CN117058075A - Crystallization detection method and device and electronic equipment - Google Patents
Crystallization detection method and device and electronic equipment Download PDFInfo
- Publication number
- CN117058075A CN117058075A CN202310904193.2A CN202310904193A CN117058075A CN 117058075 A CN117058075 A CN 117058075A CN 202310904193 A CN202310904193 A CN 202310904193A CN 117058075 A CN117058075 A CN 117058075A
- Authority
- CN
- China
- Prior art keywords
- crystallization
- image
- area
- suspected
- detection frame
- 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
- 238000002425 crystallisation Methods 0.000 title claims abstract description 203
- 230000008025 crystallization Effects 0.000 title claims abstract description 203
- 238000001514 detection method Methods 0.000 title claims abstract description 169
- 229910021421 monocrystalline silicon Inorganic materials 0.000 claims abstract description 31
- 238000013145 classification model Methods 0.000 claims abstract description 30
- 238000000034 method Methods 0.000 claims description 34
- 238000005070 sampling Methods 0.000 claims description 13
- 238000012549 training Methods 0.000 claims description 9
- 238000002372 labelling Methods 0.000 claims description 3
- 230000007547 defect Effects 0.000 abstract description 9
- 239000013078 crystal Substances 0.000 description 28
- 230000008569 process Effects 0.000 description 11
- 238000010586 diagram Methods 0.000 description 8
- 238000004590 computer program Methods 0.000 description 6
- 238000004891 communication Methods 0.000 description 5
- 230000002159 abnormal effect Effects 0.000 description 4
- 230000008859 change Effects 0.000 description 4
- 238000012216 screening Methods 0.000 description 4
- 238000004458 analytical method Methods 0.000 description 3
- 238000007689 inspection Methods 0.000 description 3
- 238000012360 testing method Methods 0.000 description 3
- 238000012795 verification Methods 0.000 description 3
- XUIMIQQOPSSXEZ-UHFFFAOYSA-N Silicon Chemical compound [Si] XUIMIQQOPSSXEZ-UHFFFAOYSA-N 0.000 description 2
- 230000002452 interceptive effect Effects 0.000 description 2
- 230000003287 optical effect Effects 0.000 description 2
- 230000000644 propagated effect Effects 0.000 description 2
- 229910052710 silicon Inorganic materials 0.000 description 2
- 239000010703 silicon Substances 0.000 description 2
- 230000006978 adaptation Effects 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 238000004422 calculation algorithm Methods 0.000 description 1
- 238000004364 calculation method Methods 0.000 description 1
- 238000013527 convolutional neural network Methods 0.000 description 1
- 238000013500 data storage Methods 0.000 description 1
- 238000005516 engineering process Methods 0.000 description 1
- 230000002708 enhancing effect Effects 0.000 description 1
- 238000000605 extraction Methods 0.000 description 1
- 239000000835 fiber Substances 0.000 description 1
- 230000006870 function Effects 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 238000012545 processing Methods 0.000 description 1
- 230000004044 response Effects 0.000 description 1
- 238000013519 translation Methods 0.000 description 1
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
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/20—Image preprocessing
- G06V10/22—Image preprocessing by selection of a specific region containing or referencing a pattern; Locating or processing of specific regions to guide the detection or recognition
-
- 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
- G06V10/764—Arrangements for image or video recognition or understanding using pattern recognition or machine learning using classification, e.g. of video objects
-
- 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
- G06V10/77—Processing image or video features in feature spaces; using data integration or data reduction, e.g. principal component analysis [PCA] or independent component analysis [ICA] or self-organising maps [SOM]; Blind source separation
- G06V10/774—Generating sets of training patterns; Bootstrap methods, e.g. bagging or boosting
-
- 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
- G06T2207/10—Image acquisition modality
- G06T2207/10004—Still image; Photographic image
-
- 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
- G06T2207/20—Special algorithmic details
- G06T2207/20081—Training; Learning
-
- 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
Abstract
The application provides a crystallization detection method, a crystallization detection device and electronic equipment, and relates to the technical field of defect detection. The crystallization detection method comprises the following steps: acquiring a current first image of monocrystalline silicon to be detected; inputting the first image into a pre-trained target classification model, and outputting a first detection frame of one or more suspected crystallization areas in the first image; acquiring a history image frame adjacent to the first image as a reference image, wherein the reference image comprises a second detection frame of the determined second crystallization area; and determining a first crystallization area of the first image from the suspected crystallization areas according to the first detection frame and the second detection frame. According to the embodiment of the application, the suspected crystallization area is screened through the second detection frame in the reference image, so that the accuracy of determining the first crystallization area is improved.
