's circumferential curved surface reference is subtracted to generate a three-dimensional height field of the grinding wheel surface. Then, by combining the three-dimensional height, grayscale texture, and topological connectivity, the abrasive grains, binder, and pore regions are separated, and the abrasive grain exposure, exposure volume, effective abrasive grain spacing, pore connectivity, and texture features are extracted. Finally, the method is compared with benchmark topographic data to output the grinding wheel surface topography evaluation results and suggestions for repair or replacement. This method solves the technical problems described in the background art.

[0009] (II) Technical Solution

[0010] To achieve the above objectives, the present invention provides the following technical solution:

[0011] A three-dimensional measurement method for the surface morphology of a grinding wheel, executed by a three-dimensional measurement device for the surface morphology of a grinding wheel, includes: controlling a rotatable grinding wheel clamping stage to step along a circumferential sector, controlling an optical imaging unit to move along the width direction of the grinding wheel and the optical axis direction of the lens, and acquiring a multi-focal image sequence with sector markers, strip markers and focal plane height markers;

[0012] Calculate the sharpness response of the same pixel in a multi-focal image sequence at different focal plane heights, determine the focus height of the pixel, and subtract the reference of the grinding wheel circumferential surface to generate a three-dimensional height field of the grinding wheel surface unfolding.

[0013] Based on the three-dimensional height field and grayscale texture map corresponding to the coordinates of the grinding wheel surface, the abrasive region, the binder region and the pore region are separated, and the abrasive exposure features, effective abrasive spacing features and pore connectivity features are extracted.

[0014] The surface geomorphology evaluation results of the grinding wheel are output by comparing the abrasive grain exposure characteristics, effective abrasive grain spacing characteristics, and pore connectivity characteristics with the benchmark geomorphological data.

[0015] Furthermore, the sharpness response of the same pixel at different focal plane heights in a multi-focal plane image sequence is calculated, including:

[0016] The images of each focal plane under the same sector and the same strip are converted to grayscale, dark field subtraction, flat field correction and edge-preserving filtering are performed; edge response and gradient response are extracted in the local neighborhood of the same pixel to form sharpness response; focus height is generated according to the peak position of sharpness response and sharpness confidence is generated simultaneously.

[0017] Furthermore, deducting the grinding wheel circumferential surface reference includes:

[0018] The focus height is converted to the initial height; the grinding wheel circumferential surface reference is generated based on the nominal radius of the grinding wheel, the unfolded coordinates of the sector center, and the unfolded coordinates of the pixels; the grinding wheel circumferential surface reference is subtracted from the initial height to obtain the terrain height;

[0019] Based on the overlapping areas corresponding to adjacent sector identifiers and adjacent strip identifiers, the terrain height is stitched together to form a three-dimensional height field unfolded on the grinding wheel surface.

[0020] Furthermore, the abrasive region, binder region, and pore region are separated, including:

[0021] Based on the three-dimensional height field, grayscale texture map, and sharpness confidence of the grinding wheel surface, abrasive evidence, bond evidence, and porosity evidence are generated; candidate abrasive regions are formed according to the abrasive evidence, porosity regions are formed according to the porosity evidence, and bond regions are formed according to the bond evidence.

[0022] The adhered candidate abrasive grain regions are divided into effective abrasive grain regions according to the height valley line and the boundary of the pore region; after spindle correction, the abrasive grain exposure features and effective abrasive grain spacing features are extracted from the effective abrasive grain regions.

[0023] Furthermore, the optical imaging unit has a preset imaging mode, which is either an area array imaging mode or a linear array imaging mode. When the preset imaging mode is an area array imaging mode, the optical imaging unit performs a single exposure for each circumferential sector and each width strip to generate a corresponding focal plane image.

[0024] When the preset imaging mode is linear array imaging mode, the angle feedback of the rotatable grinding wheel holding stage is synchronized with the line scan trigger signal, generating the corresponding focal plane image line by line.

[0025] Furthermore, when generating the three-dimensional height field of the grinding wheel surface, height registration is performed on the overlapping areas of adjacent local height fields;

[0026] When two height values ​​exist at the same spatial location in the overlapping area, the height value with higher clarity and confidence is written into the grinding wheel surface to unfold a three-dimensional height field.

[0027] When there are no two height values ​​at the same spatial location in the overlapping area, the existing height value at that spatial location is retained and written into the three-dimensional height field unfolded on the grinding wheel surface.

[0028] Furthermore, the benchmark geomorphological data includes at least one of the following: new grinding wheel benchmark geomorphological data, qualified modified geomorphological data, and historical geomorphological data of the same grinding wheel;

[0029] The comparison is aligned according to the sector identifier and the strip identifier; when the abrasive grain exposure feature is lower than the reference range and the pore connectivity feature is within the reference range, the abrasive grain dullness evaluation is output; when the pore connectivity feature is lower than the reference range, the pore blockage evaluation is output; when none of the above conditions are met, the normal evaluation is output.

[0030] Furthermore, when outputting the evaluation results of the surface topography of the grinding wheel, the three-dimensional measurement equipment for the surface topography of the grinding wheel generates a grinding wheel circumference unfolding diagram, a three-dimensional height pseudo-color diagram, abrasive boundary markers, pore boundary markers, and abnormal sector markers on the display interface;

[0031] Abnormal sector markers are bound to sector identifiers and strip identifiers; when an abnormal sector marker is selected, the display interface shows the corresponding abrasive grain exposure characteristics, effective abrasive grain spacing characteristics, and pore connectivity characteristics.

[0032] Furthermore, the three-dimensional measurement equipment for the surface morphology of the grinding wheel generates inspection data files;

[0033] The test data file includes fields for grinding wheel number, sector identifier, strip identifier, focal plane height identifier, focusing height, three-dimensional height field index of grinding wheel surface development, abrasive region number, pore region number, and grinding wheel surface topography evaluation results.

[0034] The test data file is written to the local storage area after each measurement is completed.

[0035] Furthermore, the three-dimensional measurement equipment for the surface morphology of the grinding wheel encapsulates the detection data file into a detection result message;

[0036] The test result message includes a frame header, device number, measurement task number, sector identifier set, strip identifier set, focal plane height identifier set, evaluation result field, and verification field; when the test result message is successfully sent, the three-dimensional measurement equipment for grinding wheel surface morphology records the sending completion status;

[0037] When the detection result message fails to be sent, the three-dimensional measurement equipment for the surface morphology of the grinding wheel records the pending transmission status.

[0038] (III) Beneficial Effects

[0039] This invention provides a three-dimensional measurement method for the surface morphology of a grinding wheel, which has the following advantages:

[0040] By combining a rotatable grinding wheel clamping table, a movable industrial camera, and a circumferential partitioning acquisition method, the circumferential surface of the grinding wheel is unfolded into a measurement object with a unified coordinate relationship, avoiding positional confusion in the circumferential and width directions of multi-view images, thereby providing continuous and traceable three-dimensional height field reconstruction input.

[0041] The pixel focus height is determined by multi-focal image sequences and sharpness evaluation, and the grinding wheel circumferential surface reference subtraction method separates the arc curvature of the grinding wheel itself from the actual surface undulation, so that the exposed height of abrasive grains, pore depth and local terrain volume have a clear height source; the influence of local reflection, shadow and uneven lighting on image grayscale is reduced by a combination of grayscale, illumination equalization, median filtering and background correction, while preserving abrasive grain edges and pore boundaries, so that the three-dimensional height field corresponds to the grayscale texture map space.

[0042] By using three-dimensional height, grayscale texture, height gradient, and topological connectivity to separate the abrasive region, binder region, and pore region, we can reduce the problems of abrasive adhesion, pore misidentification, and binder misidentification caused by relying solely on two-dimensional grayscale thresholds, thereby improving the stability of terrain feature extraction. By performing connectivity analysis, adhesion splitting, and principal axis correction on the effective abrasive region, we can extract abrasive grain size characterization values, abrasive exposed volume, and effective abrasive spacing under different abrasive postures and adjacent abrasive contact conditions, making the obtained parameters more consistent with the actual working surface condition of the grinding wheel.

