Intelligent number-based semiconductor device component whole-process tracking method and system

Abstract

The application discloses a semiconductor equipment part whole-process tracing method and system based on intelligent numbering, and relates to the technical field of part tracing.The application solves the problem of lacking correlation and integration of key data such as physical characteristics and image information of parts, and makes it difficult to quickly locate the root cause when an abnormality occurs.Through fusion of three core information of equipment number, surface feature point position and coding area space relationship, the application extracts part gradient contour feature points through gray processing and Sobel algorithm, and forms a surface feature reference in combination with a feature angle and a center point positioning.This double binding mode of 'coding + physical characteristics' avoids the identity confusion problem caused by the easy wear and easy tampering of traditional coding.Even when the surface coding of the part is slightly damaged, the identity can still be confirmed through gradient contour and feature point comparison, ensuring that the 'unique identification' of each part from the warehouse to the whole process of circulation is not invalid, laying a foundation for the accuracy of the tracing data.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of component traceability, in particular to a semiconductor device component whole-process traceability method and system based on intelligent numbering. BACKGROUND

[0002] In the field of semiconductor manufacturing, the quality and traceability of device components are directly related to the reliability and production yield of terminal chips.

[0003] Traditional semiconductor component traceability relies on a single barcode or serial number, which has significant technical bottlenecks:

[0004] On the one hand, semiconductor components vary greatly in size (from millimeter-level chips to meter-level substrates), and production involves multiple scenarios such as procurement, processing, testing, warehousing, and transportation. A single code is prone to failure due to wear, tampering, or environmental interference;

[0005] On the other hand, traditional traceability focuses on "node" data recording, with data scattered across different systems. There is a lack of correlation and integration of key data such as physical characteristics and image information, making it difficult to quickly locate the root cause of abnormalities, and even causing "traceability gaps".

[0006] In addition, semiconductor manufacturing requires high precision, and traditional qualitative traceability methods cannot provide the quantitative data support needed for process optimization, limiting the level of quality control refinement.

[0007] Under this background, the semiconductor device component whole-process traceability method and system based on intelligent numbering proposed in this application significantly improves the accuracy, timeliness, and reliability of semiconductor component traceability through innovative comprehensive feature number generation logic and whole-process verification mechanism, providing key technical support for quality control and risk prevention in the entire semiconductor manufacturing chain. SUMMARY

[0008] To address the shortcomings of the prior art, the present application provides a semiconductor device component whole-process traceability method and system based on intelligent numbering, which solves the problem of lack of correlation and integration of key data such as physical characteristics and image information, making it difficult to quickly locate the root cause of abnormalities.

[0009] To achieve the above purpose, the present application is implemented by the following technical scheme: a semiconductor device component whole-process traceability method based on intelligent numbering, comprising the following steps:

[0010] Step one, obtain the out-of-warehouse image of each group of semiconductor device components, and determine the associated device number in the out-of-warehouse image. Combine the surface characteristics of the components to confirm the surface comprehensive feature number of the components, specifically:

[0011] Confirming the delivery image of the semiconductor device parts, and carrying out gray processing on the delivery image, confirming the gray image associated with the delivery image: confirming the RGB value associated with different image points in the delivery image, and confirming the gray value associated with the corresponding image point by using the method of gray value = 0.114 * R + 0.587 * G + 0.299 * B, and generating the gray image associated with the delivery image according to different gray values associated with different image points;

[0012] And confirming the edge profile existing in the gray image by using the Sobel algorithm, confirming the vertical gradient and horizontal gradient associated with different gray points, and confirming the comprehensive gradient associated with the corresponding gray point, wherein ;

[0013] Recording the pixel point satisfying the comprehensive gradient >= Y1 as a gradient point, connecting the adjacent gradient points to confirm the gradient profile associated with the part gray image according to the gradient points confirmed step by step, and confirming the feature points on the gradient profile, the specific method comprising:

[0014] Recording the adjacent three groups of gradient points as undetermined points, confirming the included angle J i generated by the three groups of undetermined points, wherein i represents different continuous undetermined points, wherein Y1 is a preset value, and the included angle satisfying J i <= 90° is recorded as a feature angle, the intermediate undetermined point associated with the feature angle is recorded as a feature point, and a plurality of feature points are connected to confirm a feature polygon, and the center point of the feature polygon is confirmed in combination with the two-dimensional coordinate system, and the confirmed center point is recorded as the surface feature of the part;

[0015] Wherein, the specific method for confirming the surface comprehensive feature number of the part is:

[0016] Confirming the area where the part surface code is located in the delivery image, and marking the center point of the area, confirming the straight line distance between the marked center point and the surface feature;

[0017] Then directly extracting the part code from the area where the part surface code is located, placing the straight line distance after the part code, and generating the comprehensive feature number belonging to the current part;

[0018] Step 2, when the semiconductor device parts enter the next process, reacquire the surface image of the semiconductor device parts, and again confirm the surface comprehensive feature number, identify whether the confirmed surface comprehensive feature number is consistent with the recorded surface comprehensive feature number, if consistent, no need to do any processing, if not consistent, record this device part as an abnormal process part;

[0019] Step three, for the confirmed abnormal process component, extract the process information associated with this component, and from the process information, extract the surface feature number recorded each time, from the recorded feature number, confirm the inconsistent process, and extract the surface image, evaluate the abnormal information of the abnormal process component and display, the specific way is:

[0020] Confirm the surface feature number associated with the abnormal process component in the outbound image and record it as the main number. Then, from the process information associated with the abnormal process component, sort the confirmed surface feature numbers in chronological order, confirm the number sequence, and confirm the first surface feature number that is inconsistent with the main number from front to back in the number sequence. Record the process information associated with this surface feature number as pending information, and confirm the associated image from the pending information;

[0021] Confirm the gradient profile marked on the associated image and record it as the secondary profile. Confirm the gradient profile in the outbound image associated with the main number and record it as the main profile. Make the main profile and the secondary profile overlap and verify. Confirm the abnormal information:

[0022] If the main profile and the secondary profile do not completely overlap, and the part surface coding area completely overlaps, generate a part deformation signal for display;

[0023] If the main profile and the secondary profile do not completely overlap, and the part surface coding area also does not completely overlap, generate a multiple abnormal signal for display, which includes the part deformation signal and the surface coding replacement signal;

[0024] If the main profile and the secondary profile completely overlap, and the part surface coding area does not completely overlap, generate a surface coding replacement signal;

[0025] The specific way of making the main profile and the secondary profile overlap and verify is:

[0026] Make the main profile and the secondary profile in the same plane, mark the feature points in the main profile as main feature points, and mark the feature points in the secondary profile as secondary feature points;

[0027] Compare and process the main profile and the secondary profile to make the main feature points and the secondary feature points overlap. Execute several overlap comparison processes, and record the number of main feature points and secondary feature points associated with each overlap comparison process as process characteristics of the corresponding overlap comparison process. When the continuous pending points associated with the main feature points and the secondary feature points overlap, record it as the overlap state;

[0028] Select the group of overlap comparison processes with the largest process characteristic value as the determined process, and confirm the overlap state of the main profile and the secondary profile in the determined process. Confirm the overlap state of the surface coding area in the two profiles.

[0029] If the process features are all 0, then the maximum inner diameter line segment is confirmed from the primary profile and the secondary profile, and a set of external circles is generated by taking the maximum inner diameter line segment as the diameter, so that the primary profile and the secondary profile move in the external circles, and in the moving process, the maximum coincidence ratio is confirmed, the coincidence process associated with the maximum coincidence ratio is recorded as the determined process, and the subsequent coincidence state is evaluated.

[0030] Preferably, the coincidence ratio confirmation process comprises:

[0031] First, the proportion Z1 of the area of the primary profile overlap region in the total area of the primary profile is confirmed, then the proportion Z2 of the area of the secondary profile overlap region in the total area of the secondary profile is confirmed, and the mean value of Z1 and Z2 is confirmed, and the confirmed mean value is recorded as the coincidence ratio.