Description
Technical Field
The present application relates to the field of defect detection technologies, and in particular, to a crystallization detection method, a crystallization detection device, and an electronic device.
Background
In the equal-diameter ending process of monocrystalline silicon, abnormal crystallization defects can occur sporadically, the manual inspection efficiency is low, the detection error of the abnormal crystallization defects is large, the abnormal crystallization defects cannot be captured timely, the abnormal crystallization is more and more serious, and even the normal operation of the process is affected.
Disclosure of Invention
The embodiment of the application provides a crystallization detection method, a crystallization detection device and electronic equipment.
An embodiment of a first aspect of the present application provides a crystallization detection method, including:
acquiring a current first image of monocrystalline silicon to be detected;
inputting the first image into a pre-trained target classification model, and outputting a first detection frame of one or more suspected crystallization areas in the first image;
acquiring a history image frame adjacent to the first image as a reference image, wherein the reference image comprises a second detection frame of the determined second crystallization area;
and determining a first crystallization area of the first image from the suspected crystallization area according to the first detection frame and the second detection frame.
In one embodiment of the present application, the determining, according to the first detection frame and the second detection frame, the first crystallization area of the first image from the suspected crystallization areas includes:
calculating the similarity degree of the one or more suspected crystallization areas and each second crystallization area according to the first detection frame and the second detection frame;
and determining a first crystallization region of the first image from the suspected crystallization regions according to the similarity degree of the suspected crystallization region and each second crystallization region.
In one embodiment of the present application, the determining the first crystallization region of the first image from the suspected crystallization region according to the degree of similarity of the suspected crystallization region and each of the second crystallization regions includes:
judging whether a first similarity degree smaller than or equal to a preset threshold value exists in the similarity degree of any suspected crystallization region and each second crystallization region for any suspected crystallization region;
and if the first similarity exists, taking any suspected crystallization area as one first crystallization area.
In one embodiment of the present application, the calculating the one or more suspected crystallization areas according to the first detection frame and the second detection frame, and the similarity degree between the one or more suspected crystallization areas and each of the second crystallization areas includes:
and aiming at any first detection frame, acquiring the cross-over ratio between any first detection frame and each second detection frame, and taking the cross-over ratio as the similarity degree of the suspected crystallization area corresponding to any first detection frame and each second crystallization area.
In one embodiment of the present application, after the determining the first crystallization region of the first image from the suspected crystallization region, the method includes:
and sending out early warning reminding information based on the position of the first crystallization area.
In one embodiment of the present application, the acquiring the historical image frames adjacent to the first image as the reference image includes:
storing each historical image frame into a storage area, wherein each historical image frame in the storage area comprises a historical detection frame of a determined historical crystallization area;
historical image frames adjacent to the first image are acquired from the storage area as reference images.
In one embodiment of the present application, the training method of the object classification model includes:
acquiring a sampling image, and determining a crystallization area, a noise area and a normal area in the sampling image;
marking a crystallization area, a noise area and a normal area in the sampling image to obtain a sample image;
and training the target classification model by using the sample image to obtain the target classification model.
In one embodiment of the present application, the labeling the crystallized region, the noise region and the normal region in the sampled image includes:
generating respective corresponding target detection rectangular frames based on the position information of the crystallization region, the noise region and the normal region;
and marking the crystallization area, the noise area and the normal area by using the respective target detection rectangular frames.
An embodiment of a second aspect of the present application provides a crystallization detection device, including:
the first acquisition module is used for acquiring a current first image of the monocrystalline silicon to be detected;
the second acquisition module is used for inputting the first image into a pre-trained target classification model and outputting a first detection frame of one or more suspected crystallization areas in the first image;
a third acquisition module, configured to acquire a historical image frame adjacent to the first image as a reference image, where the reference image includes a second detection frame of the determined second crystallization area;
and the crystallization detection module is used for determining a first crystallization area of the first image from the suspected crystallization area according to the first detection frame and the second detection frame.
An embodiment of a third aspect of the present application provides an electronic device, including: a processor; a memory for storing the processor-executable instructions; wherein the processor is configured to execute the instructions to implement the crystallization detection method according to the embodiment of the first aspect of the present application.
An embodiment of a fourth aspect of the present application proposes a non-transitory computer readable storage medium, which when executed by a processor of an electronic device, enables the electronic device to perform the method proposed by the embodiment of the first aspect of the present application.
An embodiment of the fifth aspect of the present application proposes a computer program product comprising a computer program which, when executed by a processor in a communication device, implements the method proposed by the embodiment of the first aspect of the present application.