[0043] By comparing the abrasive grain exposure characteristics, effective abrasive grain spacing characteristics, pore connectivity characteristics, and texture characteristics with benchmark geomorphological data, the surface geomorphological evaluation results of the grinding wheel are output and used to distinguish between abrasive grain dulling, abrasive grain shedding, pore blockage, insufficient dressing, over-dressing, and local abnormal protrusions, providing a quantitative basis for grinding wheel dressing and replacement. Attached Figure Description

[0044] Figure 1 A schematic diagram of the overall process of the three-dimensional measurement method for the surface morphology of a grinding wheel provided in an embodiment of the present invention; (a) is a schematic diagram of the main process of zonal acquisition, multi-focal plane imaging, height reconstruction, region decomposition, feature evaluation and trimming feedback; (b) is a schematic diagram of the image recording and retesting feedback process. Figure 2 This is a schematic diagram of the equipment acquisition scene for the three-dimensional measurement device for grinding wheel surface morphology provided in an embodiment of the present invention; Figure 3 This is a schematic diagram of the partitioned acquisition of the circumferential measurement sector and radial acquisition strip of the grinding wheel provided in an embodiment of the present invention. Figure 4 A schematic diagram of a multi-focal plane imaging scene within a single measurement sector and a single acquisition strip, provided in an embodiment of the present invention; Figure 5 This is a schematic diagram of the coordinate mapping between image recording packaging and grinding wheel surface unfolding provided in an embodiment of the present invention; Figure 6 A schematic diagram of the image processing flow for reconstructing a three-dimensional height field from a multi-focal plane image sequence provided in an embodiment of the present invention; Figure 7 This is a schematic diagram of the grinding wheel circumferential surface reference deduction and adjacent height field splicing provided in an embodiment of the present invention; Among them, (a) is a schematic diagram of grinding wheel circumferential surface benchmark subtraction and terrain height generation, and (b) is a schematic diagram of reliable priority splicing of overlapping areas of adjacent sectors and adjacent strips; Figure 8 A schematic diagram of the region decomposition and abrasive geometry correction of the effective abrasive grain region, binder candidate region, and pore region on the surface of the grinding wheel provided in an embodiment of the present invention; Figure 9 This is a schematic diagram of the evaluation and repair feedback process for the surface topography of a grinding wheel provided in an embodiment of the present invention. Detailed Implementation

[0045] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0046] Please see Figures 1-9 This invention provides a three-dimensional measurement method for the surface morphology of a grinding wheel, comprising:

[0047] A rotating grinding wheel clamping table, industrial camera, lens, and light source form a unified acquisition entry point. Focal point change measurement utilizes finite depth of field and vertical scanning to obtain surface topography information. Since the surface topography of the grinding wheel is related to the distribution of abrasive grain edges, step one first solidifies the position, field of view, and focal plane height.

[0048] Step 1: Establish a unified data entry point for circumferential partitioning acquisition and multi-focal plane imaging calibration of the grinding wheel surface, so that each frame of the grinding wheel surface image has a defined sector position, strip position, focal plane height and pixel equivalent, and provides continuous input for reconstructing the three-dimensional height field in Step 2.

[0049] The surface of the grinding wheel is not a planar sample; abrasive grains, binder, and pores are distributed on a circumferential curved surface. When the industrial camera only captures images within a fixed field of view, the abrasive grains at the image edges may shift due to arc unfolding errors, and the dark areas at the pore boundaries are easily superimposed with lighting shadows. To avoid subsequent algorithms repeatedly identifying the same abrasive grain or incorrectly binding adjacent abrasive grains, the grinding wheel rotation, the industrial camera lateral movement, and the industrial camera height movement are first unified into a labeled acquisition sequence. This processing ensures that each exposure carries a transferable spatial identifier; when reading the multi-focal image sequence in step two, it is possible to directly determine which circumferential measurement sector, which radial acquisition strip, and which focal plane height the image belongs to.

[0050] This step is performed by a three-dimensional measurement device for the surface topography of a grinding wheel. This device includes a rotatable grinding wheel clamping table, an industrial camera, a camera height adjustment mechanism, a camera lateral adjustment mechanism, a lens, a ring light source, and a data acquisition controller. The grinding wheel to be measured is clamped on the coaxial clamping shaft of the rotatable grinding wheel clamping table. The clamping shaft preferably uses a steel mandrel and an end-face clamping component. An elastic washer is placed between the end-face clamping component and the side of the grinding wheel to reduce lateral sway during clamping. The industrial camera moves along the width direction of the grinding wheel via the camera lateral adjustment mechanism and along the optical axis of the lens via the camera height adjustment mechanism.

[0051] A ring light source emits light around the lens, preferably white light. When the grinding wheel surface has strong reflections, a diffuser is placed at the front of the ring light source. The diffuser converts point highlights into area illumination, preserving the transition between light and shadow at the edges of the abrasive grains. This structure forms a three-axis imaging relationship: the turntable provides the circumferential position, the lateral adjustment mechanism provides the strip position, and the height adjustment mechanism provides the focal plane position. After each action, the acquisition controller reads the turntable angle feedback, lateral displacement feedback, and height displacement feedback. Once these three feedbacks enter a hold state, exposure is triggered, ensuring that the mechanical action and image recording are completed within the same acquisition cycle.

[0052] Before starting the data acquisition, the data acquisition controller reads the nominal radius of the grinding wheel to be measured. Grinding wheel width Circumferential width of the field of view of industrial cameras Width in the field of view direction Circumferential width of the field of view The width in the field of view is obtained by multiplying the number of pixel columns by the distortion-corrected pixel equivalent. It is obtained by multiplying the number of pixel rows by the distortion-corrected pixel equivalent. Then, the acquisition controller determines the circumferential overlap rate. Calculate the circumferential step angle And based on the circumferential step angle Obtain the number of circumferential measurement sectors :

[0053]

[0054] Where: arctangent function An operator used to convert the line width of the field of view to a circular angle and round up. Used to ensure that the last segment of the circumference is covered; nominal radius of the grinding wheel. : Represents the outer radius of the grinding wheel being measured, a mechanical measurement value greater than 0, used to convert the camera's planar field of view to the circumferential unfolded length of the grinding wheel; Circumferential width of the field of view. : Represents the actual width covered by a single frame image in the circumferential unfolding direction of the grinding wheel, and the value is greater than The calibration value is used to determine the coverage range between adjacent sectors;

[0055] Circular overlap rate : Indicates the overlap ratio of adjacent circumferential measurement sectors, preferably ranging from 0.10 to 0.30, used to provide a common area for subsequent height field stitching; circumferential step angle : Indicates the angle of rotation of the rotatable grinding wheel clamping table for each rotation, with a value greater than 0 and not greater than 0. The angle value is used to control the circumferential acquisition cycle time; the number of circumferential measurement sectors. : Indicates the number of sectors required to complete one cycle of data acquisition. The value is a positive integer and is used to limit the number of acquisition cycles. The radial acquisition strip is based on the width of the grinding wheel. Width in the field of view direction Generate, set radial overlap rate for adjacent radial acquisition strips The preferred value is 0.15 to 0.40, used to cover the edge area in the width direction of the grinding wheel and retain splicing allowance.

[0056] For example, after the operator loads the grinding wheel into the clamping shaft and presses the end face clamping component, the acquisition controller first stops the rotatable grinding wheel clamping stage at the zero position, and then drives the industrial camera to move to the first strip. After the industrial camera completes the imaging of the strip in the first sector, the stage moves to the next sector, and the sector number, strip number, and angle value are simultaneously written into the image recording. If there are multiple radially acquired strips within the same sector, the acquisition controller first completes the imaging of all strips in that sector before moving to the next sector. Furthermore, the circumferential position is given by mechanical rotation, the strip position is given by lateral movement, and the overlapping area is predefined by parameters, so subsequent stitching does not require re-guessing the image source.

[0057] After each circumferential measurement sector and each radial acquisition strip is determined, the acquisition controller does not immediately move to the next position. Instead, it drives the camera height adjustment mechanism to move along the lens optical axis, acquiring a multi-focal plane image sequence around the theoretical focal point of the grinding wheel surface. The focal plane height is determined by the initial focal plane height. Focal plane step distance and high coverage bandwidth Common limitations:

[0058]

[0059] Where: Initial height of focal plane : Indicates the height of the first layer in a multi-focal plane scan. Its value is determined by the initial focus position of the lens and the theoretical position of the grinding wheel surface, and is used to provide a zero-point height for the multi-focal plane image sequence; Focal plane layer number. : Indicates the current focal plane index, with a value ranging from 0 to 1. An integer used to index different focal plane images within the same field of view; focal plane step size. : Indicates the height interval between adjacent focal planes, preferably ranging from 1 to 20 μm, used to cover variations in abrasive grain exposure and avoid excessive differences between adjacent focal planes; height coverage bandwidth : Indicates the height range covered by multifocal scanning, and the value is a set value greater than 0;

[0060] focal plane height : indicates the first The mechanical height corresponding to the focal plane layer is determined by the focal plane layer number. Incremental, used for inverting the focus height in step two; number of focal plane layers : Indicates the number of focal plane images acquired in the same field of view, and the value is a positive integer, preferably no less than 3 layers.

[0061] For example, after the industrial camera is aligned with the first strip of the first sector, the acquisition controller first lowers the camera height adjustment mechanism to the starting height of the focal plane. Once the altitude feedback stabilizes, trigger the exposure; then press the focal plane step. Move the layers up one by one and expose them one by one until the focal plane layer number is obtained. The corresponding multi-focal plane image sequence. Each frame of the image is accompanied by a focal plane layer number. and focal plane height Therefore, when calculating the sharpness evaluation value in step two, the sharp areas of the image can be directly correlated with the mechanical height. Consequently, the same abrasive tip, binder surface, and pore bottom will appear sharp in different focal planes, providing a true source of height for subsequent three-dimensional height field reconstruction.

