Gear wear state image online monitoring and evaluation method and system

Abstract

This invention discloses an online monitoring and evaluation method and system for gear wear state images, relating to the field of gear monitoring technology. The method includes: synchronously triggering and acquiring images of the target tooth surface under zoned illumination; establishing tooth number index relationships and conjugate meshing relationships according to the synchronous triggering sequence; merging the left oblique illumination image, the forward illumination image, and the right oblique illumination image according to the tooth number index to generate a current page group and a reference page group; extracting the tooth tip edge line, tooth root edge line, left and right tooth profile edge lines, and stable texture bands on the tooth surface from the current page group and the reference page group; and performing effective area clipping, tooth surface unfolding, and two-level alignment to generate an aligned unfolded page and a contact band baseline. This invention, through forward mapping, paired verification, reverse projection, and comparison of left and right adjacent tooth contact band migration, can more reliably distinguish between real wear and isolated anomalies.

Description

Technical Field

[0001] This invention relates to the field of gear monitoring technology, and in particular to a method and system for online monitoring and evaluation of gear wear condition images. Background Technology

[0002] As gear transmission devices evolve towards higher speeds, heavier loads, and longer continuous operation, monitoring methods for tooth surface condition have expanded from downtime disassembly and inspection to online sensing and image evaluation. Existing technologies typically deploy industrial cameras and auxiliary light sources in the gearbox observation window or endoscopic channel, combining them with angle encoders or speed synchronization signals to acquire tooth surface images. These images are then used to detect pitting, spalling, scratches, streaks, and local anomalies in the tooth profile through methods such as tooth surface boundary extraction, image unfolding, image registration, surface texture analysis, and contact trace recognition. Simultaneously, some methods incorporate benchmark image comparison, local tooth surface undulation recovery, meshing correspondence analysis, and temporal difference tracking to continuously characterize gear wear evolution. This technology has formed a technical path primarily based on image acquisition, geometric alignment, morphological analysis, and condition determination.

[0003] However, existing technologies still have shortcomings in practical applications: First, although images from different acquisition times, different lighting directions, and different tooth sequence positions can be acquired, the stable correspondence of the same tooth surface, the consistency of local morphology under the same unfolding coordinates, and the continuous alignment of the contact zone area are still insufficient. This can easily lead to positional drift of the wear area during comparison, thereby weakening the accuracy of wear evolution determination. Second, existing methods focus more on the identification of local anomalies on a single tooth surface, and do not make sufficient use of the meshing correspondence between the abnormal area and the paired tooth surface, the dual disturbance relationship, and the migration and expansion relationship of adjacent tooth surfaces. It is difficult to incorporate local wear, meshing re-verification, and adjacent tooth expansion into a unified determination chain. Therefore, there is still room for improvement in confirming actual wear and distinguishing wear levels. Summary of the Invention

[0004] In view of the aforementioned existing problems, the present invention is proposed.

[0005] Therefore, this invention provides an online monitoring and evaluation method for gear wear state images to solve the problems of difficulty in stable correspondence of the same tooth surface and insufficient determination of meshing correlation in online gear wear evaluation.

[0006] To solve the above-mentioned technical problems, the present invention provides the following technical solution: In a first aspect, the present invention provides an online monitoring and evaluation method for gear wear condition images, comprising, The target tooth surface is subjected to zoned illumination synchronous trigger acquisition, and tooth number index relationship and conjugate meshing relationship are established according to the synchronous triggering time sequence. The left oblique illumination image, the front illumination image and the right oblique illumination image are merged according to the tooth number index to generate the current page group and the reference page group. The current page group and the base page group are called to extract the tooth tip edge line, tooth root edge line, left and right tooth profile edge lines and tooth surface stable texture band, and perform effective area clipping, tooth surface expansion and two-level alignment to generate the aligned expansion page and contact band baseline; Dynamic illumination registration, local undulation reconstruction, curvature difference localization, and directional stre sorting are performed on the aligned unfolded pages and contact zone baselines to generate curvature abrupt change zones, stre-associated pages, and wear candidate pages. Based on the tooth number index relationship and conjugate meshing relationship, perform forward mapping, pairing verification and reverse resubmission on the wear candidate pages to determine the effective wear area, and perform contact band migration comparison on the left and right adjacent tooth expansion pages corresponding to the effective wear area to generate bidirectional verification pages and expansion pages; The system collects data from curvature abrupt change zones, streak-related pages, bidirectional return pages, and extended pages, performs tooth number sequence comparison and wear level determination, and generates online assessment results of gear wear status.

[0007] As a preferred embodiment of the online monitoring and evaluation method for gear wear state images described in this invention, the establishment of gear number index relationships and conjugate meshing relationships specifically includes: Acquire the pulse sequence of the corner encoder and divide the single-tooth visible area to generate a synchronous trigger timing sequence; Based on the tooth sequence position of each single tooth's visible interval according to the synchronous trigger timing mark, a tooth number index relationship is generated; Determine the mating tooth surface that meshes with the target tooth surface according to the tooth number index relationship, and match the corresponding tooth sequence position of the mating tooth surface; The meshing order of the target tooth surface and the paired tooth surface is arranged according to the synchronous triggering sequence to generate a conjugate meshing relationship.

[0008] As a preferred embodiment of the online monitoring and evaluation method for gear wear state images described in this invention, the generation of the current page group and the reference page group specifically includes: Within the same single-tooth visible range, continuously acquire left oblique illumination image, front illumination image and right oblique illumination image at the current acquisition time to generate the current single-tooth image group; Within the same single-tooth visible range, continuously acquire left oblique illumination image, front illumination image and right oblique illumination image at the reference acquisition time to generate a reference single-tooth image group; The current single-tooth image group and the reference single-tooth image group are arranged in order of tooth sequence position according to the tooth number index relationship, and the current page group and the reference page group are generated.

[0009] As a preferred embodiment of the online monitoring and evaluation method for gear wear state images described in this invention, the generation of the aligned unfolded page and contact band baseline specifically includes: Extract the tooth tip edge line, tooth root edge line, left and right tooth profile edge lines, and tooth surface stable texture band from the current page group and the base page group; The tooth surface is enclosed by the tooth tip edge, tooth root edge, and left and right tooth profile edges, and the working area of ​​the tooth surface is defined by the tooth surface stable texture band, thus generating an effective area boundary. Extract the tooth surface image based on the effective region boundary and generate a cropped page; The cut page is unfolded along the tooth profile direction and the tooth width direction to generate a tooth surface unfolded page, which includes the current unfolded page and the reference unfolded page; Using the baseline expanded page as the positioning page, perform overall and partial alignment on the current expanded page to generate an aligned expanded page; Extract stable contact traces from the aligned unfolded page to generate the contact band baseline.