The technical scheme provided by the embodiment of the application at least has the following beneficial effects:
the method comprises the steps of detecting a current first image of monocrystalline silicon to be detected to obtain a first detection frame of one or more suspected crystal areas, screening the one or more suspected crystal areas in the first image according to a reference image of a second detection frame of which the second crystal areas are determined, determining the first crystal areas, and improving the accuracy of distinguishing the first crystal areas by utilizing a target classification model and the twice screening of the reference image.
It is to be understood that both the foregoing general description and the following detailed description are exemplary and explanatory only and are not restrictive of the application as claimed.
Drawings
The foregoing and/or additional aspects and advantages of the application will become apparent and readily appreciated from the following description of the embodiments, taken in conjunction with the accompanying drawings, in which:
FIG. 1 is a schematic flow chart of a crystallization detection method according to an embodiment of the present application;
FIG. 2 is a flow chart of another crystallization detection method according to an embodiment of the present application;
FIG. 2A is a schematic diagram of an image annotation according to an embodiment of the present application;
FIG. 2B is a schematic diagram of matching of a detection frame during similarity calculation according to an embodiment of the present application;
FIG. 3 is a schematic diagram of a crystallization detecting apparatus according to an embodiment of the present application;
fig. 4 is a schematic structural diagram of an electronic device according to an embodiment of the present application;
fig. 5 is a schematic structural diagram of another electronic device according to an embodiment of the present application.
Detailed Description
Reference will now be made in detail to exemplary embodiments, examples of which are illustrated in the accompanying drawings. When the following description refers to the accompanying drawings, the same numbers in different drawings refer to the same or similar elements, unless otherwise indicated. The implementations described in the following exemplary embodiments do not represent all implementations consistent with embodiments of the application. Rather, they are merely examples of apparatus and methods consistent with aspects of embodiments of the application as detailed in the accompanying claims.
The terminology used in the embodiments of the application is for the purpose of describing particular embodiments only and is not intended to be limiting of embodiments of the application. As used in this application and the appended claims, the singular forms "a," "an," and "the" are intended to include the plural forms as well, unless the context clearly indicates otherwise. It should also be understood that the term "and/or" as used herein refers to and encompasses any or all possible combinations of one or more of the associated listed items.
It should be understood that although the terms first, second, third, etc. may be used in embodiments of the present application to describe various information, these information should not be limited to these terms. These terms are only used to distinguish one type of information from another. For example, the first information may also be referred to as second information, and similarly, the second information may also be referred to as first information, without departing from the scope of embodiments of the present application. The words "if" and "if" as used herein may be interpreted as "at … …" or "at … …" or "in response to a determination", depending on the context.
Reference will now be made in detail to embodiments of the present application, examples of which are illustrated in the accompanying drawings, wherein like or similar reference numerals refer to the like or similar elements throughout. The embodiments described below by referring to the drawings are illustrative and intended to explain the present application and should not be construed as limiting the application.
It should be noted that, the crystallization detection method provided in any one of the embodiments of the present application may be performed alone or in combination with possible implementation methods in other embodiments, and may also be performed in combination with any one of the technical solutions in the related art.
The following describes a crystallization detection method, a crystallization detection device and an electronic device according to an embodiment of the present application with reference to the accompanying drawings.
Fig. 1 is a schematic flow chart of a crystallization detection method according to an embodiment of the present application. As shown in fig. 1, the method includes, but is not limited to, the steps of:
s101, acquiring a current first image of monocrystalline silicon to be detected.
In some implementations, a camera may be employed to image the monocrystalline silicon to be inspected. A first image of a single crystal silicon isodiametric ending process site is acquired, for example, using a CCD camera.
In some implementations, the first image includes an area where the single crystal silicon to be detected is located.
After a current first image comprising the monocrystalline silicon to be detected is obtained, detecting whether crystal defects exist in the monocrystalline silicon to be detected according to the first image.
It is understood that the first image is an image taken in real time of the single crystal silicon to be detected in the equal diameter ending process of the single crystal silicon to be detected, that is, the first image is a current image of the single crystal silicon to be detected.
In some implementations, since the crystallization anomaly is an sporadic defect, it is difficult to capture by a human and the number of crystals is small, and thus in order to ensure accuracy of detection of the crystallization defect by using the first image, image data in the first image may be enhanced and expanded.
Optionally, the method of enhancing and expanding the image data may adopt rotation, translation, stretching or flipping, etc.
S102, inputting the first image into a pre-trained target classification model, and outputting a first detection frame of one or more suspected crystallization areas in the first image.
Alternatively, feature extraction may be performed on the first image, and a first detection frame of the first or more suspected crystalline regions in the first image may be determined based on the extracted features and the pre-trained target classification model.