[0062] After setting the acquisition path and multi-focal height sequence, the acquisition controller uses calibration components to establish the correspondence between pixel coordinates, grinding wheel surface unfolding coordinates, and focal height. The calibration components are preferably a planar calibration plate with a known grid spacing and a stepped calibration component with a known height difference; the planar calibration plate is used to calculate the pixel equivalent after distortion correction, and the stepped calibration component is used to calibrate the displacement feedback of the camera height adjustment mechanism. To allow direct access in step two, the acquisition controller encapsulates each frame of image into an image record, which at least includes an image matrix and a circumferential measurement sector number. Radial acquisition strip number Focal plane layer number Rotation angle Lateral displacement focal plane height and pixel equivalent , .

[0063] Its expanded coordinate mapping relationship is expressed as:

[0064]

[0065] Where: Circumferential expansion coordinates : Represents the position of a pixel in the unfolding direction of the grinding wheel circumference; its value is determined by the rotation angle and the pixel column position; width unfolding coordinates : Indicates the position of a pixel in the width direction of the grinding wheel; its value is determined by the pixel row position and the lateral strip displacement; focal plane calibration height. : Indicates the focal plane height corresponding to the image record, and the value is the focal plane height. ;

[0066] Rotation angle : indicates the first The worktable angle corresponding to each circumferential measurement sector, with a value ranging from 0 to... Inner angle value; pixel column coordinates and pixel row coordinates : Represents the pixel position within the image matrix, with values ​​being integers within the image size range; Pixel center column coordinates and pixel center row coordinates : Indicates the position of the center pixel of the field of view; the value is determined by the camera resolution.

[0067] Lateral displacement and strip compensation displacement : indicates the first The mechanical position compensation for each radial acquisition strip is determined by feedback from the camera's lateral adjustment mechanism; circumferential pixel equivalent. and width pixel equivalent : Represents the length of a unit pixel after distortion correction, and takes a value greater than 0.

[0068]

[0069]

[0070]

[0071] Wherein, the original pixel coordinates This represents the pixel position directly output by the industrial camera, taken as an integer coordinate within the effective pixel range of the image, and is used as input for distortion correction; the pixel coordinates after distortion correction... This represents the pixel position after lens distortion correction, using real coordinates. Its function is to perform coordinate mapping on the grinding wheel surface; lens principal point coordinates. This represents the projection point of the lens optical axis onto the image plane, taken from the calibrated pixel coordinates, and its function is to define the radial distortion center; the square of the radial distance... This represents the squared distance between the original pixel coordinates and the principal point coordinates of the lens, taking a non-negative value. Its purpose is to calculate the radial distortion correction amount.

[0072] Radial distortion coefficient and This represents the radial distortion parameter of the lens determined by the calibration plate. It is a real value and its function is to correct the coordinate offset of the lens edge region.

[0073] This mapping operator is directly calculated from the aforementioned three-row coordinate relationship, thus enabling those skilled in the art to reproduce the unfolded position of each pixel according to the fields in the image record. The image record is based on the circumferential measurement sector number. Radial acquisition strip number and focal plane layer number The data is written to the input buffer in increments; when the acquisition controller detects the focal plane layer number... If missing, re-trigger the focal plane height. Appropriate exposure is used to avoid tomography in multi-focal image sequences.

[0074] Furthermore, the image matrix, mechanical position, and focal plane height are written into the same image record, so that the same coordinate names and the same data objects can be used in step two when performing sharpness evaluation, surface reference subtraction, and local height field stitching.

[0075] Without changing the working principle of step one, the rotatable grinding wheel clamping worktable is realized by a servo rotary table, a stepper rotary table, or an indexing mechanism with an angle encoder; when the industrial camera is an area array color camera, a two-dimensional field of view is obtained in one step; when the industrial camera is an area array monochrome camera, color channel conversion is omitted but the multi-focal image sequence is retained; when the ring light source is replaced with a coaxial light source or a low-angle ring light source, the standard is still that the edge of the abrasive grain can form a stable transition between light and dark; when the camera height adjustment mechanism adopts a ball screw lifting mechanism, a linear motor lifting mechanism, or a piezoelectric lifting mechanism, it is all based on the ability to output the focal plane height. These are boundary conditions.

[0076] The preferred embodiment is as follows: a steel mandrel holds the grinding wheel, a white ring light source is used for illumination with a diffuser, an area array industrial camera covers the width direction of the grinding wheel through a horizontal adjustment mechanism, and a multi-focal image sequence is acquired through a height adjustment mechanism. The acquisition controller operates in the following sequence: zero-position confirmation, strip positioning, sector rotation, layer-by-layer focusing, and image recording and packaging.

[0077] The operator can see the workbench rotate sequentially, the industrial camera move strip by strip, and the camera height adjustment mechanism rise and fall layer by layer, ultimately obtaining a set of image records with sector number, strip number, and focal plane layer number; the corresponding technical effect is that the acquisition position is not confused, the focal plane height is not lost, and subsequent steps can be continuously called.

[0078] Step 2: Convert the output multi-focal plane image sequence with sector, strip, and focal plane height markers into a three-dimensional height field, a three-dimensional topographic map unfolded from the grinding wheel surface, a grayscale texture map, and a sharpness confidence map, so that Step 3 can directly call the topographic data with height and coordinate sources.

[0079] The surface of a grinding wheel is composed of abrasive protrusions, bond plateaus, and pore depressions. Within the same field of view, there are both bright abrasive cutting edges and dark pore shadows. If only a single frame of a two-dimensional image is segmented by grayscale thresholding, bright reflections will be mistaken for abrasive tips, and dark area edges will be mistaken for pore outlines. This step utilizes a multi-focal image sequence to create a sharpness variation curve for the same pixel location at different focal plane heights; then, by combining the sharpness peak, the grinding wheel circumferential surface reference, and the circumferential unfolded coordinates, the terrain height of each pixel relative to the local reference plane is obtained.

[0080] This step is executed collaboratively by the image processing unit and the acquisition controller on the industrial computer. The acquisition controller transmits image records, which include an image matrix and circumferential measurement sector numbers. Radial acquisition strip number Focal plane layer number focal plane height Circular expansion coordinates Width expansion coordinates Circular pixel equivalent and width pixel equivalent The image processing unit first generates a grayscale preprocessed image. Then, the focal plane response value is formed. The initial height is determined by the peak value. Then subtract the grinding wheel circumferential surface reference. Obtaining terrain height .

[0081] Therefore, the mechanical position markers and image sharpness together determine the height field, and subsequent stitching also unfolds the coordinates in the same circumferential direction. Width expansion coordinates As the standard.

[0082] Image computing unit reads the first The first circumferential measurement sector, the first The radial acquisition strip, the first The image matrix corresponds to each focal plane layer number. Since the ring light source will create local highlights on the edge of the abrasive grain and the reflection at the pores of the grinding wheel is weak, the image calculation unit first performs channel-weighted grayscale conversion on the color image, then uses the flat field image to eliminate the illumination distribution, and finally performs median filtering. The median filtering window is preferably an odd-length window, and the window side length does not exceed one-third of the projection width of the smallest abrasive grain to be identified, so that noise points are removed and the abrasive grain edges are not smoothed across.

[0083] For example, in the original field image, there are bright white spots around the abrasive grain tips and dark patches at the bottom of the pores; after this processing method, the bright spots are confined within the abrasive grain boundaries, the outline of the dark pore area is preserved, and the image recording still retains the circumferential measurement sector number. Radial acquisition strip number and focal plane layer number .

[0084]

[0085] Where: grayscale preprocessed image Indicates the first The focal plane layer number is located at the pixel position. The single-channel image at the location is used to extract non-negative grayscale values, which are then used as input for sharpness evaluation; pixel position Representing the column and row positions in the image matrix, using integer coordinates within the effective pixel range to lock the spatial location; channel weights Channel weight and channel weight These represent the weights of the red, green, and blue channels, with values ​​ranging from 0 to 1, and the sum of the three values ​​is 1. They are used to synthesize grayscale.

[0086] Red channel image Green channel images and blue channel image Represents the pixel intensity of the three channels, taking the non-negative value within the camera's output grayscale range; dark field image. This represents the sensor noise floor obtained by shading the image; a non-negative grayscale value is used to subtract fixed noise. (Plain field image) This represents the illumination distribution obtained from a uniform reflective surface, using positive grayscale values ​​to correct for uneven illumination; stability constant. To take a positive value, used to prevent the denominator from approaching zero; the median operator. Indicates radius as Neighborhood median filtering, radius A positive integer value is used to remove isolated noise and preserve abrasive grain boundaries. This, in turn, optimizes the focal plane response value. The degree of edge sharpness is determined, and the required spatial markings are preserved.