[0032] Preferably, the semiconductor equipment component full-process tracing system based on intelligent numbering comprises:

[0033] The feature processing end obtains the out-of-warehouse image of each group of semiconductor equipment components, determines the equipment number associated in the out-of-warehouse image, combines the surface features of the components, confirms the surface comprehensive feature number of the components, and binds the confirmed comprehensive feature number with the corresponding semiconductor equipment components.

[0034] The component number verification end re-obtains the surface image of the semiconductor equipment component when the component enters the next process, and again confirms the surface comprehensive feature number, identifies whether the confirmed surface comprehensive feature number is consistent with the recorded surface comprehensive feature number, and if consistent, no further processing is required, and if not consistent, the component is recorded as an abnormal process component.

[0035] The abnormal information confirmation end extracts the process information associated with the confirmed abnormal process component, extracts the surface comprehensive feature number recorded in each process information from the process information, confirms the inconsistent processes from the recorded comprehensive feature number, extracts the surface image from the inconsistent processes, evaluates the abnormal information of the abnormal process component, and displays the abnormal information.

[0036] The present application provides a semiconductor equipment component full-process tracing method and system based on intelligent numbering.

[0037] The comprehensive feature number combines the device number, surface feature point, and the spatial relationship of the coding area three core information: the gradient contour feature points of the part are extracted by gray processing and Sobel algorithm, the surface feature reference is formed by combining the feature angle and the center point positioning, and the composite number is generated by associating the straight line distance between the coding area and the feature center point. The dual binding mode of "coding + physical feature" avoids the identity confusion problem caused by the easy wear and easy tampering of traditional coding. Even when the surface coding of the part is slightly damaged, the identity can still be confirmed by comparing the gradient contour and the feature point, ensuring that the "unique identification" of each part from the warehouse to the whole process of circulation is not invalid, laying a foundation for the accuracy of the traceability data.

[0038] The surface image acquisition and comprehensive feature number comparison are automatically triggered when the part enters each process node. Compared with the traditional after-inspection mode, the abnormality is realized in real time and intercepted. When the number is inconsistent, the first abnormal process node can be quickly located through time sequence sorting, and the different fault types such as "part deformation", "coding replacement" or "multiple abnormalities" can be accurately distinguished by means of the coincidence verification of the main contour and the secondary contour (including feature point coincidence degree, maximum inner diameter line segment comparison and other multi-dimensional judgment).

[0039] Through the comprehensive feature number, the multi-source data such as warehouse image, process node image, gradient contour feature and coding information are connected to form a "image + feature + coding" three-dimensional traceability file of the whole life cycle of the part. In the after-sales operation or quality review scene, the image and feature data of each process node can be quickly retrieved through the comprehensive feature number to clearly restore the circulation path and state change of the part. Especially in the second and third embodiments, the physical features are quantified into comparable data by means of gray processing and feature polygon fitting algorithm, so that the traceability is upgraded from "qualitative description" to "quantitative analysis", providing data support for the design optimization and process improvement of semiconductor equipment parts, and helping enterprises to realize "data-driven quality control". BRIEF DESCRIPTION OF DRAWINGS

[0040] Figure 1 The figure is a schematic diagram of the method of the present application. DETAILED DESCRIPTION

[0041] The technical solutions in the embodiments of the present application will be described clearly and completely in combination with the drawings in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, not all the embodiments. Based on the embodiments in the present application, all other embodiments obtained by those skilled in the art without creative labor belong to the scope of protection of the present application.

