CN112446857A - Automatic classification marking system, establishing method and classification marking method for defect images - Google Patents
Automatic classification marking system, establishing method and classification marking method for defect images Download PDFInfo
- Publication number
- CN112446857A CN112446857A CN202011229553.6A CN202011229553A CN112446857A CN 112446857 A CN112446857 A CN 112446857A CN 202011229553 A CN202011229553 A CN 202011229553A CN 112446857 A CN112446857 A CN 112446857A
- Authority
- CN
- China
- Prior art keywords
- defect
- image
- defect image
- module
- marking
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/0002—Inspection of images, e.g. flaw detection
- G06T7/0004—Industrial image inspection
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/22—Matching criteria, e.g. proximity measures
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/70—Arrangements for image or video recognition or understanding using pattern recognition or machine learning
- G06V10/74—Image or video pattern matching; Proximity measures in feature spaces
- G06V10/75—Organisation of the matching processes, e.g. simultaneous or sequential comparisons of image or video features; Coarse-fine approaches, e.g. multi-scale approaches; using context analysis; Selection of dictionaries
- G06V10/751—Comparing pixel values or logical combinations thereof, or feature values having positional relevance, e.g. template matching
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/30—Subject of image; Context of image processing
- G06T2207/30108—Industrial image inspection
- G06T2207/30148—Semiconductor; IC; Wafer
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Computer Vision & Pattern Recognition (AREA)
- General Physics & Mathematics (AREA)
- Physics & Mathematics (AREA)
- Evolutionary Computation (AREA)
- Data Mining & Analysis (AREA)
- Artificial Intelligence (AREA)
- General Engineering & Computer Science (AREA)
- Databases & Information Systems (AREA)
- Bioinformatics & Computational Biology (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Life Sciences & Earth Sciences (AREA)
- Health & Medical Sciences (AREA)
- Computing Systems (AREA)
- Evolutionary Biology (AREA)
- General Health & Medical Sciences (AREA)
- Medical Informatics (AREA)
- Software Systems (AREA)
- Multimedia (AREA)
- Quality & Reliability (AREA)
- Investigating Materials By The Use Of Optical Means Adapted For Particular Applications (AREA)
Abstract
The invention provides a system for automatically classifying and marking defect images, an establishing method and application. The system comprises a storage module, an image conversion module, an image identification module, a rule identification module and an output module; the storage module is used for storing the defect image; the image conversion module is connected with the storage module and is used for carrying out format conversion on the stored defect image; the pattern recognition module is connected with the image conversion module and used for performing pattern recognition on the defect image after format conversion so as to perform classification marking according to the shape of the defect image; the rule identification module is connected with the image conversion module and/or the image identification module and is used for classifying and marking the defect images which cannot be classified and marked according to the shapes after format conversion according to rules; the output module is connected with the pattern recognition module and the rule recognition module and used for outputting the classification marking result. The method is beneficial to quickly and accurately finding out the problem root, saves labor power and avoids image classification difference caused by artificial subjective factors.
Description
Technical Field
The invention relates to the field of semiconductor chip production and manufacturing, in particular to defect classification in the chip manufacturing process, and particularly relates to a defect image automatic classification and marking system, an establishing method, a defect image automatic classification and marking method based on the defect image automatic classification and marking system, a control module and a storage medium.
Background
In the semiconductor chip manufacturing process, particles (particles) are generated due to environment, equipment and/or process, and the particles fall on the wafer to form various defect images, thereby reducing the production yield. With the continuous reduction of the characteristic size of the chip and the increasing of the integration level of the device, the defect image has an increasingly large influence on the quality of the device, so that the control of the chip factory on the defect image is increasingly strict, the defect image is quickly and accurately analyzed to find out the defect which really affects the manufacturing process and find out the root cause of the problem, and the method is daily important work of engineers in the semiconductor factory.
Semiconductor chip manufacturers generate a large number of defect images every day, and conventionally, engineers visually check the defect images, classify the defect images, and determine whether the images are abnormal. The manual visual inspection method not only consumes a lot of manpower, but also can generate different classifications for the same image due to artificial subjective factors, so that the problem defect and the root cause thereof are difficult to be quickly and accurately found.