[0087] After grayscale preprocessing is completed, the image calculation unit is at the same pixel position. Along the focal plane layer number Read grayscale preprocessed image And calculate the focal plane response value in the local neighborhood of that pixel. Focal surface response value Ideally, it should be composed of both Laplacian edge response and gradient direction consistency; the former reflects the clarity of the abrasive cutting edge, while the latter suppresses random noise spurious peaks.

[0088] For regions like the bottom of pores, which are darker in gray but have clear boundaries, the uniformity of gradient direction creates identifiable peaks. Image computation units are based on focal plane height. Arrangement focal plane response values Select the focal plane layer number where the peak is located. The initial height was obtained by parabolic interpolation using the adjacent focal plane layer numbers before and after the peak. When the peak is located at the end of the sequence, the image calculation unit marks the pixel as a low-confidence boundary point and calls the interpolation results of adjacent valid pixels. Wherein:

[0089]

[0090]

[0091] Where: peak focal plane layer number Indicates pixel position The focal plane layer number with the largest focal plane response value is selected as an integer to determine the peak layer; focal plane response value Indicates the first The focal plane layer number is located at the pixel position. The sharpness of the image, taken as a non-negative value, is used to evaluate focus; initial height. This indicates the pixel height before deducting the grinding wheel circumferential surface reference. It takes the value within the height range covered by the multi-focal surface scan and uses it as the surface reference deduction input.

[0092] focal plane height Indicates the peak focal plane layer number The corresponding mechanical height, taken from the calibration value in step one, is used to provide a height scale; when the peak focal plane layer number... When located on the first or last layer, three-point parabolic interpolation is not performed; instead, the initial height is set. Set to the corresponding focal plane height and the sharpness confidence of that pixel. Lower the sign; when the denominator When it is zero or close to zero, it is also used. .

[0093] Focal plane step This represents the interval between adjacent focal plane heights, taken as a positive number, used to convert the focal plane layer number difference into an actual height difference; focal plane response value. Focal surface response value and focal plane response value This indicates the sharpness of the peak layer and its adjacent layers. It is a non-negative value and is used to fit the sharpness curve near the peak.

[0094] At pixel position local neighborhood Internal calculation of focal plane response value The local neighborhood For Centered on, with radius A set of square or circular pixels. The lateral gradient is calculated using center difference. Longitudinal gradient and Laplace's response The focal plane response value is then calculated using the following formula:

[0095]

[0096] Among them, focal plane response value Indicates the first At pixel position in a focal plane image The sharpness of a point, taking a non-negative value, is used to determine the focus height; local neighborhood. Represented by pixel position The neighborhood of the center is the set of effective pixels in the image, which is used to avoid single-pixel noise affecting the sharpness evaluation; neighborhood weight. Represents pixels within the neighborhood The weighting coefficients are non-negative and the sum of the weights in the neighborhood is 1. Its function is to emphasize the central pixel while preserving local edges; horizontal gradient and longitudinal gradient These represent grayscale preprocessed images. The horizontal and vertical grayscale variations, expressed in real values, represent the sharpness of edges; Laplace response. Indicates grayscale preprocessed image The second-order transformation, taking real values, enhances subtle edges at the focal point; Laplace weights and gradient weights Both are set to 0 to 1, and their sum is 1. Their function is to adjust the contribution ratio between the second-order edge and the first-order gradient.

[0097] Furthermore, the height of the abrasive tip, binder surface, and pore bottom is obtained according to the sharpness peak, providing continuous input for the calculation of abrasive exposed volume.

[0098] Because the measured area of ​​the grinding wheel is located on its circumferential surface, pixels at the same edge of the field of view, even if belonging to an unworn reference surface, will have a height difference relative to the center of the field of view due to the curvature of the arc. The image calculation unit is based on the nominal radius of the grinding wheel. Circular expansion coordinates Expand coordinates with sector center Calculation of grinding wheel circumferential surface reference Then from the initial height Subtracting the grinding wheel circumferential surface reference The resulting landform height only represents localized abrasive grain exposure, binder undulations, and pore depressions. Subsequently, the image processing unit utilizes the overlapping areas of adjacent circumferential measurement sectors and adjacent radial acquisition strips, according to terrain height... Registration is performed in the order of continuity, consistent grayscale texture map edges, and priority given to reliable regions in the sharpness confidence map; when two height estimates of the same abrasive grain appear in the overlapping area, the pixel with the more separated peak in the sharpness confidence map is given priority.

[0099]

[0100] Where: the circumferential curved surface of the grinding wheel is the reference. Indicates the first The reference height within each circumferential measurement sector, determined by the arc of the grinding wheel, is taken as a non-negative value to differentiate the overall curvature of the grinding wheel; the nominal radius of the grinding wheel... Represents the outer radius of the grinding wheel to be measured, taken as a positive number, used to determine the circumferential surface datum; circumferential unfolded coordinates This indicates the pixel's position within the circumferential unfolding direction of the grinding wheel, using coordinate mapping values ​​for locating the arc offset; sector center unfolding coordinates. Indicates the first The position of the center of each circumferential measurement sector in the direction of the grinding wheel's circumference is used to obtain the corresponding value of the table rotation angle, which is used to provide a central reference; terrain height. Indicates pixel position The actual terrain height relative to the local surface reference is taken as a positive, zero, or negative value for region separation.

[0101] initial height This represents the obtained untrimmed curvature height, taken from the height range covered by the multifocal scan, used to provide the raw height estimate. The image processing unit also generates a sharpness confidence score. The confidence level of this sharpness From peak focal plane response value With the response value of the second peak focal plane The separation degree is obtained by normalization, with values ​​ranging from 0 to 1, and is used to indicate the reliability of the height estimation; the secondary peak focal plane response value Indicates the focal plane layer number excluding the peak focal plane. The second largest focal plane response value, taken as a non-negative value, is used to identify unstable heights caused by multi-peaks or reflections.

[0102] Furthermore, this increases the elevation of the landform. Decoupled from the circumferential curvature of the grinding wheel, overlapping areas are spliced ​​according to the same height reference; clarity and confidence. This is used in step three to distinguish between reflection, aperture depth, and boundary occlusion locations.

[0103]

[0104] Among them, the confidence level of sharpness Indicates pixel position The credibility of the high estimate is taken to Its function is to suppress reflections, occlusions, and multi-peak focal plane pixels in step three; peak focal plane response value The maximum focal plane response value is a non-negative value, serving as primary focus evidence for the current pixel; the secondary focal plane response value... Indicates the focal plane layer number excluding the peak focal plane. The second largest focal plane response value outside the main body, taken as a non-negative value, is used to identify whether the focal plane response has multiple peaks; confidence stability constant. Taking a positive number is intended to prevent the denominator from becoming invalid when the peak focal plane response value is small; maximum function and minimum function They are used to determine the confidence level of sharpness, respectively. Limited to the range of 0 to 1;

[0105] Focal surface response value Not limited to the combination of Laplacian edge response and gradient direction consistency, it also employs Tenengrad gradient, local variance gradient, or wavelet high-frequency energy generation, as long as the input is still a grayscale preprocessed image. The output is still pixel position. Non-negative clarity. Surface reference subtraction is not limited to the nominal radius of the grinding wheel. The corresponding analytical arc is also fitted using a reference plane within the overlapping region; during fitting, a sharpness confidence level is selected. The high and gently varying bond area is used as the fitting point, and the output is still named the grinding wheel circumferential surface reference. .

[0106] In a preferred embodiment, an industrial computer reads the image recordings and sequentially generates grayscale preprocessed images. Focal surface response value Initial height Landform height and clarity confidence The visible results on-site include a fully clear grayscale texture map, a three-dimensional topographic map of the grinding wheel surface, and a low-confidence area marker map. Verification indicators include the height residual of the step calibration component, the height closure difference of the overlapping area, and the peak separation degree of the same pore boundary at different focal plane layers.

[0107] Step 3: Based on the terrain elevation obtained in Step 2 Grayscale texture image, sharpness confidence level Circular pixel equivalent and width pixel equivalent The three-dimensional topographic map of the grinding wheel surface is decomposed into effective abrasive grain region, binder candidate region and pore region, and the set of topographic features required in step four is output.

[0108] The surface of the grinding wheel is not formed by a uniformly laid-out pattern of particles; the abrasive grains are held in place by a binder, and pores are trapped between the abrasive grains and the binder. A single bright grayscale value can originate from either the tip of the abrasive grain or from reflections on its side; similarly, a single dark grayscale value can originate from the bottom of the pores or from shadows cast by the abrasive grains. Therefore, step three does not use a single grayscale threshold for classification, but instead considers the terrain height... As the primary criterion, the grayscale texture image is used as the boundary criterion, and the sharpness confidence level is used as the secondary criterion. As a reliable criterion, topological connectivity is used as a structural criterion. In this way, the height source provided in step two is preserved, and the abrasive grain exposure, effective abrasive grain spacing, and pore connectivity required in step four are continuously generated from the same topographic data.