Alternatively, the object classification model may be a fast object detection algorithm or a convolutional neural network, or the like.
In some implementations, the first detection frame may be a rectangular frame, which is used to reflect a position corresponding to the suspected crystallization area.
In some implementations, the object classification model may also output an interference region in the first image that is a region that is more similar to the crystallized region and a normal region that is a region in the first image where it is determined that the crystallized region is not present.
And outputting one or more suspected crystallization areas through the target classification model, distinguishing the suspected crystallization areas again, determining whether the suspected crystallization areas are crystallization areas, and improving the accuracy of distinguishing the crystallization areas.
And S103, acquiring a history image frame adjacent to the first image as a reference image, wherein the reference image comprises a second detection frame of the determined second crystallization area.
Alternatively, the first image neighboring history image frame may be used as the reference image. For example, the previous image frame adjacent to the first image is a reference image, and the reference image is used for carrying out subsequent auxiliary analysis.
It is understood that the reference image is an image in which the crystalline region has been detected, that is, the reference image includes the second detection frame of the second crystalline region that has been determined.
It should be noted that, the monocrystalline silicon corresponding to the reference image is the same as the monocrystalline silicon to be detected currently, that is, the first image and the reference image are images of the monocrystalline silicon to be detected currently in different periods.
Alternatively, the second detection frame may be a rectangular frame for reflecting a position corresponding to the second crystallization area.
S104, determining a first crystallization area of the first image from the suspected crystallization areas according to the first detection frame and the second detection frame.
Since the silicon to be detected is in the crucible in the silicon wafer finalizing process, the crucible is in continuous rotation, and therefore if there is a crystallization area in the current first image, there is a difference between the positions of the crystallization area in the first image and the crystallization area in the previous image frame. That is, there may be a deviation in the position between the crystallized region in the first image and the second crystallized region in the reference image.
Further, position comparison is performed according to a first detection frame of one or more suspected crystallization areas in the first image and a second detection frame of a second crystallization area in the reference image.
It is understood that, since the crystals themselves are movable and the crystal morphology is also changed, the positions or shapes of the crystal defects are changed over time, so that based on the comparison between the second detection frame in the reference image and the first detection frame in the first image, it is determined whether the shapes and positions between the first detection frame and the second detection frame are changed, thereby judging whether the suspected crystal region in the first image is a real crystal region.
Optionally, if the position corresponding to any one of the first detection frames in the first image is identical to the shape and the position of the second detection frame in the second crystallization region, the suspected crystallization region corresponding to the first detection frame may be determined to be an amorphous region.
Optionally, if the shape and position of any one of the first detection frames in the first image deviate from the shape and position of the second detection frame in the second crystallization area, which means that the shape and position change occurs between the suspected crystallization area and the crystallization area in the two frames of images, it may be determined that the suspected crystallization area corresponding to the first detection frame is the crystallization area.
And comparing all the first detection frames in the first image with all the second detection frames in the reference image in pairs, and determining a crystallization area from the suspected crystallization area to be used as a first crystallization area of the first image.
Alternatively, the cross-over ratio between the first detection frame in the first image and the second detection frame in the reference image may be calculated, and the overlapping area condition of the first detection frame in the first image and the second detection frame in the reference image is reflected by the cross-over ratio, so as to reflect whether the shapes and positions of the first detection frame in the first image and the second detection frame in the reference image deviate.
In the embodiment of the application, the first detection frame of one or more suspected crystallization areas is obtained by detecting the current first image of the monocrystalline silicon to be detected, the one or more suspected crystallization areas in the first image are screened according to the reference image of the second detection frame which is already determined to be the second crystallization area, the first crystallization area is determined, and the accuracy of distinguishing the first crystallization area is improved by utilizing the target classification model and the twice screening of the reference image, so that the problems of low manual inspection efficiency and large error are solved.
Fig. 2 is a schematic flow chart of a crystallization detection method according to an embodiment of the present application. As shown in fig. 2, the method includes, but is not limited to, the steps of:
s201, acquiring a current first image of monocrystalline silicon to be detected.
In the embodiment of the present application, the implementation method of step S201 may be implemented in any manner in each embodiment of the present disclosure, which is not limited herein, and is not described herein again.
S202, inputting the first image into a pre-trained target classification model, and outputting a first detection frame of one or more suspected crystallization areas in the first image.
Alternatively, a sampling image may be acquired, and a crystallization region, a noise region, and a normal region in the sampling image may be determined; marking a crystallization area, a noise area and a normal area in the sampling image to obtain a sample image; training the classification model by using the sample image to obtain a target classification model. That is, the CCD camera can be used for collecting images in the on-site single crystal silicon isodiametric ending process as sampling images, marking the sampling images, and randomly dividing the sampling images into three sets of a training set, a verification set and a test set.
In some implementations, the content that labels the sampled image may be crystalline, noisy, and normal regions in the sampled image.
In some implementations, the noise region in the sampled image may be an interference region that affects the determination of the crystallization region, such as a region where a reflection or moire, etc., is located, as shown in fig. 2A.
Optionally, the labeling of the noise region may be a category to which the noise belongs, for example, the noise region is a ripple region or a reflection region, and the noise region in the sampled image is labeled, so that the influence of the noise region on the judgment of the crystallization region is reduced, and the accuracy of the judgment of the crystallization region is improved.
Alternatively, respective corresponding target detection rectangular frames may be generated based on the positional information of the crystallization region, the noise region, and the normal region; and marking the crystallization area, the noise area and the normal area by utilizing the respective target detection rectangular frames. That is, the crystallized region, the noise region, and the normal region in the sample image are all marked with respective target detection rectangular frames.
Randomly dividing all marked sampling images into a training set, a verification set and a test set; training the classification model by using the training set, the verification set and the test set to obtain a trained target classification model, wherein it can be understood that the target classification model is a multi-class target classification model so as to facilitate the subsequent detection and analysis of each class region in the real-time monocrystalline silicon image to be detected.
Further, a current first image of the monocrystalline silicon to be detected, which is acquired in real time, is input into a pre-trained target classification model, and a first detection frame of one or more suspected crystallization areas in the first image is output.
It can be understood that, in addition to the first detection frame of one or more suspected crystallization areas, the detection frame of the noise area and the detection frame of the normal area are also output in the target classification model. However, the application only detects the crystallization area of the monocrystalline silicon to be detected, so that one or more output suspected crystallization areas are analyzed to judge whether the output suspected crystallization areas are crystallization areas on the surface of the monocrystalline silicon to be detected.
It can be understood that the detection frame for outputting the noise area can reduce the interference of the noise area to the detection of the crystallization area, that is, the detection frame for outputting the noise area in the target classification model can improve the accuracy of the detection of the suspected crystallization area, and further judge whether the suspected crystallization area is a real crystallization area or not, so that the accuracy is higher.
In the embodiment of the present application, the implementation manner of step S202 may be implemented in any manner of each embodiment of the present disclosure, which is not limited herein, and is not repeated herein.
And S203, acquiring a history image frame adjacent to the first image as a reference image, wherein the reference image comprises a second detection frame of the determined second crystallization area.
In the single crystal silicon constant diameter head-tail process, since crystals move with the movement of the crucible and the morphology changes, and disturbance areas such as shadows and the like do not move or change, whether or not a suspected crystal area is a true crystal area can be determined by using whether or not crystals and shapes change.
Alternatively, each history image frame may be saved to a storage area, each history image frame in the storage area including a history detection frame of the determined history crystallization area; historical image frames adjacent to the first image are acquired from the storage area as reference images. That is, the image frame with the crystallization area detected in real time is stored in the storage area, and when the first image of the current monocrystalline silicon to be detected is analyzed, the reference image corresponding to the first image is cached in the storage area, so that the reference image and the second detection frame of the second crystallization area in the reference image can be directly utilized for analysis.
In the embodiment of the present application, the implementation manner of step S203 may be implemented in any manner of each embodiment of the present disclosure, which is not limited herein, and is not described herein again.
S204, calculating one or more suspected crystallization areas according to the first detection frame and the second detection frame, and calculating the similarity degree of the suspected crystallization areas and each second crystallization area.
In some implementations, for any first detection frame, the cross-over ratio between any first detection frame and each second detection frame is obtained as the similarity degree of the suspected crystallization region corresponding to any first detection frame and each second crystallization region.
For example, assume that N suspected crystalline regions are detected in the first image, where N is a positive integer, that is, N first detection frames exist in the first image; a second detection frame in which 3 second crystallization areas exist in the reference image; then for any one of the first detection frames, the cross-correlation ratio needs to be calculated with 3 second detection frames respectively, so as to obtain the similarity degree between the first detection frame and each second detection frame, as shown in fig. 2B.
It should be noted that, because the first detection frame and the second detection frame are respectively located in different images, when calculating the intersection ratio between the first detection frame and the second detection frame, the first detection frame and the second detection frame need to be placed on the same plane, for example, the first image and the reference image are overlapped, the position of the first detection frame and the position of the second detection frame are extracted on the plane, and the intersection ratio between the first detection frame and the second detection frame is determined according to the position of the first detection frame and the position of the second detection frame, as the similarity degree between the first detection frame and the second detection frame; the larger the overlap ratio is, the more the positions of the first detection frame and the second detection frame are overlapped, and the closer the first detection frame and the second detection frame are.
S205, determining a first crystallization region of the first image from the suspected crystallization regions according to the degree of similarity of the suspected crystallization region and each of the second crystallization regions.
Optionally, for any suspected crystallization region, judging whether a first similarity degree smaller than or equal to a preset threshold exists in the similarity degree of any suspected crystallization region and each second crystallization region; if the first similarity exists, any suspected crystallization area is taken as a first crystallization area. That is, for any one of the suspected crystal regions, it is determined whether or not the cross-over ratio of the suspected crystal region to each of the second crystal regions is less than or equal to the first cross-over ratio (first degree of similarity) of the threshold value, and if it is determined that the cross-over ratio of the suspected crystal region to one of the second crystal regions is smaller, that is, the overlap is smaller, the suspected crystal region may be a crystal region that moves as the crucible moves, and thus the suspected crystal region is determined to be one of the first crystal regions.
It can be understood that, for any suspected crystallization area, if there is a second similarity degree greater than a preset threshold value in the similarity degree of the suspected crystallization area and each second crystallization area, it indicates that the coincidence degree between the suspected crystallization area and a certain second crystallization area is higher, the feature of dynamic change of crystallization is not satisfied, and if the suspected crystallization area caused by possible interference factors such as ripple or shadow is false detection of the crystallization area, the suspected crystallization area is judged to be a non-crystallization area, the non-crystallization area is ignored, and no alarm or reminding operation is performed.
Therefore, according to the similarity between the first detection frame and the second detection frame, all suspected crystal areas in the first image are subjected to auxiliary judgment, the first crystal area is determined, other areas interfering with false detection are eliminated, and the accuracy of judging the crystal areas is improved.
Further, after determining that the first crystallization area exists in the first image of the monocrystalline silicon to be detected, information of the first crystallization area can be fed back to the industrial personal computer, and early warning reminding information is sent out based on the position of the first crystallization area so as to remind inspection personnel to timely process, and the crystallization is prevented from affecting follow-up processes.
In the embodiment of the application, by calculating the similarity between the first detection frame of any suspected crystallization area in the first image and each second detection frame in the reference image, whether the suspected crystallization area is a true crystallization area is analyzed, and other areas interfering with the crystallization area are screened according to the similarity by combining the movement characteristics of crystallization, so that a final first crystallization area is obtained, and the accuracy of detecting the first crystallization area is ensured.
Fig. 3 is a schematic structural diagram of a crystallization detection device according to an embodiment of the present application. As shown in fig. 3, the crystallization detection device 300 includes:
a first acquiring module 301, configured to acquire a current first image of monocrystalline silicon to be detected;
a second obtaining module 302, configured to input the first image into a pre-trained target classification model, and output a first detection frame of one or more suspected crystallization areas in the first image;
a third obtaining module 303, configured to obtain a history image frame adjacent to the first image as a reference image, where the reference image includes a second detection frame of the determined second crystallization area;
the crystallization detection module 304 is configured to determine a first crystallization area of the first image from the suspected crystallization areas according to the first detection frame and the second detection frame.
According to the embodiment of the application, the first detection frame of one or more suspected crystallization areas is obtained by detecting the current first image of the monocrystalline silicon to be detected, the one or more suspected crystallization areas in the first image are screened according to the reference image of the second detection frame of the second crystallization area, the first crystallization area is determined, and the accuracy of distinguishing the first crystallization area is improved by utilizing the target classification model and the twice screening of the reference image.
Fig. 4 is a block diagram of an electronic device, according to an example embodiment. The electronic device shown in fig. 4 is only an example and should not be construed as limiting the functionality and scope of use of the embodiments of the application.
As shown in fig. 4, the electronic device 400 includes a processor 401 that can perform various appropriate actions and processes according to a program stored in a Read Only Memory (ROM) 402 or a program loaded from a Memory 406 into a random access Memory (RAM, random Access Memory) 403. In the RAM 403, various programs and data necessary for the operation of the electronic device 400 are also stored. The processor 401, the ROM 402, and the RAM 403 are connected to each other by a bus 404. An Input/Output (I/O) interface 405 is also connected to bus 404.
The following components are connected to the I/O interface 405: a memory 406 including a hard disk and the like; and a communication section 407 including a network interface card such as a LAN (local area network ) card, a modem, or the like, the communication section 407 performing communication processing via a network such as the internet; the driver 408 is also connected to the I/O interface 405 as needed.
In particular, according to embodiments of the present application, the processes described above with reference to flowcharts may be implemented as computer software programs. For example, embodiments of the present application include a computer program embodied on a computer readable medium, the computer program containing program code for performing the method shown in the flowchart. In such an embodiment, the computer program may be downloaded and installed from the network through the communication portion 407. The above-described functions defined in the method of the application are performed when the computer program is executed by the processor 401.
In an exemplary embodiment, a storage medium is also provided, e.g., a memory, comprising instructions executable by the processor 401 of the electronic device 400 to perform the above-described method. Alternatively, the storage medium may be a non-transitory computer readable storage medium, which may be, for example, ROM, random Access Memory (RAM), CD-ROM, magnetic tape, floppy disk, optical data storage device, and the like.
In the context of this document, a computer readable storage medium may be any tangible medium that can contain, or store a program for use by or in connection with an instruction execution system, apparatus, or device. In the present application, however, the computer-readable signal medium may include a data signal propagated in baseband or as part of a carrier wave, with the computer-readable program code embodied therein. Such a propagated data signal may take any of a variety of forms, including, but not limited to, electro-magnetic, optical, or any suitable combination of the foregoing. A computer readable signal medium may also be any computer readable medium that is not a computer readable storage medium and that can communicate, propagate, or transport a program for use by or in connection with an instruction execution system, apparatus, or device. Program code embodied on a computer readable medium may be transmitted using any appropriate medium, including but not limited to: wireless, wire, fiber optic cable, RF, etc., or any suitable combination of the foregoing.
Fig. 5 is a block diagram of an electronic device, according to an example embodiment. The electronic device shown in fig. 5 is only an example and should not be construed as limiting the functionality and scope of use of the embodiments of the present application. As shown in fig. 5, the electronic device 500 includes a processor 501 and a memory 502. The memory 502 is used for storing program codes, and the processor 501 is connected to the memory 502 and is used for reading the program codes from the memory 502 to implement the crystallization detection method in the above embodiment.
Alternatively, the number of processors 501 may be one or more.
Optionally, the electronic device may further include an interface 503, and the number of the interfaces 503 may be plural. The interface 503 may be connected to an application program, and may receive data of an external device such as a sensor, or the like.
Other embodiments of the application will be apparent to those skilled in the art from consideration of the specification and practice of the application disclosed herein. This application is intended to cover any variations, uses, or adaptations of the application following, in general, the principles of the application and including such departures from the present disclosure as come within known or customary practice within the art to which the application pertains. It is intended that the specification and examples be considered as exemplary only, with a true scope and spirit of the application being indicated by the following claims.
It is to be understood that the application is not limited to the precise arrangements and instrumentalities shown in the drawings, which have been described above, and that various modifications and changes may be effected without departing from the scope thereof. The scope of the application is limited only by the appended claims.
Claims (10)
1. A crystallization detection method, comprising:
acquiring a current first image of monocrystalline silicon to be detected;
inputting the first image into a pre-trained target classification model, and outputting a first detection frame of one or more suspected crystallization areas in the first image;
acquiring a history image frame adjacent to the first image as a reference image, wherein the reference image comprises a second detection frame of the determined second crystallization area;
and determining a first crystallization area of the first image from the suspected crystallization area according to the first detection frame and the second detection frame.
2. The method of claim 1, wherein determining a first crystallized region of the first image from the suspected crystallized region based on the first detection frame and the second detection frame comprises:
calculating the similarity degree of the one or more suspected crystallization areas and each second crystallization area according to the first detection frame and the second detection frame;
and determining a first crystallization region of the first image from the suspected crystallization regions according to the similarity degree of the suspected crystallization region and each second crystallization region.
3. The method of claim 2, wherein determining a first crystallized region of the first image from the suspected crystallized region based on a degree of similarity of the suspected crystallized region to each of the second crystallized regions comprises:
judging whether a first similarity degree smaller than or equal to a preset threshold value exists in the similarity degree of any suspected crystallization region and each second crystallization region for any suspected crystallization region;
and if the first similarity exists, taking any suspected crystallization area as one first crystallization area.
4. A method according to claim 2 or 3, wherein said calculating the degree of similarity of the one or more suspected crystalline regions to each of the second crystalline regions based on the first and second detection frames comprises:
and aiming at any first detection frame, acquiring the cross-over ratio between any first detection frame and each second detection frame, and taking the cross-over ratio as the similarity degree of the suspected crystallization area corresponding to any first detection frame and each second crystallization area.
5. The method of claim 1, wherein the acquiring historical image frames adjacent to the first image as reference images comprises:
storing each historical image frame into a storage area, wherein each historical image frame in the storage area comprises a historical detection frame of a determined historical crystallization area;
historical image frames adjacent to the first image are acquired from the storage area as reference images.
6. The method of claim 1, wherein the training method of the object classification model comprises:
acquiring a sampling image, and determining a crystallization area, a noise area and a normal area in the sampling image;
marking a crystallization area, a noise area and a normal area in the sampling image to obtain a sample image;
and training the target classification model by using the sample image to obtain the target classification model.
7. The method of claim 6, wherein labeling the crystallized region, the noisy region, and the normal region in the sampled image comprises:
generating respective corresponding target detection rectangular frames based on the position information of the crystallization region, the noise region and the normal region;
and marking the crystallization area, the noise area and the normal area by using the respective target detection rectangular frames.
8. A crystallization detection device, comprising:
the first acquisition module is used for acquiring a current first image of the monocrystalline silicon to be detected;
the second acquisition module is used for inputting the first image into a pre-trained target classification model and outputting a first detection frame of one or more suspected crystallization areas in the first image;
a third acquisition module, configured to acquire a historical image frame adjacent to the first image as a reference image, where the reference image includes a second detection frame of the determined second crystallization area;
and the crystallization detection module is used for determining a first crystallization area of the first image from the suspected crystallization area according to the first detection frame and the second detection frame.
9. An electronic device, comprising:
a processor;
a memory for storing the processor-executable instructions;
wherein the processor is configured to execute the instructions to implement the method of any one of claims 1 to 7.
10. A non-transitory computer readable storage medium, wherein instructions in the storage medium, when executed by a processor of an electronic device, enable the electronic device to perform the method of any one of claims 1 to 7.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202310904193.2A CN117058075A (en) | 2023-07-20 | 2023-07-20 | Crystallization detection method and device and electronic equipment |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202310904193.2A CN117058075A (en) | 2023-07-20 | 2023-07-20 | Crystallization detection method and device and electronic equipment |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN117058075A true CN117058075A (en) | 2023-11-14 |
Family
ID=88652643
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202310904193.2A Pending CN117058075A (en) | 2023-07-20 | 2023-07-20 | Crystallization detection method and device and electronic equipment |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN117058075A (en) |
-
2023
- 2023-07-20 CN CN202310904193.2A patent/CN117058075A/en active Pending
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN110807429B (en) | Construction safety detection method and system based on tiny-YOLOv3 | |
| CN111797890A (en) | Method and system for detecting defects of power transmission line equipment | |
| CN112115897B (en) | Multi-pointer instrument alarm detection method, device, computer equipment and storage medium | |
| CN110335313B (en) | Audio acquisition equipment positioning method and device and speaker identification method and system | |
| CN109285791B (en) | Design layout-based rapid online defect diagnosis, classification and sampling method and system | |
| CN111310826B (en) | Method and device for detecting labeling abnormality of sample set and electronic equipment | |
| CN115980050B (en) | Water quality detection method and device for water outlet, computer equipment and storage medium | |
| CN111033563A (en) | Image analysis method and system for immunochromatography detection | |
| CN111008576A (en) | Pedestrian detection and model training and updating method, device and readable storage medium thereof | |
| CN114863311A (en) | Automatic tracking method and system for inspection target of transformer substation robot | |
| WO2014103617A1 (en) | Alignment device, defect inspection device, alignment method, and control program | |
| CN114331961A (en) | Method for defect detection of an object | |
| CN100433047C (en) | Device and method for detecting blurring of image | |
| CN116485779B (en) | Adaptive wafer defect detection method and device, electronic equipment and storage medium | |
| CN111626104B (en) | Cable hidden trouble point detection method and device based on unmanned aerial vehicle infrared thermal image | |
| CN116188510B (en) | Enterprise emission data acquisition system based on multiple sensors | |
| CN117372377A (en) | Broken line detection method and device for monocrystalline silicon ridge line and electronic equipment | |
| CN117058075A (en) | Crystallization detection method and device and electronic equipment | |
| CN112016387A (en) | Contraband identification method and device suitable for millimeter wave security check instrument | |
| CN116002480A (en) | Automatic detection method and system for accidental falling of passengers in elevator car | |
| CN112308010B (en) | Ship shielding detection method and device based on YOLO-V3 algorithm | |
| CN115239663A (en) | Method and system for detecting defects of contact lens, electronic device and storage medium | |
| CN114898379A (en) | Method, device and equipment for recognizing curved text and storage medium | |
| CN114140681A (en) | Remote dark and weak fixed star target detection method and device | |
| CN112967223A (en) | Artificial intelligence-based textile detection system, method and 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 |