Disclosure of Invention
In view of the above-mentioned shortcomings of the prior art, an object of the present invention is to provide a defect image automatic classification and marking system and an establishing method thereof, a defect image automatic classification and marking method based on the defect image automatic classification and marking system, a control module and a storage medium, which are used to solve the problems that a lot of manpower is consumed, and different classifications may be generated for the same image due to human subjective factors, which makes it difficult to quickly and accurately find out the problem defect and its root cause, etc. in the existing semiconductor factory, the defect image automatic classification and marking method is visually checked by engineers and manually classified.
To achieve the above and other related objects, the present invention provides a system for automatically classifying and marking defect images used in semiconductor chip fabrication, the system comprising:
the storage module is used for storing the defect image;
the image conversion module is connected with the storage module and is used for carrying out format conversion on the stored defect image;
the image conversion module is used for converting the defect image into a format of a defect image, and converting the defect image into a format of the defect image;
the rule identification module is connected with the image conversion module and/or the image identification module and is used for classifying and marking the defect images which cannot be classified and marked according to the shapes of the defect images after format conversion according to rules;
and the output module is connected with the pattern recognition module and the rule recognition module and used for outputting the classification marking result.
Optionally, the image conversion module converts the defect image from a kafka image format to a lattice format.
Optionally, the pattern recognition comprises one or more of an image template matching method, a structural similarity recognition method, a perceptual hashing algorithm and a histogram method.
Optionally, the defect image is divided into a linear shape, a circular shape, a scratch shape, and a divergent rotation shape according to the shape of the defect image.
Optionally, the defect image is classified into a normal shape and a cluster shape according to a rule identification.
More optionally, the rule identifies a distance between defect points constituting a defect image, where the defect points with a distance of <5um in the defect image are denoted as Q1, and when the number of Q1 in the same defect image is greater than 10, the defect image is classified and labeled as a cluster; and (3) recording the defects with the distance of [5um,1cm ] in the defect image as Q2, and when the number of Q2 in the same defect image is more than 30, classifying the defect image as an aggregation shape, and classifying the defect image with a non-aggregation shape as a normal shape.
The invention also provides a method for establishing the automatic defect image classification and marking system in any scheme, which comprises the following steps:
formulating classification types of the defect images, wherein the classification types comprise a type based on pattern recognition and a type based on rule recognition;
carrying out pattern recognition on the defect image, and storing the defect image into a corresponding pattern library according to different shapes;
and performing rule identification on the defect images which are not classified after the pattern identification, and storing the defect images which accord with different identification rules into a corresponding pattern library.
The invention also provides a method for automatically classifying and marking the defect image by using the system for automatically classifying and marking the defect image, which comprises the following steps:
storing the defect images to be classified and marked and performing format conversion;
and performing pattern recognition on the defect image after format conversion so as to perform classification marking according to the shape of the defect image, and if the defect image cannot be subjected to classification marking according to the shape, performing rule recognition on the defect image so as to perform classification marking according to the rule on the defect image which cannot be subjected to classification marking according to the shape.
The invention also provides a control module, which comprises a processor and a memory; the memory is used for storing a computer program; the processor is used for executing the computer program stored in the memory so as to enable the control module to execute the automatic defect image classification and marking method in any scheme.
The invention also provides a storage medium, on which a computer program is stored, which when executed by a processor implements the method for automatically classifying and marking defect images as described in any of the above aspects.
As described above, the present invention provides a system and a method for automatically classifying and marking defective images based on computer implementation, which can rapidly and accurately classify and analyze the defective images, help to rapidly and accurately find the root cause of problems, and not only help to save manpower, but also avoid image classification differences caused by human subjective factors, and help to improve production yield, compared with the conventional method of manually inspecting and classifying by eyes.
Drawings
Fig. 1 is a schematic structural diagram of an automatic defect image classification and marking system according to an embodiment of the present invention.
Fig. 2 and 3 are schematic diagrams showing a defect image.
Fig. 4 is a schematic structural diagram of a control module according to an embodiment of the invention.