[0109] This step is performed by the image processing unit. The image processing unit first reads the three-dimensional topographic map of the grinding wheel surface and then unfolds it in circumferential coordinates. Width expansion coordinates As a global planar location, based on terrain height This represents local undulations. Subsequently, the image computation unit at each pixel location... A local neighborhood is formed around the abrasive grains, and the height response, boundary response, texture response, and sharpness confidence response are calculated to obtain abrasive grain evidence maps, binder evidence maps, and porosity evidence maps. After completing the three types of evidence maps, the image processing unit first extracts candidate abrasive grain regions, then uses height valley lines and pore boundaries to separate bonded abrasive grains, and finally performs PCA principal axis correction to obtain grain size, exposure, and spacing parameters. The entire processing sequence remains evidence generation, region separation, geometric correction, and parameter encapsulation, ensuring that the output corresponds to the evaluation fields in step four.

[0110] Image computing units based on terrain height The system identifies abrasive grains by locating raised areas, pores by identifying low-lying, dark textured areas, and candidate binder regions by locating areas with gentle heights between abrasive grains and pores. To avoid misjudgments due to high grayscale caused by reflections, abrasive grain determination does not directly use the maximum grayscale value, but instead requires the simultaneous participation of height response, boundary response, and sharpness confidence response. To prevent pore shadows from being absorbed by the abrasive grain sidewalls, pore determination simultaneously reads the low height response, dark texture response, and connected closed boundaries.

[0111] For example, the operator sees on the display screen that the previously glued bright area is broken down into several raised marks, the dark recessed area is retained as pore marks, and the light-colored platform between them is retained as a candidate area for the binder.

[0112]

[0113] Where: Abrasive evidence value Indicates pixel position Comprehensive evidence belonging to the candidate abrasive grain region, with a value ranging from 0 to 1, is used to generate candidate abrasive grain regions; highly responsive. Indicates the elevation of the landform The normalized protrusion degree relative to the local binder reference height, with a value of 0 to 1, is used to highlight the abrasive grain exposure;

[0114] Boundary response Indicates the elevation of the landform The height gradient normalization value within the local neighborhood, ranging from 0 to 1, is used to preserve the edges of the abrasive grains; texture response. This represents the normalized texture intensity corresponding to the rough reflection of the abrasive surface in the grayscale texture image, with a value ranging from 0 to 1, used to help distinguish between abrasive grains and the binder; Sharpness confidence level. This indicates the high degree of confidence of the output from step two, with a value ranging from 0 to 1, used to suppress low-confidence pixels caused by reflections and occlusions;

[0115] High weight Boundary weights Texture weights and confidence weight Each value is between 0 and 1, and the sum of the four values ​​is 1, used to limit the value of each piece of evidence in the abrasive evidence category. The proportion of contribution.

[0116] The image computing unit uses abrasive grain evidence values The high-value connected regions form candidate abrasive grain regions, the low-height and dark-textured connected regions form pore regions, and the remaining flat regions form binder candidate regions.

[0117]

[0118] Among them, high response Indicates pixel position The degree of protrusion relative to the local binder reference is taken as follows: to Its function is to highlight the abrasive grains; localized binder reference height. Indicates pixel position The height benchmark of the candidate region for the smooth binder within the neighborhood is taken as a real number height value, which serves as the zero point for the convex height; local peak height. Indicates pixel position The representative height of the reliable bulge within the neighborhood, taken as a real number, serves to normalize the height response; height stability constant. Taking a positive number is intended to prevent the denominator from becoming invalid when the local peak height is close to the local binder reference height.

[0119] The boundary response is:

[0120]

[0121] Among them, boundary response Indicates the elevation of the landform At pixel position Edge strength, take to Its function is to identify abrasive grain edges; upper limit of local height gradient. The upper bound of the elevation gradient within the local neighborhood is represented by a positive number, and its function is to normalize the boundary response; the gradient stability constant. Taking a positive number serves to maintain calculation stability;

[0122] Texture response:

[0123]

[0124] Among them, texture response Represents the grayscale texture image at the pixel position The local texture intensity, take to Its function is to distinguish between the rough reflection of abrasive particles and the smooth areas of the binder; local texture energy and local texture contrast Obtained from the gray-level co-occurrence matrix, taking non-negative values, its function is to describe the local texture state; the upper limit of local texture energy. and local texture contrast limit Represents the texture normalization reference within the neighborhood, taking a positive value; its function is to normalize the texture response; texture stability constant. Taking a positive number is to prevent the denominator from becoming invalid.

[0125] Furthermore, the abrasive region is defined by both height and sharpness, the pore region by both low height and dark texture, and the binder candidate region by spatial residual relationship. The three types of regions have common coordinates and boundary traceability.

[0126] After candidate abrasive grain regions are formed, the image processing unit performs connected component labeling and checks whether there are more than two height peaks within the connected component of each candidate abrasive grain. If adjacent abrasive grains are in contact on the two-dimensional projection, a simple connected component will treat multiple abrasive grains as a single object; this processing method finds the boundary along the height valley line, which is formed by the terrain height between adjacent height peaks. Decreased and sharpness confidence The effective positional composition is still maintained. The image computing unit performs marker watershed splitting using height peaks as seeds and height valleys as segmentation boundaries to obtain non-overlapping effective abrasive grain regions. .

[0127] Subsequently, for each effective abrasive region Extracting the circumferential unfolding coordinates of pixels Width expansion coordinates Constructing a confidence level with clarity Shape matrix of weights And take the direction with the largest feature as the direction of the abrasive grain spindle.

[0128]

[0129] Where: morphology matrix Indicates the first Effective abrasive zone The weighted planar distribution matrix, taken as a second-order square matrix, is used to calculate the principal axis direction of the abrasive particles; the effective abrasive particle region. Indicates the first Each split abrasive grain region is used to define a pixel set to limit the geometric measurement range;

[0130] Confidence of sharpness Indicates pixel position The high confidence level, ranging from 0 to 1, is used to reduce the influence of low confidence boundary points on the principal axis direction; circumferential unfolded coordinates. Width expansion coordinates Indicates pixel position The actual coordinates in the plane of the grinding wheel are taken from the mapping value in step one, which is used to convert the pixel size into the actual geometric size.

[0131] Circumferential coordinates of abrasive grain center and abrasive grain center width coordinates Indicates the effective abrasive grain area Internal Confidence Based on Sharpness The weighted center coordinates are used to determine the actual length value, which is then used to eliminate the influence of regional position offset on the principal axis direction. The image calculation unit calculates the maximum length along the abrasive grain principal axis direction and the minimum width along the direction perpendicular to the principal axis direction, and saves the rotated boundary projection as the abrasive grain size characterization value.

[0132] Define the effective abrasive zone The outer ring zone is :

[0133]

[0134] Among them, the outer ring zone Indicates the effective abrasive grain area The surrounding pixel set used to estimate the binder height; this pixel set provides a local reference height; distance function. Indicates pixel position To the effective abrasive zone The nearest boundary distance, taken as the non-negative pixel distance, is used to define the range of the ring band; the inner ring radius... and outer ring radius All are positive numbers and Its function is to exclude abrasive grain edges and limit the binder sampling range; binder candidate region This represents the set of binder pixels obtained from the initial separation in step three. The purpose of taking this set of pixels is to prevent pores or abrasive particles from entering the reference calculation.

[0135] The reference height for local bonding agent is determined by the following formula:

[0136]

[0137] Among them, the median operator Indicates taking the outer expansion ring. The median value of the ground surface height serves to reduce the impact of individual abrasive residues or pore points on the reference height.

[0138] Furthermore, the adhered abrasive grains are first separated by the height valley line, and then the measurement direction is corrected by the PCA spindle. The abrasive grain posture will not change the grain size reading method, and the center point of the subsequent effective abrasive grain spacing also has a unified geometric source.

[0139] Image computing unit in each effective abrasive region The binder candidate region is read from the outer edge, and a local binder reference height is formed by pixels with a gently sloping height within the ring. Subsequently, the image processing unit calculates the effective abrasive grain area. The inner height is higher than the local binder reference height. The surface accumulation of the portion is used to obtain the exposed volume of the abrasive particles. Simultaneously read the effective abrasive grain area Maximum terrain elevation and local binder reference elevation The difference is used to obtain the height of the exposed abrasive grains.