[0010] As a preferred embodiment of the online monitoring and evaluation method for gear wear state images described in this invention, the generation of curvature abrupt change bands, streak-related pages, and wear candidate pages specifically includes: Perform position correction and brightness correspondence on the partitioned lighting images in the alignment expansion page to generate a lighting registration page; Call the lighting registration page to restore the local undulations of the tooth surface and generate the current curvature page; Retrieve the reference curvature page corresponding to the reference page group and perform difference localization to generate curvature abrupt change bands; In the region where the curvature abrupt change zone coincides with the contact zone baseline, the stre direction, stre connected segments, and stre bifurcation segments are extracted to generate stre-associated pages; Abnormal regions in curvature change bands and streak-related pages are aggregated to generate wear candidate pages.

[0011] As a preferred embodiment of the online monitoring and evaluation method for gear wear state images described in this invention, the determination of the effective wear area specifically includes: The tooth number index relationship is used to locate the candidate abnormal area in the wear candidate page, and the corresponding expansion area of ​​the paired tooth surface is determined according to the conjugate meshing relationship; The candidate anomaly region is projected onto the corresponding expanded region and contact band interruption, contact band narrowing, local offset and dual streak are extracted to generate a paired verification region. According to the conjugate meshing relationship, the paired back-affirmation area is back-projected onto the alignment unfolding page of the target tooth surface to generate the reverse back-projection area; Compare the overlap range between the candidate abnormal area and the reverse return area, and filter out the common area that falls into the wear candidate page at the same time to generate the effective wear area.

[0012] As a preferred embodiment of the online monitoring and evaluation method for gear wear state images described in this invention, the generation of bidirectional return pages and extended pages specifically includes: Retrieve the unfolded pages of the left and right adjacent teeth corresponding to the tooth number of the effective wear area, and locate the contact band segment on the same side as the effective wear area in the unfolded pages of the left and right adjacent teeth to generate the adjacent tooth contact segment; The paired return verification area, reverse return area, effective wear area and adjacent tooth contact segment are spliced ​​together according to tooth sequence to generate adjacent tooth comparison page; In the adjacent tooth comparison page, the contact band direction, contact band width and stripe connection status are compared, and the migration segments extending from the effective wear area to the left and right adjacent teeth are extracted to generate an extended sequence; The effective wear area, the paired return verification area, and the reverse return area are superimposed on the alignment unfolding page of the target tooth surface to generate a bidirectional return verification page; Arrange the bidirectional return pages and extended sequences accordingly to generate extended pages.

[0013] As a preferred embodiment of the online monitoring and evaluation method for gear wear state images described in this invention, the step of performing tooth number sequence comparison specifically includes: Arrange the curvature abrupt change zone, the streak-related page, the bidirectional return page, and the extended page according to the tooth number index relationship, and generate the current tooth number sequence page; Retrieve the historical tooth number sequence page corresponding to the previous online detection and the baseline tooth number sequence page corresponding to the baseline page group, and generate a comparison page group; Compare the contact band occupancy, stripe continuity, and return range in the current tooth number sequence page, historical tooth number sequence pages, and reference tooth number sequence pages one by one to generate a sequence difference page.

[0014] As a preferred embodiment of the online monitoring and evaluation method for gear wear state images described in this invention, the wear level determination specifically includes: The abnormal distributions in the sequence difference page, the bidirectional return confirmation page, and the extended page are collected to generate the wear determination page; Based on the location of the abnormal distribution in the contact zone, the corresponding disturbance of the paired tooth surface, and the continuous migration of the left and right adjacent teeth in the wear judgment page, mild judgment results, moderate judgment results, and severe judgment results are generated and summarized to obtain the online evaluation results of gear wear status.

[0015] Secondly, the present invention provides an online monitoring and evaluation system for gear wear state images, including a relationship establishment module, which performs partitioned illumination synchronous triggering acquisition on the target tooth surface, establishes tooth number index relationship and conjugate meshing relationship according to the synchronous triggering sequence, merges the left oblique illumination image, the forward illumination image and the right oblique illumination image according to the tooth number index, and generates the current page group and the reference page group. The alignment module is expanded by calling the current page group and the base page group to extract the tooth tip edge line, tooth root edge line, left and right tooth profile edge lines and tooth surface stable texture band, and performing effective area clipping, tooth surface expansion and two-level alignment to generate the aligned expanded page and contact band baseline; The candidate extraction module performs dynamic illumination registration, local undulation reconstruction, curvature difference localization, and directional stre sorting on the aligned unfolded page and contact zone baseline to generate curvature change zone, stre-associated page, and wear candidate page; The return certificate extension module performs forward mapping, pairing return certificates, and reverse return on the wear candidate pages according to the tooth number index relationship and conjugate meshing relationship, determines the effective wear area, and performs contact band migration comparison on the left and right adjacent tooth expansion pages corresponding to the effective wear area to generate bidirectional return certificate pages and expansion pages; The condition assessment module combines the curvature change zone, the scratch-related page, the bidirectional return page, and the extended page to perform tooth number sequence comparison and wear level determination, and generates online assessment results of gear wear condition.

[0016] The beneficial effects of this invention are as follows: By synchronously triggering acquisition with partitioned illumination, establishing tooth number index relationships, and establishing conjugate meshing relationships, stable correspondences are formed between tooth surface images at different acquisition times in tooth sequence positions; through effective region clipping, tooth surface unfolding, and overall and local alignment processing, the current page group and the reference page group obtain directly comparable aligned unfolded pages and contact zone baselines under unified unfolding coordinates, thereby improving the spatial consistency of tooth surface comparison; through dynamic illumination registration, local undulation reconstruction, curvature difference positioning, and directional streak sorting, surface morphology changes and streak evolution within the contact zone are jointly characterized, which helps to improve the targeting of wear candidate region extraction; through forward mapping, paired verification, reverse projection, and left and right adjacent tooth contact zone migration comparison, the target tooth surface anomaly, paired tooth surface disturbance, and adjacent tooth expansion state are connected, enabling more reliable differentiation between real wear and isolated anomalies; through tooth number sequence comparison and wear level determination, continuous and structured evaluation results are formed for online monitoring scenarios, thereby improving the accuracy and consistency of gear wear state identification. Attached Figure Description

[0017] To more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings used in the following description of the embodiments will be briefly introduced. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0018] Figure 1 This is a flowchart of an online monitoring and evaluation method for gear wear condition images.