Description of the element reference numerals
11 memory module
12 image conversion module
13 pattern recognition module
14 rule identification module
15 output module
21 processor
22 memory
Detailed Description
The embodiments of the present invention are described below with reference to specific embodiments, and other advantages and effects of the present invention will be easily understood by those skilled in the art from the disclosure of the present specification. The invention is capable of other and different embodiments and of being practiced or of being carried out in various ways, and its several details are capable of modification in various respects, all without departing from the spirit and scope of the present invention. It is to be noted that the features in the following embodiments and examples may be combined with each other without conflict.
It should be noted that the drawings provided in the following embodiments are only for illustrating the basic idea of the present invention, and the components related to the present invention are only shown in the drawings rather than drawn according to the number, shape and size of the components in actual implementation, and the type, quantity and proportion of the components in actual implementation may be changed freely, and the layout of the components may be more complicated.
As shown in fig. 1, the present invention provides a defect image automatic classification marking system for semiconductor chip preparation, namely, the defect image automatic classification marking system is used for automatically classifying and marking defect images generated in the semiconductor chip preparation process, and the system includes a storage module 11, an image conversion module 12, a pattern recognition module 13, a rule recognition module 14 and an output module 15; the storage module 11 is used for storing a defect image; the image conversion module 12 is connected to the storage module 11, and is configured to perform format conversion on the stored defect image; the pattern recognition module 13 is connected to the image conversion module 12, and is configured to perform pattern recognition on the defect image after format conversion, so as to perform classification and marking according to the shape of the defect image; the rule identification module 14 is connected to the image conversion module 12 and/or the pattern identification module 13, and is configured to classify and mark, according to rules, a defect image that cannot be classified and marked according to shape after format conversion; the output module 15 is connected to the pattern recognition module 13 and the rule recognition module 14, and is configured to output a classification mark result, and the output module 15 includes one or a combination of a wired output module and a wireless output module. The automatic defect image classification and marking system can be loaded on a computer, so that the automatic defect image classification and marking system can automatically classify and mark defect images, can quickly and accurately classify and analyze the defect images compared with the traditional mode of manual visual inspection and classification, is favorable for quickly and accurately finding out the root cause of problems, is favorable for saving manpower, can avoid image classification difference caused by artificial subjective factors, and is favorable for improving the production yield.
It should be noted that the above modules may be only one functional division rather than a physical division, for example, a single or multiple of the foregoing functional modules may be loaded on the same computer.
In this embodiment, as an example, the image conversion module 12 converts the defect image from the kafka image format to the lattice format, that is, the original defect image may be stored in the kafka image format, but may also be stored in other formats such as png to tif, jpg, etc., and what is important is that the original storage format is suitable for being subsequently converted into a two-dimensional lattice format for computer processing, such as bmp format, etc. In the embodiment, the initial defect image is stored in the kafka image format, which is beneficial to reducing the storage capacity and meeting the real-time processing requirement of large-capacity data. The defect images may be automatically captured by defect capture cameras deployed in a semiconductor fab or manually obtained by engineers, and the obtained defect images are assembled and classified to create a system that facilitates automatic classification and marking. The defect image may be preprocessed after being acquired, that is, the automatic defect image classification and marking system may include a preprocessing module, for example, to adjust the definition of the defect image, or to eliminate an interference image.