[0140] For porous regions, the image processing unit traces connected paths along pore boundaries, recording the number of pores, pore area ratio, pore depth, and Euler number. For the overall terrain, the image processing unit reads texture contrast, texture energy, texture homogeneity, texture correlation, and texture entropy from the grayscale texture image and encapsulates them with geometric features. Among these:

[0141]

[0142] Where: the volume of abrasive grains exposed Indicates the first Effective abrasive zone Higher than the local binder reference height The volume, taking a non-negative value, is used to characterize the amount of protrusion actually involved in grinding by the abrasive grains; terrain height. Indicates pixel position The three-dimensional terrain height, taken as positive, zero, or negative, is used to provide the volume integral height; local binder reference height. Indicates the first Effective abrasive zone The reference height of the surrounding binder candidate area is taken as the actual height value and used as the zero point for exposure calculation;

[0143] Circular pixel equivalent and width pixel equivalent These represent the actual lengths corresponding to a unit pixel in the circumferential and width directions, respectively. They are positive numbers and are used to convert pixel accumulation into actual area.

[0144]

[0145] Where: effective abrasive grain spacing Indicates the first The shortest distance from each effective abrasive grain center to its nearest neighbor is taken as a non-negative value to characterize the density of abrasive grain distribution; the circumferential coordinates of the abrasive grain centers are also considered. and abrasive grain center width coordinates Indicates the first Effective abrasive zone The center position is determined by the actual length value; the circumferential coordinates of the abrasive grain center are... and abrasive grain center width coordinates Indicates the first Effective abrasive zone The center position is taken as the actual length value; the first... Valid abrasive grain serial number and the Valid abrasive grain serial number Both are positive integers and are different, used to distinguish different effective abrasive grain regions;

[0146] Furthermore, the exposed volume of abrasive particles The effective abrasive grain spacing is determined by both height and pixel equivalent. The porosity parameter is determined by the center of the unfolded coordinates, the pore size parameter is determined by the connected path, the texture parameter is determined by the grayscale texture map, and all output fields can be traced back to the input data in step two.

[0147] In an alternative implementation, the candidate abrasive region is not based on abrasive evidence values. Determined individually, and with a high degree of response. and clarity confidence The intersection is determined; the splitting of adherent abrasive grains is not limited to marked watersheds, but also uses high valley line cutting or minimal curvature line cutting, and the output is still the effective abrasive grain region. Principal axis correction is not limited to the topography matrix. The characteristic direction is also adopted using the direction of the minimum bounding rectangle, and the output is still the direction of the abrasive spindle.

[0148] In a preferred embodiment, after the operator completes step two, the image processing unit first displays three types of region maps: the abrasive region appears as raised blocks, the pore region as recessed blocks, and the binder candidate region is located between the two. Then, it displays the abrasive grain splitting line and the PCA principal axis. Finally, it generates a graph containing the effective number of abrasive grains, abrasive surface density, abrasive grain size characterization value, abrasive grain exposure height, and abrasive grain exposure volume. Effective abrasive grain spacing This is a set of geomorphic features including pore quantity, pore area ratio, pore depth, pore connectivity, Euler number, and texture characteristics. The verification method is as follows: Typical abrasive grains and pores are manually selected on a 3D geomorphic map unfolded on the grinding wheel surface. The system output boundaries are checked to see if they conform to the contours of high protrusions and depressions, and whether the same abrasive grain in the overlapping area generates consistent abrasive grain center coordinates.

[0149] Step 4: Transform the set of grinding wheel surface geomorphic features obtained in Step 3 into grinding wheel surface geomorphic evaluation results, geomorphic anomaly types, geomorphic change trends, and repair suggestions, so that the measurement results can be transformed from geometric descriptions into actionable grinding wheel maintenance judgments.

[0150] Wheel dressing depends not only on the highest individual abrasive grain, nor solely on the visibility of pores. A reduction in the effective number of abrasive grains leads to sparse cutting edges, a decrease in abrasive grain exposure height results in duller grinding, reduced pore connectivity causes chip space to be occupied by abrasive chips, and a decrease in texture energy reflects a smoother surface edge. If these parameters are observed separately, the operator may easily only see localized bumps or dark spots, neglecting the combined changes in abrasive grains, bond material, and pores within the same circumferential measurement sector.

[0151] The various features in step three are reorganized according to the working mechanism of the grinding wheel to form a continuous chain of benchmark comparison, state judgment, and dressing feedback, so that the output of the three-dimensional measurement method of grinding wheel surface morphology can directly serve the selection, re-inspection and dressing of grinding wheels.

[0152] This step is performed by the geomorphological evaluation calculation unit. The geomorphological evaluation calculation unit reads the geomorphological feature set output from step three, and simultaneously reads one or more of the following: baseline geomorphological data, historical geomorphological data, and post-dressing geomorphological data. The baseline geomorphological data corresponds to the geomorphological features of a new grinding wheel or a properly dressed grinding wheel; the historical geomorphological data corresponds to the geomorphological features of the same grinding wheel during previous testing; and the post-dressing geomorphological data corresponds to the geomorphological features re-measured after this dressing process. The geomorphological evaluation calculation unit first aligns each feature according to the circumferential measurement sector number and the radial acquisition strip number, then calculates the feature deviation, subsequently generating sub-values ​​for abrasive grain exposure, effective abrasive grain spacing, pore connectivity, and texture state, and finally outputs the grinding wheel surface geomorphological evaluation result. The display interface receives the grinding wheel surface geomorphological evaluation result and marks abnormal sectors on the grinding wheel unfolding diagram; the dressing equipment operation interface receives dressing suggestions and displays one of the following: increase dressing depth, adjust dressing feed, shorten dressing interval, partial re-inspection, or replace the grinding wheel.

[0153] The geomorphological evaluation calculation unit does not directly average the entire surface of the grinding wheel. Instead, it uses the circumferential measurement sector number and radial acquisition strip number from step one to place the current geomorphological features and the baseline geomorphological features into the same unfolded coordinate unit. This approach preserves the spatial position of the grinding wheel's width edge, circumferential joints, and local trimming marks.

[0154] For example, after the operator completes a measurement on-site, the display interface shows a circumferential unfolded diagram of the grinding wheel. The exposed volume of abrasive grains, effective abrasive grain spacing, and pore area ratio within the same sector are recorded in the same row, with the corresponding record of the reference grinding wheel displayed beside it. The operator can see which section of the current grinding wheel has a significant depression or blockage. In terms of recording format, each unfolded coordinate unit saves the current geomorphic features, reference geomorphic features, source time stamp, and acquisition sector mark. When a unit lacks reference geomorphic features, the geomorphic evaluation calculation unit calls the reference geomorphic features of adjacent units in the same strip to fill the gap in one go, and prompts the operator to re-examine the location with a light-colored border on the display interface. This recording method ensures that data gaps do not interrupt the evaluation process, while retaining a visible re-examination entry point on-site. The feature deviation is formed by the following formula:

[0155]

[0156] Where: characteristic deviation : indicates the first The degree of deviation of each geomorphic feature from the baseline geomorphic feature is represented by a real value, used to unify feature variations across different dimensions; the current geomorphic feature... : indicates the output of step three. Each feature is represented by its actual numerical value, which is used to provide the current evaluation input; benchmark geomorphological features. : Represents the geomorphic features in baseline geomorphic data, historical geomorphic data, or modified geomorphic data that are consistent with the current geomorphic features. Corresponding features, using the same unit of measurement, are used to provide a comparison reference; feature number Takes positive integers for indexing abrasive grain exposure, spacing, porosity, and texture fields; absolute value operator. : Indicates the baseline geomorphic features Take a non-negative amplitude value to unify the normalized denominator under the same positive and negative signs;

[0157] Stability constant Take a positive number, used for the baseline geomorphic features It maintains computational stability when approaching zero. Furthermore, spatial alignment avoids mixing edge anomalies into the global mean, reducing feature deviation. This allows geomorphic features from different sources to be evaluated on a unified scale, and historical and modified data can be directly spliced ​​with current measurement data.

[0158] After obtaining the feature deviation, the geomorphological evaluation calculation unit does not directly give a normal or abnormal conclusion, but first generates four evaluation sub-values. The abrasive grain exposure evaluation sub-value is formed by the abrasive grain exposure height, abrasive grain exposure volume, and effective abrasive grain number; the effective abrasive grain spacing evaluation sub-value is formed by the spacing between adjacent effective abrasive grains and abrasive grain surface density; the pore connectivity evaluation sub-value is formed by the pore area ratio, pore depth, pore connectivity, and Euler number; and the texture state evaluation sub-value is formed by the texture contrast, texture energy, texture homogeneity, texture correlation, and texture entropy. Taking the abrasive grain exposure evaluation sub-value as an example:

[0159]

[0160] Abrasive grain exposure evaluation sub-value : Indicates the exposure state of abrasive grains participating in grinding, taking a non-negative value, used to judge abrasive grain dulling, insufficient bond coating, and inadequate dressing; average abrasive grain exposure height. : Represents the average effective abrasive grain exposure height within the current measurement area, taking a non-negative height value, used to characterize the degree of abrasive grain protrusion;

[0161] Benchmark Excavation Height : Represents the reference value for abrasive grain outcrop height in the corresponding measurement area of ​​the benchmark geomorphological data, taking a positive height value to provide a height reference; average abrasive grain outcrop volume : Represents the average volume of exposed abrasive grains within the current measurement area, using a non-negative volume value, used to characterize the effective cutting protrusion amount; reference exposed volume. : Represents the exposed volume reference value for the corresponding measurement area in the benchmark geomorphological data, taking a positive volume value, used to provide a volume reference; effective abrasive grain count : Indicates the number of effective abrasive grain areas within the current measurement region, taken as a non-negative integer, used to characterize the number of cutting edges.