[0019] Figure 2 Flowchart for unfolding alignment and baseline generation.

[0020] Figure 3 This is a flowchart for curvature abrupt change and streak recognition.

[0021] Figure 4Flowchart for verification of evidence extension and wear confirmation. Detailed Implementation

[0022] To make the above-mentioned objects, features and advantages of the present invention more apparent and understandable, the specific embodiments of the present invention will be described in detail below with reference to the accompanying drawings.

[0023] Many specific details are set forth in the following description in order to provide a full understanding of the invention. However, the invention may also be practiced in other ways different from those described herein, and those skilled in the art can make similar extensions without departing from the spirit of the invention. Therefore, the invention is not limited to the specific embodiments disclosed below.

[0024] Secondly, the term "one embodiment" or "embodiment" as used herein refers to a specific feature, structure, or characteristic that may be included in at least one implementation of the present invention. The phrase "in one embodiment" appearing in different places in this specification does not necessarily refer to the same embodiment, nor is it a single or selective embodiment that is mutually exclusive with other embodiments.

[0025] Reference Figures 1-4 As one embodiment of the present invention, this embodiment provides an online monitoring and evaluation method for gear wear condition images, including the following steps: S1. Perform zoned illumination synchronous trigger acquisition on the target tooth surface, and establish tooth number index relationship and conjugate meshing relationship according to the synchronous triggering sequence. Merge the left oblique illumination image, the forward illumination image and the right oblique illumination image according to the tooth number index to generate the current page group and the reference page group.

[0026] S1.1. Fix an endoscopic industrial camera, a ring-shaped partition light source, and a corner encoder at the gearbox viewing window. Adjust the shooting direction of the endoscopic industrial camera so that its imaging center covers the working tooth surface area of ​​the target tooth surface. Adjust the left oblique illumination area, the forward illumination area, and the right oblique illumination area of ​​the ring-shaped partition light source so that the left oblique illumination direction, the forward illumination direction, and the right oblique illumination direction fall sequentially within the same imaging area of ​​the target tooth surface. Then, read the pulse sequence of the corner encoder and record the pulse segments when the target tooth surface enters and leaves the viewing window, generating a single-tooth viewing interval. The single-tooth viewing interval is used to define the acquisition window corresponding to each tooth surface, and subsequent synchronous triggering timing is established within the single-tooth viewing interval.

[0027] The single-tooth visible intervals corresponding to the pulse sequence of the corner encoder are arranged sequentially. A synchronous triggering sequence is established according to the starting pulse position of the single-tooth visible interval. The left oblique illumination triggering time, the forward illumination triggering time, and the right oblique illumination triggering time are sequentially corresponding to each single-tooth visible interval, so that the left oblique illumination image, the forward illumination image, and the right oblique illumination image all fall within the same single-tooth visible interval.

[0028] S1.2. Based on the current angular position of the target tooth surface corresponding to the synchronous triggering sequence, and combined with the number of teeth of the target gear, mark the tooth sequence position for each single-tooth visible interval, and generate a tooth number index relationship. The tooth number index relationship is established by using a method where each single-tooth visible interval corresponds to one tooth sequence position. The tooth sequence positions are numbered consecutively starting from the first tooth surface of the target gear that enters the visible area of ​​the observation window. Each single-tooth visible interval in the synchronous triggering sequence corresponds to a unique tooth sequence position.

[0029] The matching tooth surfaces that mesh with the target tooth surface are determined according to the tooth number index relationship, and the corresponding tooth sequence positions of the matching tooth surfaces are matched. The meshing order of the target tooth surface and the matching tooth surface is arranged along the synchronous triggering time sequence to generate a conjugate meshing relationship.

[0030] To further explain, the establishment of the conjugate meshing relationship is completed using a slow rotation verification method: First, a known tooth sequence position in the tooth number index relationship is brought into the center of the visible area of ​​the observation window. Then, the mating tooth surface in contact with the known tooth sequence position is confirmed along the gear meshing direction and recorded as the corresponding tooth sequence position. Subsequently, the target gear continues to rotate along the synchronous triggering sequence, and the meshing order of the target tooth surface and the mating tooth surface is recorded tooth by tooth according to the continuous meshing sequence of the gear pair, forming a correspondence of "target tooth surface tooth sequence position - mating tooth surface tooth sequence position - meshing order position". The same recording is completed for all tooth sequence positions to obtain the complete conjugate meshing relationship.

[0031] S1.3. Under synchronous trigger timing control, continuously acquire the left oblique illumination image, the front illumination image, and the right oblique illumination image at the current acquisition time within the same single-tooth visible range. Then, map the left oblique illumination image, the front illumination image, and the right oblique illumination image to the same tooth sequence position according to the tooth number index relationship to generate the current single-tooth image group. Next, at the reference acquisition time after the equipment is initially installed or after maintenance, using the same synchronous trigger timing, the same single-tooth visible range, and the same tooth number index relationship as the current acquisition time, continuously acquire the left oblique illumination image, the same front illumination image range, and the same tooth number index relationship. Map the left oblique illumination image, the same front illumination image, and the right oblique illumination image to the same tooth sequence position to generate the reference single-tooth image group. Arrange all current single-tooth image groups in the tooth sequence position order corresponding to the tooth number index relationship to generate the current page group. Finally, arrange all reference single-tooth image groups in the tooth sequence position order corresponding to the tooth number index relationship to generate the reference page group.

[0032] S2. Call the current page group and the base page group to extract the tooth tip edge line, tooth root edge line, left and right tooth profile edge lines and tooth surface stable texture band, and perform effective area clipping, tooth surface expansion and two-level alignment to generate the aligned expanded page and contact band baseline.

[0033] S2.1. Call the current page group and the reference page group, read the current single tooth image group and the reference single tooth image group at the corresponding tooth sequence position according to the tooth number index relationship, and extract the tooth tip edge line, tooth root edge line, left tooth profile edge line and right tooth profile edge line from the forward illumination image of the current single tooth image group and the reference single tooth image group.

[0034] Among them, the tooth tip edge line refers to the continuous boundary line of the working tooth surface near the tooth tip circle, the tooth root edge line refers to the continuous boundary line of the working tooth surface near the tooth root transition zone, and the left tooth profile edge line and the right tooth profile edge line refer to the tooth profile boundary lines on both sides of the working tooth surface.