By way of example, the pattern recognition includes, but is not limited to, one or more of an image template matching method, a structural similarity recognition method, a perceptual hashing algorithm, and a histogram method. For example, the plurality of defect images are matched by using an image template matching method, so that the defect images which are matched in a consistent way are classified into the same type. In the matching process, when the size of the defect image is not consistent, a sliding matching window is usually needed to scan the whole image to obtain the best matching patch. I.e. to the same type when the shape of the defect images is the same or substantially the same but different in size. The method can also be used for pattern recognition by various methods, for example, a histogram method is firstly used for rapid recognition, if the histogram method cannot recognize the defect image, the structural similarity recognition method is continuously used for recognition, or the method can also be used for pattern recognition of the same defect image by various methods, and only when the defect image passes through the recognition of a plurality of patterns, the defect image is classified into a certain image type, which is beneficial to improving the recognition accuracy. Although there may be a plurality of shapes of the defect image, after analyzing all defects collected in the factory in a long-term work, the inventor finds that the shape of the defect image can be roughly divided into four types, namely, a line shape (line _ shape), a ring shape (ring _ map), a scratch shape (scratch) and a divergent spin shape (spin), and when the pattern recognition module recognizes the defect image, the pattern recognition module classifies and marks the defect image according to the four types. Specifically, as shown in the position marked by the dashed line frame of fig. 2, the line shape indicates that the defect image is substantially in a single line shape, the ring shape indicates that the defect pixel points constituting the defect image are substantially located in the circumferential direction of a circle, the scratch shape indicates that the defect image is substantially in a plurality of irregular line shapes, and the divergent rotation shape indicates that the defect image is substantially diffused outward at a certain central point. The inventors have analyzed the causes of the defect images having the same shape, and found that the causes of the defect images having the same shape are generally common, for example, a ring-shaped defect may be generated in a process requiring rotation, a linear defect is easily generated in a translation process of a wafer, and the like. Therefore, the defect images are accurately classified, so that the generation reasons of the defect images can be rapidly analyzed, the defect sources can be improved, and the production yield can be improved.
If the defect image can not be classified according to the specific shape, namely when the image recognition module can not recognize the defect image, the rule recognition module continues to recognize the defect image. In the present embodiment, the defect image is further classified into a normal shape (normal) and a cluster shape (cluster) according to the rule identification in consideration of the difference in the size range of the defect image and the degree of influence on the yield and the like. And when the rule identification module identifies the defect images which cannot be classified and marked according to the shapes, the rule identification module classifies and marks the defect images according to the two types. In a specific example, the rule identifies the distance between defect points constituting a defect image, and a defect point with a distance of <5um in the defect image is denoted as Q1 (which can also be described as a point with the defect point as a fixed point and the distance between the point and other defect points being <5 μm, then the point is denoted as Q1), and when the number of Q1 in the same defect image is greater than 10, the defect image is classified as a cluster shape; defects with a distance of [5um,1cm ] (i.e. a distance of 1-5 μm and including end points) in the defect image are marked as Q2, when the number of Q2 in the same defect image is more than 30, the defect image is classified as cluster, the defect image without cluster is classified as normal, the left side in fig. 3 is a cluster defect image and the right side is a normal defect image. Generally, different types of defect images are caused by different causes, and there are many commonalities among defect images of the same type. Therefore, the defect images which cannot be classified according to the specific shapes are further subdivided, and the defect sources are further and rapidly checked.
Of course, it should be noted that the category of the defect image may have other definitions according to different situations, and the definition is not strictly defined in this embodiment, but the above classification can meet the production requirements in most semiconductor factories. In addition, the invention preferably carries out pattern recognition first and then carries out rule recognition, thus improving the classification accuracy.
For example, the automatic defect image classification and marking system may further include a correction and/or update module, so that after the automatic defect image classification and marking system is established, correction and/or update may be performed, for example, a defect image classified inaccurately may be reclassified by manual correction, or a subsequently collected defect image may be classified and marked by the automatic defect image classification and marking system and then stored in the system for updating the database.
The present invention further provides a method for establishing the automatic defect image classification and marking system according to any of the above embodiments, so that the above description of the automatic defect image classification and marking system is incorporated herein by reference in its entirety, and will not be described in detail. The establishing method comprises the following steps:
formulating classification types of the defect images, wherein the classification types comprise a type based on pattern recognition and a type based on rule recognition;
carrying out pattern recognition on the defect image, and storing the defect image into a corresponding pattern library according to different shapes;
and performing rule identification on the defect images which are not classified after the pattern identification, and storing the defect images which accord with different identification rules into a corresponding pattern library.
It should be noted that the actual storage space of the defect image may be loaded on the same hardware, that is, stored in the same database, and the same database is simply divided into a plurality of storage units according to different classification types.