[0162] Reference effective number of abrasive grains : Represents the effective number of abrasive grains in the corresponding measurement area of ​​the benchmark topographic data, taken as a positive integer, used to provide a quantitative reference; height weight Volume weight and quantity weight All to And the three and are , used to define the contribution ratio of height, volume and quantity.

[0163] Stability constant Take a positive number, used for the reference exposure height. , benchmark exposed volume Or the reference effective number of abrasive grains The calculation remains stable when the abrasive grains are close to zero. In this processing method, the evaluation sub-value is evaluated when the abrasive grains are exposed. When the effective abrasive grain spacing evaluation sub-value decreases and the abrasive grain surface density decreases, the geomorphological evaluation calculation unit outputs abrasive grain detachment; when the abrasive grain exposure evaluation sub-value... When the texture state evaluation sub-value decreases, the output indicates abrasive dulling or binder coating; when the pore connectivity evaluation sub-value decreases and the pore area ratio decreases, the output indicates pore blockage; when the highest local exposure height is consistently higher than the adjacent area and the sharpness confidence is valid, the output indicates local anomaly protrusion. Furthermore, each anomaly type is supported by multiple fields, and abrasive particles, pores, and textures will not be misled by a single bright spot or dark spot, ensuring that the state determination corresponds to the physical characteristics in step three.

[0164] After obtaining four evaluation sub-values, the geomorphological evaluation calculation unit generates a comprehensive evaluation value and outputs it along with the anomaly type. The comprehensive evaluation value expresses the overall usability of the current geomorphological conditions of the grinding wheel, while the anomaly type indicates the direction of trimming. For example, when the display interface marks a circumferential measurement sector as clogged on the grinding wheel unfolded diagram, the operator sees text prompts on the trimming equipment operation interface suggesting shortening the trimming interval or increasing the trimming intensity. After trimming, the system re-executes steps one through three, and the new geomorphological feature set re-enters step four, displaying it alongside the geomorphological data before trimming. The area where the anomaly marker disappears is recorded as the effective trimming area. The comprehensive evaluation value is formed by the following formula:

[0165]

[0166] Where: Comprehensive evaluation value : Represents the overall evaluation value of the surface topography of the grinding wheel, taking a non-negative value, used to output normal, re-inspection, or repair suggestions; Exposure weight Spacing weight Pore ​​weight and texture weight All values ​​are between 0 and 1, and the sum of the four values ​​is 1, used to limit the contribution ratio of the four evaluation sub-values; abrasive grain exposure evaluation sub-value : Represents the abrasive grain exposure state, taking a non-negative value, used to characterize the degree of abrasive grain protrusion and cutting participation; effective abrasive grain spacing evaluation sub-value : Represents the distribution of abrasive grains, takes a non-negative value, and is used to characterize the density of the cutting edge;

[0167] Pore ​​connectivity evaluation sub-values : Represents the pore volume and connectivity, taking a non-negative value, used to characterize the risk of clogging;

[0168]

[0169] Among them, the pore connectivity evaluation sub-value This indicates the pore volume and connectivity, and takes a non-negative value. Its function is to determine pore blockage.

[0170] Current pore area ratio This represents the ratio of the area of ​​the pore region to the total area measured in the current measurement region, taken as... to Its function is to characterize the area of ​​the chip space; the reference pore area ratio Represents the porosity in the baseline geomorphological data, taking... to And it is not zero; its function is to provide an area reference; the current pore connectivity path length. This represents the total length or main connected path length of the pores within the current measurement area. It takes a non-negative length value and is used to characterize the degree of pore connectivity.

[0171] Reference pore connectivity path length This represents the length of the porous connectivity path in the baseline geomorphological data, taking a positive length value; its function is to provide a connectivity reference; the current Euler number. and the benchmark Euler number All values ​​are integers, reflecting changes in pore topology; area weight. Connectivity weights and topological weights All values ​​are between 0 and 1, and the sum of the three is 1. This serves to limit the contribution ratio of pore area, pore connectivity, and topological change.

[0172] Texture state evaluation subvalue : Represents the texture state of the terrain, takes a non-negative value, and is used to characterize the changes in the edges and surfaces of abrasive particles;

[0173]

[0174] Among them, texture state evaluation sub-value This indicates the state of texture change on the grinding wheel surface, taking a non-negative value. Its purpose is to help determine whether the abrasive grains are dull, the binder coating is excessive, or the dressing process is inadequate; current texture contrast. Current texture energy Current texture homogeneity and current texture entropy The gray-level co-occurrence matrices from step three are used to characterize changes in terrain texture; the baseline texture contrast... Baseline texture energy Homogeneity of baseline texture and baseline texture entropy These are derived from baseline topographic data and serve to provide texture references; texture contrast weights. Texture energy weight Texture homogeneity weight and texture entropy weight Each value is between 0 and 1, and the sum of the four values ​​is 1. This is used to limit the contribution ratio of each texture item.

[0175]

[0176] Among them, the effective abrasive grain spacing evaluation sub-value This represents the abrasive grain distribution state, taking a non-negative value, and its function is to determine the abrasive grain sparseness or shedding trend; the current average effective abrasive grain spacing. This represents the average effective abrasive grain spacing within the current measurement area, using a non-negative length value. Its function is to characterize the density of the abrasive grain center distribution; the reference effective abrasive grain spacing. This represents the effective abrasive grain spacing in the corresponding area of ​​the benchmark geomorphological data. It takes a positive length value and serves to provide a spacing reference.

[0177] Current abrasive surface density This represents the effective number of abrasive grains per unit area within the current measurement region, taking a non-negative value. Its function is to characterize the density of the cutting edge; the reference abrasive grain surface density. This represents the abrasive surface density of the corresponding area in the baseline geomorphological data, taking a positive value, and its function is to provide a density reference; spacing weight. and density weight Both are set to 0 to 1, and their sum is 1. Their function is to limit the contribution ratio of spacing and density.

[0178] Furthermore, the comprehensive evaluation value It accepts four types of evaluation sub-values, with anomaly types accepting specific feature fields, and adjustment suggestions accepting anomaly types. The retested data then returns to the benchmark alignment stage, forming the same inventive concept for collection, reconstruction, extraction, evaluation, and adjustment feedback.

[0179] In alternative implementations, the benchmark topographic data is not limited to new grinding wheel data; it can also use data from the same batch of qualified grinding wheels or topographic data from the most recent qualified dressing. The weights of the four evaluation sub-values ​​are determined manually, by process cards, or by calibration using historical qualified samples, but the output field remains the comprehensive evaluation value. Types of landform anomalies and repair recommendations.

[0180] In a preferred embodiment, the geomorphological evaluation calculation unit is installed in an industrial computer. The display interface shows the abnormal sectors using a grinding wheel circumferential unfolding diagram. The trimming equipment operation interface only receives text suggestions and sector locations, without directly altering the mechanical actions of the trimming equipment. After operator confirmation, trimming is performed, and then the method is used for retesting. Evaluation and verification indicators include the consistency of abnormal sector locations and the comprehensive evaluation value before and after trimming. The direction of change, the fading of pore blockage markers in the retest map, and the consistency between local abnormal protrusion markers and the peak position of the three-dimensional height field.

[0181] In one embodiment, an exposure threshold is set. Spacing threshold Pore ​​threshold Texture threshold and comprehensive threshold Each threshold is determined by a set of qualified benchmark samples. Generate, for example:

[0182]

[0183] Among them, the exposure threshold This represents the minimum acceptable boundary for the abrasive grain exposure evaluation sub-value in a qualified grinding wheel, taking a non-negative value. Its function is to determine insufficient abrasive grain exposure; the qualified benchmark sample set. This represents a set of topographical samples of new grinding wheels, qualified dressed grinding wheels, or historically qualified grinding wheels. It is a finite set of samples and serves to provide a source of evaluation thresholds.

[0184] No. Abrasive exposure evaluation sub-values ​​for each qualified sample Indicates the first in the qualified benchmark sample set one sample It takes a non-negative value and its function is to form the exposure threshold.

[0185] Similarly, spacing threshold Pore ​​threshold Texture threshold and comprehensive threshold All are composed of qualified benchmark sample sets The corresponding evaluation sub-value or comprehensive evaluation value is determined. If there is no qualified benchmark sample set, a fixed threshold is given by the process card. Anomaly judgment is suggested as follows:

[0186] when and At that time, the output abrasive particles are dulled or coated with a binder. And the current abrasive surface density Lower than the reference abrasive surface density At that time, the abrasive particles are shed. And the current pore area ratio Pore ​​area ratio below the benchmark At that time, the output orifice becomes blocked.