[0042] First embodiment

[0043] Please refer to Figure 1The application provides a semiconductor equipment part whole-process tracing method based on intelligent numbering, comprising the following steps:

[0044] Step one, obtain the delivery image of each group of semiconductor equipment parts, determine the associated equipment number in the delivery image, combine the surface features of the parts, confirm the surface comprehensive feature number of the parts, bind the confirmed comprehensive feature number with the corresponding semiconductor equipment parts, specifically, the comprehensive feature number here not only includes the surface features of the parts, but also includes the corresponding equipment number, according to the surface image of the corresponding parts, confirm the feature corner points existing on the surface, and then lock the built-in midpoint according to the comprehensive performance of the feature corner points, so as to confirm the overall comprehensive feature number of the corresponding parts by combining the feature points in the corresponding number area;

[0045] Step two, when the semiconductor equipment parts enter the next process, re-obtain the surface image of the semiconductor equipment parts, and confirm the surface comprehensive feature number again, identify whether the confirmed surface comprehensive feature number is consistent with the recorded surface comprehensive feature number, if yes, no further treatment is needed, if not, the equipment part is recorded as an abnormal process part, specifically, the corresponding semiconductor equipment part is associated with the corresponding processing process, each process is associated with different process images, but the feature numbers of each process image are consistent, if there is inconsistency, it means that the corresponding part has an abnormality in the transportation process, so subsequent traceability verification treatment is needed to identify whether the corresponding part has missed related processes;

[0046] Step three, for the confirmed abnormal process part, extract the process information associated with the part, and extract the surface comprehensive feature number recorded by each process information from the process information, confirm the inconsistent processes from the recorded comprehensive feature number, and extract the surface image from the inconsistent processes, evaluate and display the abnormal information of the abnormal process part, specifically, there is an entry image in each process during the transmission determination of the corresponding part, the feature number confirmation process is needed in the entry process, when the feature number of the corresponding process is abnormal, abnormal identification is needed to determine and display the abnormal information.

[0047] The semiconductor equipment part whole-process tracing system based on intelligent numbering comprises:

[0048] The feature processing end obtains the delivery image of each group of semiconductor equipment parts, determines the associated equipment number in the delivery image, combines the surface features of the parts, confirms the surface comprehensive feature number of the parts, and binds the confirmed comprehensive feature number with the corresponding semiconductor equipment parts;

[0049] The component number verification end re-acquires the surface image of the semiconductor equipment component when it enters the next process, and reconfirms the surface comprehensive feature number. It identifies whether the confirmed surface comprehensive feature number is consistent with the recorded surface comprehensive feature number. If they are consistent, no processing is required. If they are inconsistent, the component is recorded as an abnormal process component.

[0050] The anomaly information confirmation end extracts the process information associated with the confirmed abnormal process component, extracts the surface comprehensive feature number recorded for each process information from the process information, confirms inconsistent processes from the recorded comprehensive feature numbers, extracts surface images from them, evaluates and displays the anomaly information of the abnormal process component.

[0051] Second Embodiment

[0052] As a further embodiment of Embodiment 1, this embodiment, in its specific implementation process, mainly focuses on the confirmation process of the comprehensive feature number of the corresponding component compared to the first embodiment;

[0053] The specific method for confirming the surface features of components is as follows:

[0054] Confirm the outbound images of semiconductor equipment components and perform grayscale processing on the outbound images. Confirm the grayscale images associated with the outbound images: confirm the RGB values ​​associated with different image points in the outbound images, and use the following method: grayscale value = 0.114×R + 0.587×G + 0.299×B to confirm the grayscale values ​​associated with the corresponding image points. Based on the different grayscale values ​​associated with different image points, generate the grayscale images associated with the outbound images.

[0055] The Sobel algorithm was used to confirm the edge contours within the grayscale image, and the vertical and horizontal gradients associated with different grayscale points were identified. The overall gradient associated with each grayscale point was then determined. ;

[0056] Pixels that satisfy the comprehensive gradient ≥ Y1 are recorded as gradient points. Based on the gradient points confirmed step by step, adjacent gradient points are connected to confirm the gradient contour associated with the grayscale image of the component, and feature points are confirmed on the gradient contour. Three adjacent sets of gradient points are recorded as undetermined points, and the built-in angle J generated by the three sets of undetermined points is confirmed. i The included angle determines the direction as the internal direction of the gradient profile, where i represents different continuous undetermined points. If , where Y1 is a preset value, its specific value is determined by the operator based on experience, and J satisfies iAn internal angle ≤90° is denoted as a feature angle. The intermediate undetermined point associated with the feature angle is denoted as a feature point. Several feature points are connected to confirm a set of feature polygons. Then, combined with a two-dimensional coordinate system, the center point of the feature polygon is confirmed. The confirmed center point is denoted as the surface feature of the part. Specifically, when the feature polygon is placed inside a set of two-dimensional coordinate systems, each point on the line connecting the feature polygons is associated with a set of two-dimensional coordinates. Then, the average value of several sets of two-dimensional coordinates is processed to confirm a set of average coordinates. The point where the average coordinates are located is the corresponding center point. Therefore, the confirmed center point is the surface feature of the corresponding part.

[0057] The specific method for confirming the comprehensive feature number of the component surface is as follows:

[0058] Within the outbound image, identify the area where the component surface code is located, and mark the center point of the area to confirm the straight-line distance between the marked center point and the surface feature.

[0059] Then, extract the component code directly from the area where the component surface code is located, place the straight-line distance after the component code, and generate a comprehensive feature number belonging to the current component. For example, if the component code is date + cumulative serial number, such as 251005-0001, and the confirmed straight-line distance is 20mm, then the generated comprehensive feature number is: 251005-0001-20.

[0060] Third Embodiment

[0061] This embodiment is a further embodiment of embodiment one, mainly for the confirmation process of whether there is missing information in the abnormal process component, and compares and verifies the process information associated with the abnormal process component.

[0062] The specific method for comparing and verifying process information is as follows:

[0063] Identify the surface comprehensive feature number associated with the abnormal process component in the outbound image and record it as the main number. Then, sort the surface comprehensive feature numbers confirmed each time in the process information associated with the abnormal process component according to the time sequence, confirm the number sequence, and confirm the first surface comprehensive feature number that does not match the main number from the beginning to the end in the number sequence. Record the process information associated with this surface comprehensive feature number as pending information, and confirm the associated image from the pending information.

[0064] The gradient contours marked on the associated images are confirmed and recorded as secondary contours. The gradient contours in the outbound images associated with the main image number are also confirmed and recorded as primary contours. The primary and secondary contours are then checked for overlap to identify any anomalies.

[0065] If the primary profile and the secondary profile do not completely coincide, and the surface coding area of the part completely coincides, a part deformation signal is generated for display;

[0066] If the primary profile and the secondary profile do not completely coincide, and the surface coding area of the part also does not completely coincide, a multiple abnormal signal is generated for display, including the part deformation signal and the surface coding replacement signal;

[0067] If the primary profile and the secondary profile completely coincide, and the surface coding area of the part does not completely coincide, a surface coding replacement signal is generated;

[0068] The specific way of making the primary profile and the secondary profile coincide is as follows:

[0069] The primary profile and the secondary profile are placed in the same plane, the feature points marked in the primary profile are recorded as primary feature points, and the feature points marked in the secondary profile are recorded as secondary feature points;

[0070] The primary profile and the secondary profile are compared and processed to make the primary feature points and the secondary feature points coincide, a plurality of coincidence comparison processes are performed, and the number of coincidences of the primary feature points and the secondary feature points associated with each coincidence comparison process is recorded as the process characteristics of the corresponding coincidence comparison process. When the continuous undetermined points associated with the primary feature points and the secondary feature points coincide, it is recorded as a coincidence state;

[0071] The coincidence comparison process with the largest process characteristic value is selected as the determination process, and the coincidence state of the primary profile and the secondary profile is confirmed in the determination process, and the coincidence state of the surface coding area in the two profiles is confirmed;

[0072] If the process characteristics are all 0, the largest inner diameter line segment (that is, the longest line segment inside, with both end points located on the corresponding profile) is confirmed from the primary profile and the secondary profile, and a set of external circles is generated with the largest inner diameter line segment as the diameter. The primary profile and the secondary profile move within the external circles, and in the moving process, the maximum coincidence ratio is confirmed, the coincidence process associated with the maximum coincidence ratio is recorded as the determination process, and the subsequent coincidence state evaluation is performed.

[0073] The so-called maximum coincidence ratio is to first confirm the proportion Z1 of the primary profile coincidence area in the total area of the primary profile, then confirm the proportion Z2 of the secondary profile coincidence area in the total area of the secondary profile, and then confirm the average of Z1 and Z2 as the coincidence ratio. Different moving processes are associated with different coincidence ratios, so the maximum coincidence ratio can be selected to confirm the coincidence state.

[0074] Some data in the above formula are dimensionless for numerical calculation, and the contents not described in detail in the specification are all prior art known by those skilled in the art.

[0075] The above examples are only used to illustrate the technical method of the present application but not limit the present application. Although the present application is described in detail with reference to the preferred embodiments, those skilled in the art should understand that the technical method of the present application can be modified or replaced equivalently without departing from the spirit and scope of the technical method of the present application.

Claims

1. A method for end-to-end traceability of semiconductor equipment components based on intelligent numbering, characterized in that: Includes the following steps: Step 1: Obtain the outbound image of each group of semiconductor equipment components and determine the equipment number associated with the outbound image. Combine this with the surface features of the components to confirm the comprehensive surface feature number of the components. The specific method is as follows: Within the outbound image, identify the area where the component surface code is located, and mark the center point of the area to confirm the straight-line distance between the marked center point and the surface feature. Then, extract the component code directly from the area where the component surface code is located, place the straight-line distance after the component code, and generate the surface comprehensive feature number belonging to the current component. Step 2: When the semiconductor equipment component enters the next process, re-acquire the surface image of the semiconductor equipment component and reconfirm the surface integrated feature number. Identify whether the confirmed surface integrated feature number is consistent with the recorded surface integrated feature number. If they are consistent, no processing is required. If they are inconsistent, record this equipment component as an abnormal process component. Step 3: For the identified abnormal process components, extract the process information associated with these components, and extract the surface composite feature numbers recorded for each process step from the process information. From the recorded surface composite feature numbers, identify inconsistent processes, extract surface images from them, assess the abnormal information of the abnormal process components, and display it. The specific method is as follows: Identify the surface comprehensive feature number associated with the abnormal process component in the outbound image and record it as the main number. Then, sort the surface comprehensive feature numbers confirmed each time in the process information associated with the abnormal process component according to the time sequence, confirm the number sequence, and confirm the first surface comprehensive feature number that does not match the main number from the beginning to the end in the number sequence. Record the process information associated with this surface comprehensive feature number as pending information, and confirm the associated image from the pending information. The gradient contours marked on the associated images are confirmed and recorded as secondary contours. The gradient contours in the outbound images associated with the main image number are also confirmed and recorded as primary contours. The primary and secondary contours are then checked for overlap to identify any anomalies. If the main contour and the secondary contour do not completely overlap, but the coding area on the surface of the component completely overlaps, then a component deformation signal is generated for display. If the main contour and the secondary contour do not completely overlap, and the surface coding area of ​​the component does not completely overlap, multiple abnormal signals will be generated for display. The multiple abnormal signals include component deformation signals and surface coding replacement signals. If the main contour and the secondary contour completely overlap, but the surface coding area of ​​the component does not completely overlap, a surface coding replacement signal is generated.

2. The method for end-to-end traceability of semiconductor equipment components based on intelligent numbering according to claim 1, characterized in that, In step one, the specific method for confirming the surface features of the parts is as follows: Confirm the outbound images of semiconductor equipment components and perform grayscale processing on the outbound images. Confirm the grayscale images associated with the outbound images: confirm the RGB values ​​associated with different image points in the outbound images, and use the following method: grayscale value = 0.114×R + 0.587×G + 0.299×B to confirm the grayscale values ​​associated with the corresponding image points. Based on the different grayscale values ​​associated with different image points, generate the grayscale images associated with the outbound images. The Sobel algorithm was used to confirm the edge contours within the grayscale image, and the vertical and horizontal gradients associated with different grayscale points were identified. The overall gradient associated with each grayscale point was then determined. ; Pixels that satisfy the comprehensive gradient ≥ Y1 are recorded as gradient points. Based on the gradient points confirmed step by step, adjacent gradient points are connected to confirm the gradient contour associated with the grayscale image of the component, and feature points are confirmed on the gradient contour.