[0035] After extracting the tooth tip edge line, tooth root edge line, left tooth profile edge line, and right tooth profile edge line, the tooth surface stable texture band is extracted from the forward illumination images of the current single tooth image group and the reference single tooth image group. The tooth surface stable texture band refers to the texture area located in the middle of the working tooth surface, which is continuously distributed at different acquisition times and is not affected by the tooth tip reflective area and the tooth root shadow area. The tooth surface stable texture band is used to define the boundaries of the subsequent effective area and the local alignment area.

[0036] The tooth surface is bounded by the tooth tip edge, tooth root edge, left tooth profile edge, and right tooth profile edge. A stable texture band on the tooth surface is used to define the working area, resulting in the effective region boundaries for the current single-tooth image group and the reference single-tooth image group. These effective region boundaries are used to remove highlighted areas at the tooth tip, deep shadow areas at the tooth root, and external background areas on the tooth flank.

[0037] S2.2. Based on the effective region boundary corresponding to the current single-tooth image group, perform the same boundary cropping on the left oblique illumination image, the front illumination image, and the right oblique illumination image in the current single-tooth image group to obtain the current cropping page; based on the effective region boundary corresponding to the reference single-tooth image group, perform the same boundary cropping on the left oblique illumination image, the front illumination image, and the right oblique illumination image in the reference single-tooth image group to obtain the reference cropping page. The current cropping page and the reference cropping page maintain the same tooth sequence position, the same boundary range, and the same illumination order, and the correspondence is used for subsequent tooth surface unfolding.

[0038] The current cut page and the reference cut page are unfolded along the tooth profile direction and the tooth width direction to generate a tooth surface unfolded page. The tooth profile direction refers to the direction from the tooth root edge line to the tooth tip edge line, and the tooth width direction refers to the direction extending along the tooth surface width. The tooth surface unfolded page includes the current unfolded page and the reference unfolded page, where the current unfolded page is obtained by unfolding the current cut page, and the reference unfolded page is obtained by unfolding the reference cut page.

[0039] To ensure consistent coordinates for the same tooth position at different acquisition times, the tooth tip and root edges are mapped to the upper and lower boundaries of the unfolded page, respectively, while the left and right tooth profile edges are mapped to the left and right boundaries, respectively. This ensures that the current unfolded page and the reference unfolded page maintain the same arrangement in both the tooth profile and tooth width directions. After tooth surface unfolding is complete, the left oblique illumination unfolded image, the front illumination unfolded image, and the right oblique illumination unfolded image in the current unfolded page have a point-to-point correspondence with their corresponding unfolded images in the reference unfolded page. Subsequent two-level alignment is performed around the current unfolded page and the reference unfolded page.

[0040] S2.3. Using the reference unfolded page as the positioning page, perform overall alignment on the current unfolded page. During overall alignment, first use the tooth tip edge line, tooth root edge line, left tooth profile edge line, and right tooth profile edge line in the reference unfolded page to determine the outer contour position of the reference unfolded page. Then use the corresponding edge lines in the current unfolded page to adjust the translation position and rotation angle of the current unfolded page so that the outer contour position of the current unfolded page coincides with the outer contour position of the reference unfolded page, thus obtaining the overall aligned page. The overall aligned page is used to eliminate the overall positional deviation between the current acquisition time and the reference acquisition time.

[0041] Within the corresponding areas of the overall aligned page and the reference unfolded page, local alignment is performed using a tooth surface stable texture band. During local alignment, a continuous texture area is selected within the tooth surface stable texture band, and the texture positions of the overall aligned page and the reference unfolded page are compared segment by segment to ensure that the texture undulation positions of the tooth surface stable texture band are consistent in the tooth profile direction and tooth width direction, thus obtaining the aligned unfolded page. The aligned unfolded page maintains the lighting order of the current unfolded page and is located in the same unfolding coordinate system as the reference unfolded page.

[0042] After completing the alignment and unfolding of the page, stable contact traces are extracted from the forward-illuminated unfolded image of the aligned and unfolded page. Stable contact traces refer to contact band-like traces that are continuously distributed along the tooth width direction, have minimal positional variation along the tooth profile direction, and maintain consistent positions in adjacent tooth sequences. The centerline of the contact band is extracted along the middle position of the stable contact trace to obtain the contact band baseline.

[0043] S3. Perform dynamic illumination registration, local undulation reconstruction, curvature difference localization, and directional streak sorting on the aligned unfolded pages and contact zone baselines to generate curvature change zones, streak-related pages, and wear candidate pages.

[0044] S3.1. Invoke the alignment unfolded page and contact band baseline. Read the left oblique illumination unfolded image, the front illumination unfolded image, and the right oblique illumination unfolded image corresponding to the same tooth sequence position, according to the tooth number index. Confine these images to the tooth surface area where the contact band baseline is located, forming a set of partitioned illumination images for the same tooth sequence position. This set of partitioned illumination images is used to eliminate positional offsets and brightness deviations between different illumination directions.

[0045] Position correction is performed on the left oblique illumination unfolded image, the front illumination unfolded image, and the right oblique illumination unfolded image to ensure that the tooth surface boundary positions in the left oblique illumination unfolded image, the front illumination unfolded image, and the right oblique illumination unfolded image coincide within the same unfolded coordinate system. After position correction, brightness correspondence is performed on the left oblique illumination unfolded image, the front illumination unfolded image, and the right oblique illumination unfolded image to ensure that the grayscale changes at the same tooth surface position in the three unfolded images only retain the difference in illumination direction, thus obtaining the illumination registration page. Each pixel position in the illumination registration page corresponds simultaneously to the left oblique illumination brightness, the front illumination brightness, and the right oblique illumination brightness. The left oblique illumination brightness, the front illumination brightness, and the right oblique illumination brightness together characterize the light and dark response of the same tooth surface position under different illumination directions. Subsequent local undulation reconstruction is performed point by point according to the illumination registration page.

[0046] S3.2. When calling the illumination registration page to restore the local undulations of the tooth surface, read the left oblique illumination intensity, the forward illumination intensity, and the right oblique illumination intensity point by point along the illumination registration page, and restore the local height changes of the tooth surface based on the relative changes between the left oblique illumination intensity, the forward illumination intensity, and the right oblique illumination intensity to obtain the current curvature page. Each position in the current curvature page corresponds to a local undulation change result, which is used to represent the surface curvature changes of the tooth surface in the tooth profile direction and the tooth width direction.