[0187] When the maximum local exposure height is higher than the corresponding height benchmark of the adjacent sector and the clarity confidence level is... Reaching the confidence threshold When multiple abnormal conditions are met simultaneously, the primary abnormality type is output in the order of pore blockage, local abnormal protrusion, abrasive grain shedding, abrasive grain dulling, or binder coating, while other abnormality types are retained as secondary abnormality types. This supplement gives the rule model definite decision boundaries and output rules.

[0188] In one embodiment, the set of geomorphic anomaly types is denoted as The set of repair suggestions is denoted as Establish a trimming suggestion mapping operator:

[0189]

[0190] Among them, the trimming suggestion mapping operator This represents the mapping relationship from geomorphic anomaly types to remediation recommendations, and its function is to transform evaluation results into operational suggestions. (Geomorphic anomaly type set) This includes abrasive dulling or bond coating, abrasive grain detachment, pore blockage, insufficient dressing, over-dressing, and localized abnormal protrusions. A collection of dressing recommendations is provided. This includes increasing dressing depth, adjusting dressing feed, shortening dressing intervals, partial re-inspection, re-dressing, and replacing the grinding wheel.

[0191] Those skilled in the art will recognize that the units and algorithm steps of the various examples described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are implemented in hardware or software depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for each specific application, but such implementation should not be considered beyond the scope of this application.

[0192] Those skilled in the art will understand that, for the sake of convenience and brevity, the specific working processes of the systems, devices, and units described above can be referred to the corresponding processes in the foregoing method embodiments, and will not be repeated here.

[0193] In the several embodiments provided in this application, it should be understood that the disclosed systems, apparatuses, and methods can be implemented in other ways. For example, the apparatus embodiments described above are merely illustrative; for instance, the division of units is only a logical functional division, and in actual implementation, there may be other division methods. For example, multiple units or components may be combined or integrated into another system, or some features may be ignored or not executed. Furthermore, the coupling or direct coupling or communication connection shown or discussed may be through some interfaces; the indirect coupling or communication connection between apparatuses or units may be electrical, mechanical, or other forms.

[0194] The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the units can be selected to achieve the purpose of this embodiment according to actual needs.

[0195] The above description is merely a specific embodiment of this application, but the scope of protection of this application is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the scope of the technology disclosed in this application should be included within the scope of protection of this application. Therefore, the scope of protection of this application should be determined by the scope of the claims.

Claims

1. A method for three-dimensional measurement of the surface morphology of a grinding wheel, executed by a three-dimensional measurement device for the surface morphology of a grinding wheel, characterized in that: include: The rotatable grinding wheel clamping stage is controlled to step in the circumferential sector, and the optical imaging unit is controlled to move along the width direction of the grinding wheel and the optical axis of the lens to acquire a multi-focal plane image sequence with sector markers, strip markers and focal plane height markers. Calculate the sharpness response of the same pixel in a multi-focal image sequence at different focal plane heights, determine the focus height of the pixel, and subtract the reference of the grinding wheel circumferential surface to generate a three-dimensional height field of the grinding wheel surface unfolding. Based on the three-dimensional height field and grayscale texture map corresponding to the coordinates of the grinding wheel surface, the abrasive region, the binder region and the pore region are separated, and the abrasive exposure features, effective abrasive spacing features and pore connectivity features are extracted. The surface geomorphology evaluation results of the grinding wheel are output by comparing the abrasive grain exposure characteristics, effective abrasive grain spacing characteristics, and pore connectivity characteristics with the benchmark geomorphological data.

2. The method for three-dimensional measurement of grinding wheel surface morphology according to claim 1, characterized in that: Calculate the sharpness response of the same pixel at different focal plane heights in a multi-focal plane image sequence, including: The images of each focal plane under the same sector and the same strip are converted to grayscale, dark field subtraction, flat field correction and edge-preserving filtering are performed; edge response and gradient response are extracted in the local neighborhood of the same pixel to form sharpness response; focus height is generated according to the peak position of sharpness response and sharpness confidence is generated simultaneously.

3. The method for three-dimensional measurement of grinding wheel surface morphology according to claim 2, characterized in that: Subtracting the grinding wheel circumferential surface reference includes: The focus height is converted to the initial height; the grinding wheel circumferential surface reference is generated based on the nominal radius of the grinding wheel, the unfolded coordinates of the sector center, and the unfolded coordinates of the pixels; the grinding wheel circumferential surface reference is subtracted from the initial height to obtain the terrain height; Based on the overlapping areas corresponding to adjacent sector identifiers and adjacent strip identifiers, the terrain height is stitched together to form a three-dimensional height field unfolded on the grinding wheel surface.

4. The method for three-dimensional measurement of grinding wheel surface morphology according to claim 3, characterized in that: Separate the abrasive region, binder region, and pore region, including: Based on the three-dimensional height field, grayscale texture map, and sharpness confidence of the grinding wheel surface, abrasive evidence, bond evidence, and porosity evidence are generated; candidate abrasive regions are formed according to the abrasive evidence, porosity regions are formed according to the porosity evidence, and bond regions are formed according to the bond evidence. The adhered candidate abrasive grain regions are divided into effective abrasive grain regions according to the height valley line and the boundary of the pore region; after spindle correction, the abrasive grain exposure features and effective abrasive grain spacing features are extracted from the effective abrasive grain regions.

5. The method for three-dimensional measurement of grinding wheel surface morphology according to claim 1, characterized in that: The optical imaging unit has a preset imaging mode, which is either an area array imaging mode or a linear array imaging mode. When the preset imaging mode is an area array imaging mode, the optical imaging unit performs a single exposure for each circumferential sector and each width strip to generate the corresponding focal plane image. When the preset imaging mode is linear array imaging mode, the angle feedback of the rotatable grinding wheel holding stage is synchronized with the line scan trigger signal, generating the corresponding focal plane image line by line.

6. The method for three-dimensional measurement of grinding wheel surface morphology according to claim 1, characterized in that: When generating a three-dimensional height field for the surface of the grinding wheel, height registration is performed on the overlapping areas of adjacent local height fields; When two height values ​​exist at the same spatial location in the overlapping area, the height value with higher clarity and confidence is written into the grinding wheel surface to unfold a three-dimensional height field. When there are no two height values ​​at the same spatial location in the overlapping area, the existing height value at that spatial location is retained and written into the three-dimensional height field unfolded on the grinding wheel surface.

7. The method for three-dimensional measurement of grinding wheel surface morphology according to claim 6, characterized in that: The benchmark geomorphological data includes at least one of the following: new grinding wheel benchmark geomorphological data, qualified modified geomorphological data, and historical geomorphological data of the same grinding wheel; The comparison is aligned according to the sector identifier and the strip identifier; when the abrasive grain exposure feature is lower than the reference range and the pore connectivity feature is within the reference range, the abrasive grain dullness evaluation is output; when the pore connectivity feature is lower than the reference range, the pore blockage evaluation is output; when none of the above conditions are met, the normal evaluation is output.

8. The method for three-dimensional measurement of grinding wheel surface morphology according to claim 7, characterized in that: When outputting the evaluation results of the surface topography of the grinding wheel, the three-dimensional measurement equipment for the surface topography of the grinding wheel generates a grinding wheel circumference unfolding diagram, a three-dimensional height pseudo-color diagram, abrasive boundary markers, pore boundary markers, and abnormal sector markers on the display interface. Abnormal sector markers are bound to sector identifiers and strip identifiers; when an abnormal sector marker is selected, the display interface shows the corresponding abrasive grain exposure characteristics, effective abrasive grain spacing characteristics, and pore connectivity characteristics.

9. The method for three-dimensional measurement of grinding wheel surface morphology according to claim 8, characterized in that: The three-dimensional measurement equipment for grinding wheel surface morphology generates inspection data files; The test data file includes fields for grinding wheel number, sector identifier, strip identifier, focal plane height identifier, focusing height, three-dimensional height field index of grinding wheel surface development, abrasive region number, pore region number, and grinding wheel surface topography evaluation results. The test data file is written to the local storage area after each measurement is completed.

10. The method for three-dimensional measurement of grinding wheel surface morphology according to claim 1, characterized in that: The three-dimensional measurement equipment for grinding wheel surface morphology encapsulates the detection data file into a detection result message; The test result message includes a frame header, device number, measurement task number, sector identifier set, strip identifier set, focal plane height identifier set, evaluation result field, and verification field; when the test result message is successfully sent, the three-dimensional measurement equipment for grinding wheel surface morphology records the sending completion status; When the detection result message fails to be sent, the three-dimensional measurement equipment for the surface morphology of the grinding wheel records the pending transmission status.

Images