In particular, the classification type may be formulated empirically by an engineer. In this embodiment, the pattern recognition-based types include a line shape, a loop shape, a scratch shape, and a divergent rotation shape, and the rule recognition-based types include a convergence shape and a normal shape. Before the pattern recognition is performed on the defect image, the defect image can be preprocessed, for example, the definition and the like of the defect image are adjusted, or an interference image is eliminated. After the automatic defect image classification and marking system is established, correction and/or updating can be performed, for example, the defect images which are classified inaccurately are reclassified by manual correction, or the subsequently collected defect images are classified and marked by the automatic defect image classification and marking system and then are stored in the system for updating the database.
The present invention also provides an application of the system for automatically classifying and marking defect images in any of the above-mentioned schemes, that is, an automatic classifying and marking method for defect images based on the system for automatically classifying and marking defect images in any of the above-mentioned schemes, so that the introduction to the system for automatically classifying and marking defect images can be incorporated herein in its entirety. The automatic classification marking method for the defect images comprises the following steps:
storing the defect images to be classified and marked and performing format conversion;
and performing pattern recognition on the defect image after format conversion so as to perform classification marking according to the shape of the defect image, and if the defect image cannot be subjected to classification marking according to the shape, performing rule recognition on the defect image so as to perform classification marking according to the rule on the defect image which cannot be subjected to classification marking according to the shape.
Specifically, as an example, the image conversion module converts the defect image from the kafka image format to the lattice format, i.e. the original defect image may be stored in the kafka image format, but may also be stored in other formats such as png or even tif, jpg, etc., and what is important is that the original storage format is suitable for being subsequently converted into a two-dimensional lattice format for computer processing, such as bmp format, etc. In the embodiment, the initial defect image is stored in the kafka image format, which is beneficial to reducing the storage capacity and meeting the real-time processing requirement of large-capacity data. The defect image may be automatically captured by a defect capture camera disposed in a semiconductor factory, or may be obtained manually by an engineer, and the source of the defect image is not limited in this embodiment. After the defect image is acquired, preprocessing may be performed, such as adjusting the sharpness of the defect image, or eliminating an interference image.
By way of example, the pattern recognition includes, but is not limited to, one or more of an image template matching method, a structural similarity recognition method, a perceptual hashing algorithm, and a histogram method. For example, an image template matching method is adopted to match a defect image to be classified and marked with the defect image automatic classification and marking system to determine whether the defect image to be classified and marked belongs to one of a linear shape (ine _ shape), a ring shape (ring _ map), a scratch shape (scratch) and a divergent rotation shape (spin). And according to the needs, can adopt multiple methods to carry on the figure recognition to the same defective map at the same time, will be classified into a certain image type only when the defective map passes multiple figure recognition, help to improve and discern the accuracy. If the defect image to be classified and marked does not conform to the image types, further rule identification can be carried out according to the following rules: identifying the distance between the defect points forming the defect image, wherein the defect points with the distance of <5um in the defect image are marked as Q1 (which can also be described as Q1 if the distance between the defect point and other defect points is <5 mu m by taking the defect point as a fixed point), and when the number of Q1 in the same defect image is more than 10, the defect image is classified and marked as a cluster shape; defects with a distance of [5um,1cm ] (i.e. a distance of 1-5 μm and including end points) in the defect image are marked as Q2, when the number of Q2 in the same defect image is more than 30, the defect image is classified as cluster, and the defect image with non-cluster shape is classified as normal.
The automatic defect image classifying and marking method is not only suitable for classifying and marking the defect images generated in the semiconductor chip preparation process, but also suitable for any image classifying and marking needing to be classified or identified.
The invention also provides a control module, which comprises a processor 21 and a memory 22; the memory 22 is used for storing a computer program; the processor 21 is configured to execute the computer program stored in the memory 22 to make the control module execute the method for automatically classifying and marking defect images as described in any one of the above aspects.