[0047] To ensure the current curvature page maintains the same judgment scale as the reference state, the reference unfolded page corresponding to the same tooth sequence position in the reference page group is called according to the tooth number index relationship. The left oblique illumination unfolded image, the front illumination unfolded image, and the right oblique illumination unfolded image in the reference unfolded page are processed using the same position correction and brightness correspondence methods as the illumination registration page to obtain the reference illumination registration page. Based on the reference illumination registration page, the local undulations of the tooth surface in the reference state are recovered to obtain the reference curvature page. The current curvature page and the reference curvature page maintain consistency in tooth sequence position, unfolded coordinates, and illumination order. Subsequent curvature difference localization is performed around the current curvature page and the reference curvature page.

[0048] By continuously merging the curvature difference values ​​at all tooth surface development coordinate positions, curvature abrupt change zones are obtained. Curvature abrupt change zones refer to band-shaped regions that are continuously distributed along the tooth profile direction or tooth width direction in the tooth surface development page and show obvious morphological changes relative to the reference curvature page.

[0049] S3.3. Overlay the curvature change zone with the contact zone baseline, and extract the tooth surface area where the curvature change zone and the contact zone baseline coincide to obtain the streak combing area. The streak combing area refers to the tooth surface area that simultaneously satisfies the curvature difference change and the contact zone passage position. The streak combing area is used to avoid invalid streak extraction in non-contact areas and areas without curvature change.

[0050] Gray-scale sequences were extracted along different directions within the streak region, and permutation entropy was calculated for the gray-scale sequences in different directions to determine the direction with the minimum permutation entropy as the streak direction. The direction with the minimum permutation entropy indicates the direction with the most stable gray-scale permutation pattern. When the direction with the minimum permutation entropy coincides with the extension direction of the abrasive sliding streak, the streak boundary continuity in the left oblique illumination unfolded image, the front illumination unfolded image, and the right oblique illumination unfolded image is the clearest. Subsequent streak connected segments and streak bifurcation segments are extracted along the direction with the minimum permutation entropy.

[0051] To further explain, the expression for calculating the permutation entropy of grayscale sequences in different directions is as follows: ; in, The permutation entropy of a grayscale sequence in a certain direction; Represents the number of grayscale sequences. The probability of a certain permutation pattern occurring; This indicates the total number of permutation patterns.

[0052] It should be noted that the permutation entropy calculation uses the entropy calculation method in information theory, which is used to characterize the degree of disorder in the arrangement of gray-level sequences in a certain direction. The smaller the permutation entropy, the stronger the continuity of the gray-level sequence along the corresponding direction, and the closer the corresponding direction is to the stripe direction.

[0053] After determining the streak direction, continuously varying grayscale regions are extracted along the streak direction to obtain connected streak segments. Grayscale bifurcation regions are extracted at locations where the streak direction separates to obtain bifurcation segments. Within connected streak segments, regions that extend continuously but experience midway grayscale breaks are identified to obtain the locations of broken streaks. The streak direction, connected streak segments, broken streak locations, and bifurcation segments are then merged to obtain streak companion pages. These companion pages are used to characterize the extension, breakage, and bifurcation relationships of streaks within the same abnormal area on the tooth surface.

[0054] The abnormal regions in the curvature abrupt change zone and the streak-related abnormal regions in the streak-associated page are merged. Abnormal regions appearing simultaneously in both the curvature abrupt change zone and the streak-associated page are retained as primary abnormal regions. Abnormal regions appearing only in the curvature abrupt change zone and connected to the boundary of the streak-associated page are retained as associated abnormal regions. These primary and associated abnormal regions are then uniformly organized to obtain the wear candidate page. All abnormal regions in the wear candidate page correspond to the same tooth sequence position, the same unfolding coordinates, and the same contact zone region.

[0055] S4. Perform forward mapping, pairing verification, and reverse resubmission on the wear candidate pages according to the tooth number index relationship and conjugate meshing relationship to determine the effective wear area, and perform contact zone migration comparison on the left and right adjacent tooth expansion pages corresponding to the effective wear area to generate bidirectional verification pages and expansion pages.

[0056] S4.1. Call the wear candidate page, read the wear candidate page corresponding to the same tooth sequence position one by one according to the tooth number index relationship, and extract continuous abnormal areas in the wear candidate page to obtain the candidate abnormal area. The candidate abnormal area refers to the set of abnormal areas located in the same tooth sequence position, the same unfolding coordinate and the same contact zone area. Each abnormal area in the candidate abnormal area is kept in one-to-one correspondence with the curvature change zone and the streak companion page.

[0057] After extracting the candidate anomaly area, the conjugate meshing relationship is invoked. The tooth sequence position of the paired tooth surface that meshes with the target tooth surface is found according to the tooth sequence position of the target tooth surface corresponding to the candidate anomaly area. Then, the aligned unfolded page corresponding to the tooth sequence position of the paired tooth surface is retrieved according to the tooth number index relationship to obtain the corresponding unfolded area of ​​the paired tooth surface. The corresponding unfolded area of ​​the paired tooth surface refers to the tooth surface area that has a meshing sequence relationship with the target tooth surface where the candidate anomaly area is located. The corresponding unfolded area of ​​the paired tooth surface and the candidate anomaly area maintain a meshing correspondence in the tooth surface unfolded coordinate.

[0058] Based on the recorded meshing sequence in the conjugate meshing relationship, the candidate abnormal area is mapped along the tooth profile and tooth width directions to the corresponding unfolded area of ​​the mating tooth surface, resulting in the positive mapping area. The positive mapping area is used to define the inspection range of the mating tooth surface that has a direct meshing correspondence with the candidate abnormal area.

[0059] S4.2. Within the positive mapping area, examine the changes in contact band morphology and the corresponding relationship of streaks segment by segment, extracting contact band interruptions, contact band narrowing, local offsets, and paired streaks to obtain the paired verification area. Specifically, a contact band interruption refers to a band-shaped cessation area appearing along the continuous direction of the contact band; a contact band narrowing refers to a position where the width of the contact band decreases relative to the adjacent normal area; a local offset refers to a position where the center position of the contact band deviates from the adjacent area of ​​the same tooth sequence; and paired streaks refer to streak areas distributed along opposite contact sliding directions at the corresponding positions on the paired tooth surface and the candidate abnormal area.

[0060] To ensure a direct correspondence between the paired verification areas and the candidate anomaly areas, the contact band interruption region and the contact band narrowing region are first extracted within the forward mapping region. Then, the local offset region and the dual streak region are extracted. Subsequently, the positions of the contact band interruption region, the contact band narrowing region, the local offset region, and the dual streak region are merged to obtain the paired verification areas. Each verification area in the paired verification area corresponds to an anomaly region within the candidate anomaly area.

[0061] Based on the conjugate meshing relationship, the paired verification area is projected back from the corresponding unfolded area of ​​the paired tooth surface to the aligned unfolded page of the target tooth surface, resulting in the reverse projection area. The reverse projection area refers to the corresponding region after the paired verification area returns to the target tooth surface along the meshing correspondence between the tooth sequence position of the paired tooth surface and the tooth sequence position of the target tooth surface. The reverse projection area and the candidate anomaly area are located within the same target tooth surface unfolded coordinate system.

[0062] S4.3. Overlay the candidate anomaly area and the reverse projection area onto the aligned unfolded page of the same target tooth surface. Compare the boundary overlap range and region overlap range of the candidate anomaly area and the reverse projection area one by one, and then filter out the overlapping areas that fall into the wear candidate page at the same time to obtain the effective wear area. The effective wear area refers to the anomaly area that simultaneously satisfies the existence of target tooth surface anomaly, the existence of paired tooth surface verification, and the existence of reverse projection overlap. Each area in the effective wear area has a triple correspondence relationship of forward mapping, paired verification, and reverse projection.

[0063] To avoid misclassifying isolated noise areas or non-meshing anomaly areas as effective wear areas, isolated edge areas in candidate anomaly areas, isolated edge areas in reverse feedback areas, and overlapping areas not included in the wear candidate page range are not retained. The retained overlapping areas are then uniformly organized to obtain the effective wear area. The effective wear area is used to characterize the true wear area that maintains the meshing correspondence after being checked by the paired tooth surfaces.

[0064] The effective wear area, the mating verification area, and the reverse re-projection area are simultaneously overlaid onto the aligned unfolded page of the target tooth surface, preserving their boundary positions and area ranges, resulting in a bidirectional verification page. The bidirectional verification page is used to characterize the bidirectional correspondence between the abnormal areas of the target tooth surface, the mating tooth surface, and the re-projection abnormal area.

[0065] S4.4. Using the tooth index relationship corresponding to the effective wear area, retrieve the aligned unfolded pages corresponding to the tooth sequence positions to the left and right of the effective wear area, respectively, to obtain the left adjacent tooth unfolded page and the right adjacent tooth unfolded page. Locate the contact band segment on the same side as the effective wear area within the left and right adjacent tooth unfolded pages to obtain the adjacent tooth contact segment. The adjacent tooth contact segment refers to the contact band area located in the left and right adjacent tooth unfolded pages, distributed along the tooth width direction on the same side as the effective wear area.

[0066] The pairing verification area, reverse return area, effective wear area, and adjacent tooth contact segment are spliced ​​together according to tooth sequence to obtain the adjacent tooth comparison page. The adjacent tooth comparison page sequentially retains the adjacent tooth contact segment of the left adjacent tooth unfolded page, the effective wear area and reverse return area of ​​the target tooth surface, the pairing verification area of ​​the paired tooth surface, and the adjacent tooth contact segment of the right adjacent tooth unfolded page, so that the contact band change and streak change on the same tooth sequence chain are in a continuous arrangement state.

[0067] By comparing the contact band direction, contact band width, and streak concatenation status in the adjacent tooth comparison pages, the migration segments continuously extending from the effective wear area to the left or right adjacent tooth unfolded pages are extracted to obtain the extended sequence. The contact band direction refers to the extension direction of the contact band centerline along the tooth width and tooth profile direction; the contact band width refers to the distribution width of the contact band in the tooth width direction; the streak concatenation status refers to the continuous connection relationship between the connected segments of the streak on the target tooth surface and the left and right adjacent teeth; and the migration segment refers to the continuous abnormal segment that starts from the effective wear area and extends to the left or right adjacent tooth unfolded page.

[0068] After completing the extended sequence, the bidirectional return pages and the extended sequence are arranged in accordance with the tooth sequence position to obtain the extended pages. The extended pages simultaneously retain the effective wear area of ​​the target tooth surface, the pairing return area of ​​the paired tooth surface, the reverse return area of ​​the target tooth surface, and the migration segments in the expansion pages of the left and right adjacent teeth. The extended pages are used to characterize whether the wear is limited to a single tooth or has formed an extended sequence along the adjacent tooth surface.

[0069] S5. Collect curvature change zone, streak associated page, bidirectional return page and extension page, perform tooth number sequence comparison and wear level determination, and generate online evaluation results of gear wear status.

[0070] S5.1. Read the curvature abrupt change zone, streak-related page, bidirectional return confirmation page, and extension page corresponding to the same tooth sequence position one by one according to the tooth number index relationship, and arrange the curvature abrupt change zone, streak-related page, bidirectional return confirmation page, and extension page in the order of tooth sequence position to obtain the current tooth number sequence page. The current tooth number sequence page retains the contact zone occupancy, streak continuity, paired return confirmation range, and adjacent tooth extension position corresponding to each tooth sequence position. The current tooth number sequence page is used for tooth-by-tooth comparison with the historical tooth number sequence page and the reference tooth number sequence page.

[0071] After completing the current tooth number sequence page, retrieve the historical tooth number sequence page corresponding to the previous online detection, and retrieve the baseline tooth number sequence page corresponding to the baseline page group to obtain the comparison page group. The historical tooth number sequence page refers to the set of tooth sequence pages arranged according to the tooth number index relationship during the previous online detection, and the baseline tooth number sequence page refers to the set of tooth sequence pages arranged by the expanded results of the baseline state corresponding to the baseline page group. The current tooth number sequence page, historical tooth number sequence page, and baseline tooth number sequence page in the comparison page group maintain a corresponding relationship in tooth number position, expanded coordinates, and contact zone position.

[0072] The sequence difference page is obtained by comparing the contact band position, streak continuity, and return verification range in the current tooth number sequence page, historical tooth number sequence pages, and reference tooth number sequence pages. Contact band position refers to the location range of the contact band baseline and contact band area in the tooth surface unfolded page. Streak continuity refers to the continuity of connected streak segments within the same tooth sequence position and the degree of connection between adjacent tooth sequence positions. Return verification range refers to the area range corresponding to the effective wear zone, paired return verification zone, and reverse return zone in the bidirectional return verification page. The sequence difference page retains the changes in each tooth sequence position relative to the historical tooth number sequence page and the reference tooth number sequence page.

[0073] S5.2. The changes in contact band occupancy, streak continuity, and return range in the sequence difference page are combined with the corresponding bidirectional regions in the bidirectional return page and the adjacent tooth migration segments in the extended page to obtain the wear determination page. Each tooth sequence position in the wear determination page simultaneously retains the current abnormal distribution, the corresponding disturbance distribution of paired tooth surfaces, and the continuous migration distribution of left and right adjacent teeth. The wear determination page is used to uniformly distinguish different wear levels.

[0074] The wear assessment is categorized based on the location of the abnormal distribution in the contact zone, the corresponding disturbance on the paired tooth surface, and the continuous migration of the left and right adjacent teeth: A mild assessment result is defined as an abnormal distribution confined to the edge of the contact zone, a corresponding disturbance on the paired tooth surface limited to the corresponding position of the candidate abnormal area, and a tooth sequence position where the left and right adjacent teeth do not form continuous migration segments. A moderate assessment result is defined as an abnormal distribution entering the main contact zone, a corresponding disturbance on the paired tooth surface covering the corresponding position of the main contact zone, and a tooth sequence position where the left and right adjacent teeth exhibit unilateral continuous migration segments. A severe assessment result is defined as an abnormal distribution crossing the main contact zone, a corresponding disturbance on the paired tooth surface continuously expanding along the contact zone, and a tooth sequence position where the left and right adjacent teeth simultaneously exhibit continuous migration segments.

[0075] The results of mild, moderate, and severe gear wear assessments are summarized according to the tooth number index to obtain the online gear wear condition assessment results. The online gear wear condition assessment results retain the wear level distribution corresponding to each tooth sequence position, and also retain the correspondence between the current tooth number sequence page, sequence difference page, bidirectional verification page, and extension page.

[0076] This embodiment also provides an online monitoring and evaluation system for gear wear condition images, including: The relationship establishment module performs zoned illumination synchronous trigger acquisition on the target tooth surface, and establishes tooth number index relationship and conjugate meshing relationship according to the synchronous triggering sequence. It merges the left oblique illumination image, the forward illumination image and the right oblique illumination image according to the tooth number index to generate the current page group and the reference page group. The alignment module is expanded by calling the current page group and the base page group to extract the tooth tip edge line, tooth root edge line, left and right tooth profile edge lines and tooth surface stable texture band, and performing effective area clipping, tooth surface expansion and two-level alignment to generate the aligned expanded page and contact band baseline; The candidate extraction module performs dynamic illumination registration, local undulation reconstruction, curvature difference localization, and directional stre sorting on the aligned unfolded page and contact zone baseline to generate curvature change zone, stre-associated page, and wear candidate page; The return certificate extension module performs forward mapping, pairing return certificates, and reverse return on the wear candidate pages according to the tooth number index relationship and conjugate meshing relationship, determines the effective wear area, and performs contact band migration comparison on the left and right adjacent tooth expansion pages corresponding to the effective wear area to generate bidirectional return certificate pages and expansion pages; The condition assessment module combines the curvature change zone, the scratch-related page, the bidirectional return page, and the extended page to perform tooth number sequence comparison and wear level determination, and generates online assessment results of gear wear condition.

[0077] In summary, this invention establishes a stable correspondence between tooth surface images at different acquisition times in terms of tooth sequence position through partitioned illumination synchronous trigger acquisition, tooth number index relationship establishment, and conjugate meshing relationship establishment; through effective region clipping, tooth surface unfolding, and overall and local alignment processing, the current page group and the reference page group obtain directly comparable aligned unfolded pages and contact band baselines under unified unfolding coordinates, thereby improving the spatial consistency of tooth surface comparison; through dynamic illumination registration, local undulation reconstruction, curvature difference localization, and directional streak sorting, surface morphology changes and streak evolution within the contact band are jointly characterized, which helps to improve the targeting of wear candidate region extraction; through forward mapping, paired verification, reverse projection, and left and right adjacent tooth contact band migration comparison, the target tooth surface anomaly, paired tooth surface disturbance, and adjacent tooth expansion state are connected, which can more reliably distinguish between real wear and isolated anomalies; through tooth number sequence comparison and wear level determination, a continuous and structured evaluation result for online monitoring scenarios is formed, thereby improving the accuracy and consistency of gear wear state identification.

[0078] It should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit it. Although the present invention has been described in detail with reference to preferred embodiments, those skilled in the art should understand that modifications or equivalent substitutions can be made to the technical solutions of the present invention without departing from the spirit and scope of the technical solutions of the present invention, and all such modifications or substitutions should be covered within the scope of the claims of the present invention.

Claims

1. A method for online monitoring and evaluation of gear wear condition using images, characterized in that: include, The target tooth surface is subjected to zoned illumination synchronous trigger acquisition, and tooth number index relationship and conjugate meshing relationship are established according to the synchronous triggering time sequence. The left oblique illumination image, the front illumination image and the right oblique illumination image are merged according to the tooth number index to generate the current page group and the reference page group. The current page group and the base page group are called to extract the tooth tip edge line, tooth root edge line, left and right tooth profile edge lines and tooth surface stable texture band, and perform effective area clipping, tooth surface expansion and two-level alignment to generate the aligned expansion page and contact band baseline; Dynamic illumination registration, local undulation reconstruction, curvature difference localization, and directional stre sorting are performed on the aligned unfolded pages and contact zone baselines to generate curvature abrupt change zones, stre-associated pages, and wear candidate pages. Based on the tooth number index relationship and conjugate meshing relationship, perform forward mapping, pairing verification and reverse resubmission on the wear candidate pages to determine the effective wear area, and perform contact band migration comparison on the left and right adjacent tooth expansion pages corresponding to the effective wear area to generate bidirectional verification pages and expansion pages; The system collects data from curvature abrupt change zones, streak-related pages, bidirectional return pages, and extended pages, performs tooth number sequence comparison and wear level determination, and generates online assessment results of gear wear status.

2. The online monitoring and evaluation method for gear wear condition images as described in claim 1, characterized in that: The establishment of tooth number index relationships and conjugate meshing relationships specifically includes: Acquire the pulse sequence of the corner encoder and divide the single-tooth visible area to generate a synchronous trigger timing sequence; Based on the tooth sequence position of each single tooth's visible interval according to the synchronous trigger timing mark, a tooth number index relationship is generated; Determine the mating tooth surface that meshes with the target tooth surface according to the tooth number index relationship, and match the corresponding tooth sequence position of the mating tooth surface; The meshing order of the target tooth surface and the paired tooth surface is arranged according to the synchronous triggering sequence to generate a conjugate meshing relationship.

3. The online monitoring and evaluation method for gear wear condition images as described in claim 2, characterized in that: The generation of the current page group and the base page group specifically includes: Within the same single-tooth visible range, continuously acquire left oblique illumination image, front illumination image and right oblique illumination image at the current acquisition time to generate the current single-tooth image group; Within the same single-tooth visible range, continuously acquire left oblique illumination image, front illumination image and right oblique illumination image at the reference acquisition time to generate a reference single-tooth image group; The current single-tooth image group and the reference single-tooth image group are arranged in order of tooth sequence position according to the tooth number index relationship, and the current page group and the reference page group are generated.

4. The online monitoring and evaluation method for gear wear condition images as described in claim 3, characterized in that: The generation of the aligned unfolded page and contact strip baseline specifically includes: Extract the tooth tip edge line, tooth root edge line, left and right tooth profile edge lines, and tooth surface stable texture band from the current page group and the base page group; The tooth surface is enclosed by the tooth tip edge, tooth root edge, and left and right tooth profile edges, and the working area of ​​the tooth surface is defined by the tooth surface stable texture band, thus generating an effective area boundary. Extract the tooth surface image based on the effective region boundary and generate a cropped page; The cut page is unfolded along the tooth profile direction and the tooth width direction to generate a tooth surface unfolded page, which includes the current unfolded page and the reference unfolded page; Using the baseline expanded page as the positioning page, perform overall and partial alignment on the current expanded page to generate an aligned expanded page; Extract stable contact traces from the aligned unfolded page to generate the contact band baseline.

5. The online monitoring and evaluation method for gear wear condition images as described in claim 4, characterized in that: The generation of curvature abrupt change bands, streak-related pages, and wear candidate pages specifically includes: Perform position correction and brightness correspondence on the partitioned lighting images in the alignment expansion page to generate a lighting registration page; Call the lighting registration page to restore the local undulations of the tooth surface and generate the current curvature page; Retrieve the reference curvature page corresponding to the reference page group and perform difference localization to generate curvature abrupt change bands; In the region where the curvature abrupt change zone coincides with the contact zone baseline, the stre direction, stre connected segments, and stre bifurcation segments are extracted to generate stre-associated pages; Abnormal regions in curvature change bands and streak-related pages are aggregated to generate wear candidate pages.

6. The online monitoring and evaluation method for gear wear condition images as described in claim 5, characterized in that: The determination of the effective wear zone specifically includes: The tooth number index relationship is used to locate the candidate abnormal area in the wear candidate page, and the corresponding expansion area of ​​the paired tooth surface is determined according to the conjugate meshing relationship; The candidate anomaly region is projected onto the corresponding expanded region and contact band interruption, contact band narrowing, local offset and dual streak are extracted to generate a paired verification region. According to the conjugate meshing relationship, the paired back-affirmation area is back-projected onto the alignment unfolding page of the target tooth surface to generate the reverse back-projection area; Compare the overlap range between the candidate abnormal area and the reverse return area, and filter out the common area that falls into the wear candidate page at the same time to generate the effective wear area.

7. The online monitoring and evaluation method for gear wear condition images as described in claim 6, characterized in that: The generation of the bidirectional verification page and the extended page specifically includes: Retrieve the unfolded pages of the left and right adjacent teeth corresponding to the tooth number of the effective wear area, and locate the contact band segment on the same side as the effective wear area in the unfolded pages of the left and right adjacent teeth to generate the adjacent tooth contact segment; The paired return verification area, reverse return area, effective wear area and adjacent tooth contact segment are spliced ​​together according to tooth sequence to generate adjacent tooth comparison page; In the adjacent tooth comparison page, the contact band direction, contact band width and stripe connection status are compared, and the migration segments extending from the effective wear area to the left and right adjacent teeth are extracted to generate an extended sequence; The effective wear area, the paired return verification area, and the reverse return area are superimposed on the alignment unfolding page of the target tooth surface to generate a bidirectional return verification page; Arrange the bidirectional return pages and extended sequences accordingly to generate extended pages.

8. The online monitoring and evaluation method for gear wear condition images as described in claim 7, characterized in that: The comparison of the tooth number sequence specifically includes: Arrange the curvature abrupt change zone, the streak-related page, the bidirectional return page, and the extended page according to the tooth number index relationship, and generate the current tooth number sequence page; Retrieve the historical tooth number sequence page corresponding to the previous online detection and the baseline tooth number sequence page corresponding to the baseline page group, and generate a comparison page group; Compare the contact band occupancy, stripe continuity, and return range in the current tooth number sequence page, historical tooth number sequence pages, and reference tooth number sequence pages one by one to generate a sequence difference page.

9. The online monitoring and evaluation method for gear wear condition images as described in claim 8, characterized in that: The wear level determination specifically includes: The abnormal distributions in the sequence difference page, the bidirectional return confirmation page, and the extended page are collected to generate the wear determination page; Based on the location of the abnormal distribution in the contact zone, the corresponding disturbance of the paired tooth surface, and the continuous migration of the left and right adjacent teeth in the wear judgment page, mild judgment results, moderate judgment results, and severe judgment results are generated and summarized to obtain the online evaluation results of gear wear status.

10. A gear wear condition image online monitoring and evaluation system, based on the gear wear condition image online monitoring and evaluation method according to any one of claims 1 to 9, characterized in that: include, The relationship establishment module performs zoned illumination synchronous trigger acquisition on the target tooth surface, and establishes tooth number index relationship and conjugate meshing relationship according to the synchronous triggering sequence. It merges the left oblique illumination image, the forward illumination image and the right oblique illumination image according to the tooth number index to generate the current page group and the reference page group. The alignment module is expanded by calling the current page group and the base page group to extract the tooth tip edge line, tooth root edge line, left and right tooth profile edge lines and tooth surface stable texture band, and performing effective area clipping, tooth surface expansion and two-level alignment to generate the aligned expanded page and contact band baseline; The candidate extraction module performs dynamic illumination registration, local undulation reconstruction, curvature difference localization, and directional stre sorting on the aligned unfolded page and contact zone baseline to generate curvature change zone, stre-associated page, and wear candidate page; The return certificate extension module performs forward mapping, pairing return certificates, and reverse return on the wear candidate pages according to the tooth number index relationship and conjugate meshing relationship, determines the effective wear area, and performs contact band migration comparison on the left and right adjacent tooth expansion pages corresponding to the effective wear area to generate bidirectional return certificate pages and expansion pages; The condition assessment module combines the curvature change zone, the scratch-related page, the bidirectional return page, and the extended page to perform tooth number sequence comparison and wear level determination, and generates online assessment results of gear wear condition.

Images