3. The method for end-to-end traceability of semiconductor equipment components based on intelligent numbering according to claim 2, characterized in that, The confirmation process for the feature points includes: Record three adjacent sets of gradient points as points to be determined, and confirm the built-in angle J generated by the three sets of points to be determined. i The included angle determines the direction as the internal direction of the gradient profile, where i represents different continuous undetermined points. If, where Y1 is a preset value, and J satisfies... i An internal angle ≤90° is recorded as a feature angle. The intermediate undetermined point associated with the feature angle is recorded as a feature point. Several feature points are connected to confirm a set of feature polygons. Then, in combination with a two-dimensional coordinate system, the center point of the feature polygon is confirmed. The confirmed center point is recorded as the surface feature of the component.

4. The method for end-to-end traceability of semiconductor equipment components based on intelligent numbering according to claim 1, characterized in that, In step three, the specific method for verifying the overlap between the main contour and the secondary contour is as follows: Place the main contour and the secondary contour in the same plane, and record the feature points marked in the main contour as the main feature points, and the feature points marked in the secondary contour as the secondary feature points. The main contour and the secondary contour are compared and overlapped to make the main feature points and the secondary feature points overlap. Several comparison and overlap processes are executed, and the number of overlaps between the main feature points and the secondary feature points associated with each comparison and overlap process is recorded as the process feature of the corresponding comparison and overlap process. When the continuous undetermined points associated with the main feature points and the secondary feature points all overlap, it is recorded as an overlap state. Select the process with the largest overlap ratio and record it as the determined process. Within the determined process, confirm the overlap status of the main contour and the secondary contour, and confirm the overlap status of the surface coding areas of the two contours. If all process features are 0, then the maximum inner diameter line segment is identified from the main contour and the secondary contour, and a set of external circles is generated using the maximum inner diameter line segment as the diameter. The main contour and the secondary contour move within the external circles, and the maximum overlap ratio is identified during the movement process. The overlap process associated with the maximum overlap ratio is recorded as the determination process, and the subsequent overlap state is evaluated.

5. The method for end-to-end traceability of semiconductor equipment components based on intelligent numbering according to claim 4, characterized in that, The process for confirming the overlap ratio includes: First, confirm the proportion Z1 of the area of ​​the overlapping region of the main contour to the total area of ​​the main contour. Then, confirm the proportion Z2 of the area of ​​the overlapping region of the secondary contour to the total area of ​​the secondary contour. Then, calculate the average proportion of Z1 and Z2 and record the average proportion as the overlap ratio. Different movement processes are associated with different overlap ratios.

6. A semiconductor equipment component end-to-end traceability system based on intelligent numbering, wherein the system operates according to the semiconductor equipment component end-to-end traceability method based on intelligent numbering as described in any one of claims 1-5, characterized in that, include: The feature processing end acquires the outbound image of each group of semiconductor equipment components, determines the equipment number associated with the outbound image, and confirms the comprehensive surface feature number of the component by combining the surface features of the component. The confirmed comprehensive surface feature number is then bound to the corresponding semiconductor equipment component. The component number verification end re-acquires the surface image of the semiconductor equipment component when it enters the next process, and reconfirms the surface comprehensive feature number. It identifies whether the confirmed surface comprehensive feature number is consistent with the recorded surface comprehensive feature number. If they are consistent, no processing is required. If they are inconsistent, the component is recorded as an abnormal process component. The abnormal information confirmation end extracts the process information associated with the confirmed abnormal process component, extracts the surface comprehensive feature number recorded for each process information from the process information, confirms inconsistent processes from the recorded surface comprehensive feature number, extracts the surface image from it, evaluates the abnormal information of the abnormal process component and displays it.

Images