It should be noted that the division of each functional unit of the above modules is only a logical division, and the actual implementation may be wholly or partially integrated into one physical entity, or may be physically separated. And these units can be implemented entirely in software, invoked by a processing element; or may be implemented entirely in hardware; the method can also be realized partly in the form of calling software by the processing element and partly in the form of hardware. For example, the control module may be a processing element separately set up, or may be implemented by being integrated in a chip of the apparatus, or may be stored in a memory of the apparatus in the form of program code, and the function of the control module may be called and executed by a processing element of the apparatus. Other modules are implemented similarly. In addition, all or part of the modules can be integrated together or can be independently realized. The processing element described herein may be an integrated circuit having signal processing capabilities. In implementation, each step of the above method or each module above may be implemented by an integrated logic circuit of hardware in a processor element or an instruction in the form of software.
For example, the above modules may be one or more integrated circuits configured to implement the above methods, such as: one or more Application Specific Integrated Circuits (ASICs), or one or more microprocessors (DSPs), or one or more Field Programmable Gate Arrays (FPGAs), among others. For another example, when one of the above modules is implemented in the form of a Processing element scheduler code, the Processing element may be a general-purpose processor, such as a Central Processing Unit (CPU) or other processor capable of calling program code. For another example, these modules may be integrated together and implemented in the form of a system-on-a-chip (SOC).
As shown in fig. 4, in an embodiment, the control module of the present invention includes: a processor 21 and a memory 22.
The memory 22 is used for storing a computer program.
The memory 22 includes: various media that can store program codes, such as ROM, RAM, magnetic disk, U-disk, memory card, or optical disk.
The processor 21 is connected to the memory 22, and is configured to execute the computer program stored in the memory 22, so that the control module executes the method for automatically classifying and marking defect images according to any one of the foregoing embodiments.
Preferably, the Processor 21 may be a general-purpose Processor, including a Central Processing Unit (CPU), a Network Processor (NP), and the like; the Integrated Circuit may also be a Digital Signal Processor (DSP), an Application Specific Integrated Circuit (ASIC), a Field Programmable Gate Array (FPGA) or other Programmable logic device, discrete Gate or transistor logic device, or discrete hardware components.
The present invention also provides a storage medium having a computer program stored thereon, wherein the computer program, when executed by a processor, implements the method for automatically classifying and marking a defect image according to any one of the above aspects.
Preferably, the storage medium includes various media that can store program codes, such as ROM, RAM, a magnetic disk, a usb disk, a memory card, or an optical disk.
In summary, the present invention provides a defect image automatic classification marking system and a method for creating the same, a defect image automatic classification marking method based on the defect image automatic classification marking system, a control module and a storage medium. The automatic defect image classification and marking system comprises a storage module, an image conversion module, an image identification module, a rule identification module and an output module; the storage module is used for storing the defect image; the image conversion module is connected with the storage module and is used for carrying out format conversion on the stored defect image; the image recognition module is connected with the image conversion module and is used for carrying out image recognition on the defect image after format conversion so as to carry out classification marking according to the shape of the defect image; the rule identification module is connected with the image conversion module and/or the image identification module and is used for classifying and marking the defect images which cannot be classified and marked according to shapes after format conversion according to rules; and the output module is connected with the pattern recognition module and the rule recognition module and is used for outputting a classification marking result. The automatic defect image classification and marking system can be loaded on a computer, so that the automatic defect image classification and marking system can automatically classify and mark defect images, can quickly and accurately classify and analyze the defect images compared with the traditional mode of manual visual inspection and classification, is favorable for quickly and accurately finding out the root cause of problems, is favorable for saving manpower, can avoid image classification difference caused by artificial subjective factors, and is favorable for improving the production yield.
The foregoing embodiments are merely illustrative of the principles and utilities of the present invention and are not intended to limit the invention. Any person skilled in the art can modify or change the above-mentioned embodiments without departing from the spirit and scope of the present invention. Accordingly, it is intended that all equivalent modifications or changes which can be made by those skilled in the art without departing from the spirit and technical spirit of the present invention be covered by the claims of the present invention.
Claims (10)
1. An automatic classification marking system for defect images in semiconductor chip preparation, comprising:
the storage module is used for storing the defect image;
the image conversion module is connected with the storage module and is used for carrying out format conversion on the stored defect image;
the image conversion module is used for converting the defect image into a format of a defect image, and converting the defect image into a format of the defect image;
the rule identification module is connected with the image conversion module and/or the image identification module and is used for classifying and marking the defect images which cannot be classified and marked according to the shapes of the defect images after format conversion according to rules;
and the output module is connected with the pattern recognition module and the rule recognition module and used for outputting the classification marking result.
2. The automatic classification marking system for defect images according to claim 1, characterized in that: the image conversion module converts the defect image from a kafka image format to a lattice format.
3. The automatic classification marking system for defect images according to claim 1, characterized in that: the pattern recognition comprises one or more of an image template matching method, a structural similarity recognition method, a perceptual hash algorithm and a histogram method.
4. The automatic classification marking system for defect images according to claim 1, characterized in that: the defect image is divided into a linear shape, a circular shape, a scratch shape, and a divergent rotation shape according to the shape of the defect image.
5. The automatic classification marking system for defect images according to claim 1, characterized in that: the defect image is classified into a normal shape and an aggregate shape according to rule identification.
6. The system of claim 5, wherein the rules are identified as: identifying the distance between defect points forming a defect image, wherein the defect points with the distance of <5um in the defect image are marked as Q1, and when the number of Q1 in the same defect image is more than 10, the defect image is classified and marked as a cluster; and (3) recording the defects with the distance of [5um,1cm ] in the defect image as Q2, and when the number of Q2 in the same defect image is more than 30, classifying the defect image as an aggregation shape, and classifying the defect image with a non-aggregation shape as a normal shape.
7. A method for establishing a defect image automatic classification and marking system according to any one of claims 1 to 6, comprising:
formulating classification types of the defect images, wherein the classification types comprise a type based on pattern recognition and a type based on rule recognition;
carrying out pattern recognition on the defect image, and storing the defect image into a corresponding pattern library according to different shapes;
and performing rule identification on the defect images which are not classified after the pattern identification, and storing the defect images which accord with different identification rules into a corresponding pattern library.
8. A method for automatically classifying and marking a defect image by using the system for automatically classifying and marking a defect image according to any one of claims 1 to 6, comprising:
storing the defect images to be classified and marked and performing format conversion;
and performing pattern recognition on the defect image after format conversion so as to perform classification marking according to the shape of the defect image, and if the defect image cannot be subjected to classification marking according to the shape, performing rule recognition on the defect image so as to perform classification marking according to the rule on the defect image which cannot be subjected to classification marking according to the shape.
9. A control module, characterized by: the method comprises the following steps: a processor and a memory;
the memory is used for storing a computer program;
the processor is configured to execute the computer program stored in the memory to cause the control module to perform the method of automatically classifying and marking a defect image according to claim 8.
10. A storage medium having stored thereon a computer program, wherein the computer program, when executed by a processor, implements the method for automatic classification marking of defect images according to claim 8.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202011229553.6A CN112446857A (en) | 2020-11-06 | 2020-11-06 | Automatic classification marking system, establishing method and classification marking method for defect images |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202011229553.6A CN112446857A (en) | 2020-11-06 | 2020-11-06 | Automatic classification marking system, establishing method and classification marking method for defect images |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN112446857A true CN112446857A (en) | 2021-03-05 |
Family
ID=74736126
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202011229553.6A Pending CN112446857A (en) | 2020-11-06 | 2020-11-06 | Automatic classification marking system, establishing method and classification marking method for defect images |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN112446857A (en) |
Citations (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JP2001168160A (en) * | 1999-12-07 | 2001-06-22 | Sony Corp | System for inspecting semiconductor wafer |
| JP2001250397A (en) * | 1999-12-27 | 2001-09-14 | Mitsubishi Electric Corp | Fault analyzing method, derivation method for degeneration threshold value, and recording medium |
| JP2006352173A (en) * | 2001-04-10 | 2006-12-28 | Hitachi Ltd | Method of analyzing defect data and device thereof |
| JP2012174896A (en) * | 2011-02-22 | 2012-09-10 | Lasertec Corp | Inspection device and defect inspection method |
| US20130279794A1 (en) * | 2012-04-19 | 2013-10-24 | Applied Materials Israel Ltd. | Integration of automatic and manual defect classification |
| US20150168311A1 (en) * | 2013-12-18 | 2015-06-18 | Lasertec Corporation | Defect classifying method and inspection apparatus |
| KR20190111644A (en) * | 2018-03-23 | 2019-10-02 | (주)브릭 | Apparatus and method of classifying defect type of repetition pattern image |
| TW202037906A (en) * | 2019-01-10 | 2020-10-16 | 美商蘭姆研究公司 | Defect classification and source analysis for semiconductor equipment |
-
2020
- 2020-11-06 CN CN202011229553.6A patent/CN112446857A/en active Pending
Patent Citations (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JP2001168160A (en) * | 1999-12-07 | 2001-06-22 | Sony Corp | System for inspecting semiconductor wafer |
| JP2001250397A (en) * | 1999-12-27 | 2001-09-14 | Mitsubishi Electric Corp | Fault analyzing method, derivation method for degeneration threshold value, and recording medium |
| JP2006352173A (en) * | 2001-04-10 | 2006-12-28 | Hitachi Ltd | Method of analyzing defect data and device thereof |
| JP2012174896A (en) * | 2011-02-22 | 2012-09-10 | Lasertec Corp | Inspection device and defect inspection method |
| US20130279794A1 (en) * | 2012-04-19 | 2013-10-24 | Applied Materials Israel Ltd. | Integration of automatic and manual defect classification |
| US20150168311A1 (en) * | 2013-12-18 | 2015-06-18 | Lasertec Corporation | Defect classifying method and inspection apparatus |
| KR20190111644A (en) * | 2018-03-23 | 2019-10-02 | (주)브릭 | Apparatus and method of classifying defect type of repetition pattern image |
| TW202037906A (en) * | 2019-01-10 | 2020-10-16 | 美商蘭姆研究公司 | Defect classification and source analysis for semiconductor equipment |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN109977808B (en) | Wafer surface defect mode detection and analysis method | |
| TWI755613B (en) | Pattern grouping method based on machine learning | |
| WO2020177432A1 (en) | Multi-tag object detection method and system based on target detection network, and apparatuses | |
| CN109829914A (en) | The method and apparatus of testing product defect | |
| CN109840889A (en) | High-precision vision measurement method, device and system based on bionic Algorithm | |
| JP2019087181A (en) | Device and method for image inspection | |
| CN113554631B (en) | Chip surface defect detection method based on improved network | |
| WO2024021461A1 (en) | Defect detection method and apparatus, device, and storage medium | |
| CN114862814A (en) | Solar cell panel defect detection method and system, storage medium and terminal | |
| US12013918B2 (en) | Method and apparatus for clustering images | |
| WO2021101851A1 (en) | Finding semiconductor defects using convolutional context attributes | |
| CN115131596A (en) | Defect classification device, method, and program | |
| CN108537792B (en) | Power defect image identification method based on convolutional neural network | |
| TW202242794A (en) | Image processing method and system, and non-transitory computer readable medium | |
| CN112446857A (en) | Automatic classification marking system, establishing method and classification marking method for defect images | |
| CN117910073A (en) | Artwork package design optimization system and method based on 3D printing technology | |
| CN114120318B (en) | Dark field image target point accurate extraction method based on integrated decision tree | |
| CN116109627A (en) | Defect detection method, device and medium based on migration learning and small sample learning | |
| CN114841980A (en) | Insulator defect detection method and system based on line patrol aerial image | |
| CN115082449A (en) | Electronic component defect detection method | |
| KR20220101373A (en) | Apparatus and method for detecting defects in pcb based on video | |
| Tsai et al. | Enhancing the data analysis in IC testing by machine learning techniques | |
| CN117115143B (en) | Method and device for detecting wafer core particle, computer equipment and storage medium | |
| CN118501159B (en) | Automobile part defect detection method and system based on machine vision | |
| WO2024011756A1 (en) | Image acquisition parameter adjustment method and system, electronic device, and storage medium |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination |