CN116493282A - Rapid single-product silicon wafer sorting system and method based on machine vision - Google Patents
Rapid single-product silicon wafer sorting system and method based on machine vision Download PDFInfo
- Publication number
- CN116493282A CN116493282A CN202310382245.4A CN202310382245A CN116493282A CN 116493282 A CN116493282 A CN 116493282A CN 202310382245 A CN202310382245 A CN 202310382245A CN 116493282 A CN116493282 A CN 116493282A
- Authority
- CN
- China
- Prior art keywords
- sorting
- silicon wafer
- experience
- preset
- speaking
- 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
- XUIMIQQOPSSXEZ-UHFFFAOYSA-N Silicon Chemical compound [Si] XUIMIQQOPSSXEZ-UHFFFAOYSA-N 0.000 title claims abstract description 285
- 229910052710 silicon Inorganic materials 0.000 title claims abstract description 285
- 239000010703 silicon Substances 0.000 title claims abstract description 285
- 238000000034 method Methods 0.000 title claims abstract description 16
- 235000012431 wafers Nutrition 0.000 claims abstract description 218
- 238000010276 construction Methods 0.000 claims abstract description 16
- 238000005192 partition Methods 0.000 claims description 19
- 238000004458 analytical method Methods 0.000 claims description 18
- 238000012795 verification Methods 0.000 claims description 12
- 238000005516 engineering process Methods 0.000 claims description 10
- 238000003062 neural network model Methods 0.000 claims description 7
- 238000000605 extraction Methods 0.000 claims description 6
- 238000012549 training Methods 0.000 claims description 6
- 239000000284 extract Substances 0.000 claims description 3
- 230000009286 beneficial effect Effects 0.000 description 6
- 239000000047 product Substances 0.000 description 4
- 238000010586 diagram Methods 0.000 description 2
- 238000004519 manufacturing process Methods 0.000 description 2
- 238000012986 modification Methods 0.000 description 2
- 230000004048 modification Effects 0.000 description 2
- 239000013589 supplement Substances 0.000 description 2
- 230000004075 alteration Effects 0.000 description 1
- 238000001514 detection method Methods 0.000 description 1
- 239000002245 particle Substances 0.000 description 1
- 238000012546 transfer Methods 0.000 description 1
Classifications
-
- B—PERFORMING OPERATIONS; TRANSPORTING
- B07—SEPARATING SOLIDS FROM SOLIDS; SORTING
- B07C—POSTAL SORTING; SORTING INDIVIDUAL ARTICLES, OR BULK MATERIAL FIT TO BE SORTED PIECE-MEAL, e.g. BY PICKING
- B07C5/00—Sorting according to a characteristic or feature of the articles or material being sorted, e.g. by control effected by devices which detect or measure such characteristic or feature; Sorting by manually actuated devices, e.g. switches
- B07C5/34—Sorting according to other particular properties
-
- B—PERFORMING OPERATIONS; TRANSPORTING
- B07—SEPARATING SOLIDS FROM SOLIDS; SORTING
- B07C—POSTAL SORTING; SORTING INDIVIDUAL ARTICLES, OR BULK MATERIAL FIT TO BE SORTED PIECE-MEAL, e.g. BY PICKING
- B07C5/00—Sorting according to a characteristic or feature of the articles or material being sorted, e.g. by control effected by devices which detect or measure such characteristic or feature; Sorting by manually actuated devices, e.g. switches
- B07C5/02—Measures preceding sorting, e.g. arranging articles in a stream orientating
-
- B—PERFORMING OPERATIONS; TRANSPORTING
- B07—SEPARATING SOLIDS FROM SOLIDS; SORTING
- B07C—POSTAL SORTING; SORTING INDIVIDUAL ARTICLES, OR BULK MATERIAL FIT TO BE SORTED PIECE-MEAL, e.g. BY PICKING
- B07C5/00—Sorting according to a characteristic or feature of the articles or material being sorted, e.g. by control effected by devices which detect or measure such characteristic or feature; Sorting by manually actuated devices, e.g. switches
- B07C5/36—Sorting apparatus characterised by the means used for distribution
- B07C5/361—Processing or control devices therefor, e.g. escort memory
-
- 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
- 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
- G06V10/82—Arrangements for image or video recognition or understanding using pattern recognition or machine learning using neural networks
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
- Y02P—CLIMATE CHANGE MITIGATION TECHNOLOGIES IN THE PRODUCTION OR PROCESSING OF GOODS
- Y02P90/00—Enabling technologies with a potential contribution to greenhouse gas [GHG] emissions mitigation
- Y02P90/30—Computing systems specially adapted for manufacturing
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Evolutionary Computation (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Computing Systems (AREA)
- Databases & Information Systems (AREA)
- Health & Medical Sciences (AREA)
- General Health & Medical Sciences (AREA)
- Medical Informatics (AREA)
- Software Systems (AREA)
- Artificial Intelligence (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Multimedia (AREA)
- Crystals, And After-Treatments Of Crystals (AREA)
Abstract
The invention provides a machine vision-based single-wafer rapid sorting system and a machine vision-based single-wafer rapid sorting method, wherein the machine vision-based single-wafer rapid sorting system comprises: the silicon wafer sorting experience library construction module is used for constructing a silicon wafer sorting experience library and continuously updating the silicon wafer sorting experience library; the silicon wafer image acquisition module is used for acquiring a first silicon wafer image of the silicon wafer to be sorted; the sorting type determining module is used for determining the sorting type of the silicon wafer based on the first silicon wafer image and the silicon wafer sorting experience library; and the sorting module is used for sorting the silicon wafers based on the sorting type. According to the system and the method for rapidly sorting the single-product silicon wafers based on the machine vision, sorting staff is not required to sort the silicon wafers based on the shot silicon wafer images, so that the labor cost is greatly reduced, and in addition, the problem that sorting staff may cause silicon wafer sorting errors due to fatigue, boring and the like is further avoided.
Description
Technical Field
The invention relates to the technical field of silicon wafer detection, in particular to a system and a method for rapidly sorting single-product silicon wafers based on machine vision.
Background
At present, in the production and warehouse-out process of silicon wafers, the silicon wafers need to be sorted. Normally, silicon wafer sorting is completed in a manual semi-automatic mode, silicon wafers to be sorted are fed into a shooting position, a camera shoots the silicon wafers located in the shooting position, silicon wafer images are returned to a display screen in front of sorting staff, and sorting staff checks the silicon wafer images and then selects sorting types on an operation interface of the display screen. However, there are two problems to do so: firstly, the production amount of the silicon wafer is generally large, the required sorting personnel are much, and the labor cost is high; secondly, the efforts of sorting staff are limited, and as the sorting time length is increased and the repeatability of the sorting work is high, the problem of wrong sorting of the silicon wafers can occur due to fatigue, boring and the like. For this, a solution is needed.
Disclosure of Invention
The invention aims to provide a machine vision-based single-product silicon wafer rapid sorting system, which does not need sorting staff to sort silicon wafers based on shot silicon wafer images, so that the labor cost is greatly reduced, and in addition, the problem that sorting staff may cause silicon wafer sorting errors due to fatigue, boring and the like is further avoided.
The embodiment of the invention provides a machine vision-based single-product silicon wafer rapid sorting system, which comprises the following components:
the silicon wafer sorting experience library construction module is used for constructing a silicon wafer sorting experience library and continuously updating the silicon wafer sorting experience library;
the silicon wafer image acquisition module is used for acquiring a first silicon wafer image of the silicon wafer to be sorted;
the sorting type determining module is used for determining a first sorting type of the silicon wafer based on the first silicon wafer image and the silicon wafer sorting experience library;
and the sorting module is used for sorting the silicon wafers based on the first sorting type.
Preferably, the silicon wafer sorting experience library construction module constructs a silicon wafer sorting experience library, which comprises:
acquiring a second silicon wafer image and a second sorting type uploaded by a first expert person;
acquiring a preset first image feature searching template corresponding to the second sorting type;
searching for first image features representing a second classification type from the second silicon wafer image based on the first image feature search template;
pairing the first image features with the second sorting types and then using the paired first image features and the second sorting types as silicon chip sorting experience for warehousing;
and/or the number of the groups of groups,
extracting silicon wafer sorting experience description shared by a second expert from a preset silicon wafer sorting experience sharing scene;
Determining that a second expert person shares a third sort type involved and a second image feature representing the third sort type based on the silicon wafer sort experience description;
and pairing the second image characteristic with the third sorting type, and then warehousing as silicon chip sorting experience.
Preferably, the silicon wafer sorting experience library construction module extracts a silicon wafer sorting experience description shared by a second expert from a preset silicon wafer sorting experience sharing scene, and the method comprises the following steps:
verifying the credibility of the silicon chip sorting experience sharing scene;
when the verification is passed, acquiring speaking content and speaking time of a second expert in a silicon chip sorting experience sharing scene;
distributing the speaking content on a preset time axis based on speaking time;
determining a speech content cluster from the time axis; the speaking content clusters are formed by arranging a plurality of speaking contents, and the time distance between every two adjacent speaking contents in the speaking content clusters is smaller than or equal to a preset time distance threshold;
combining the speaking content clusters which meet the speaking content cluster combining conditions on the time axis and using the speaking content clusters as silicon chip sorting experience description;
the rest uncombined speaking content clusters are used as silicon chip sorting experience description;
The speech content cluster combining conditions include:
the minimum distance between the two speaking content clusters is within a preset first distance interval;
the maximum distance between the two speaking content clusters falls within a preset second distance interval;
the difference of the number of the speaking contents in the previous speaking content cluster and the speaking contents in the next speaking content cluster in the two speaking content clusters is larger than or equal to a preset number difference threshold;
the forefront one of the two speaking content clusters is matched and matched with one standard speaking content in a preset sharing preamble library.
Preferably, the silicon chip sorting experience library construction module verifies the credibility of the silicon chip sorting experience sharing scene, and the method comprises the following steps:
determining scene partitions in which the second expert person speaks from the silicon slice sorting experience sharing scene;
determining whether browsing user comments exist in the scene partition;
if not, acquiring the browsing progress of the browsing scene partition of the browsing user, and if the browsing progress is greater than or equal to a preset browsing progress threshold, determining whether the corresponding browsing user has comment permission, if so, counting by a preset first weight;
if yes, carrying out semantic analysis on the browsed user comments based on a semantic analysis technology to obtain comment semantics;
Matching comment semantics with objection semantics in a preset objection semantics library;
when the matching coincidence is achieved, a preset second weight and a preset third weight corresponding to the matched coincidence objection semantics are obtained, and browsing user comments corresponding to the matched coincidence comment semantics are used as target browsing user comments;
counting with a second weight when the target browsing user comments meet the first comment condition;
counting with a third weight when the target browsing user comments meet the second comment condition;
accumulating the sum of all the counts to obtain the credibility;
when the credibility is larger than or equal to a preset credibility threshold, passing verification;
wherein the first comment condition includes:
the target browsing user comments are replied by the second expert personnel;
the final reply of the browsing user who publishes the target browsing user comment to the second expert personnel is replied;
the second expert person browses the final reply of the browsing user who publishes the target browsing user comment but does not reply;
wherein the second comment condition includes:
the target browsing user comments are replied by the second expert personnel;
the final reply of the browsing user who publishes the target browsing user comment to the second expert person is browsed but not replied;
Wherein the second weight is less than the first weight and less than the third weight.
Preferably, the silicon wafer sorting experience library construction module determines, based on the silicon wafer sorting experience description, that the second expert personnel share the third sorting type involved and the second image feature representing the third sorting type, including:
based on semantic analysis technology, performing semantic analysis on silicon chip sorting experience description to obtain description semantics;
acquiring a preset sorting type determining library; the sort type determination library includes: a plurality of groups of first standard description semantics and type items which are in one-to-one correspondence;
matching the description semantics with any first standard description semantics;
when the matching is met, the type item corresponding to the first standard description semantics of the matching is used as a third sorting type;
determining whether a third silicon wafer image exists in the silicon wafer sorting experience description;
if yes, acquiring a preset second image feature searching template corresponding to the third sorting type, and searching for second image features from the third silicon wafer image based on the second image feature searching template;
if not, acquiring a preset image feature determination library; the image feature determination library includes: a plurality of groups of second standard description semantics and characteristic items which are in one-to-one correspondence;
Matching the description semantics which are not matched with the first standard description semantics with any second standard description semantics;
and when the matching is met, the feature item corresponding to the second standard description semantics of the matching is taken as the second image feature.
Preferably, the sorting type determining module determines a first sorting type of the silicon wafer based on the first silicon wafer image and a silicon wafer sorting experience library, including:
determining a template based on a preset experience application rule, and generating an experience application rule according to silicon wafer sorting experience in a silicon wafer sorting experience library;
performing feature extraction on the first silicon wafer image based on a preset feature extraction template to obtain a third image feature;
executing an empirical application rule on the third image feature to determine a first sort type;
and/or the number of the groups of groups,
training the neural network model by taking the silicon chip sorting experience library as a training sample until convergence;
inputting the first silicon chip image into a neural network model, and determining a first sorting type.
The machine vision-based single-wafer rapid sorting method provided by the embodiment of the invention is characterized by comprising the following steps of:
step S1: building a silicon wafer sorting experience library and continuously updating;
step S2: acquiring a first silicon wafer image of a silicon wafer to be sorted;
Step S3: determining a first sorting type of the silicon wafer based on the first silicon wafer image and a silicon wafer sorting experience library;
step S4: the silicon wafers are sorted based on the first sort type.
Preferably, step S1: building a silicon wafer sorting experience library, which comprises the following steps:
acquiring a second silicon wafer image and a second sorting type uploaded by a first expert person;
acquiring a preset first image feature searching template corresponding to the second sorting type;
searching for first image features representing a second classification type from the second silicon wafer image based on the first image feature search template;
pairing the first image features with the second sorting types and then using the paired first image features and the second sorting types as silicon chip sorting experience for warehousing;
and/or the number of the groups of groups,
extracting silicon wafer sorting experience description shared by a second expert from a preset silicon wafer sorting experience sharing scene;
determining that a second expert person shares a third sort type involved and a second image feature representing the third sort type based on the silicon wafer sort experience description;
and pairing the second image characteristic with the third sorting type, and then warehousing as silicon chip sorting experience.
Preferably, the extracting the silicon slice sorting experience description shared by the second expert from the preset silicon slice sorting experience sharing scene comprises the following steps:
Verifying the credibility of the silicon chip sorting experience sharing scene;
when the verification is passed, acquiring speaking content and speaking time of a second expert in a silicon chip sorting experience sharing scene;
distributing the speaking content on a preset time axis based on speaking time;
determining a speech content cluster from the time axis; the speaking content clusters are formed by arranging a plurality of speaking contents, and the time distance between every two adjacent speaking contents in the speaking content clusters is smaller than or equal to a preset time distance threshold;
combining the speaking content clusters which meet the speaking content cluster combining conditions on the time axis and using the speaking content clusters as silicon chip sorting experience description;
the rest uncombined speaking content clusters are used as silicon chip sorting experience description;
the speech content cluster combining conditions include:
the minimum distance between the two speaking content clusters is within a preset first distance interval;
the maximum distance between the two speaking content clusters falls within a preset second distance interval;
the difference of the number of the speaking contents in the previous speaking content cluster and the speaking contents in the next speaking content cluster in the two speaking content clusters is larger than or equal to a preset number difference threshold;
the forefront one of the two speaking content clusters is matched and matched with one standard speaking content in a preset sharing preamble library.
Preferably, verifying the credibility of the silicon chip sorting experience sharing scene comprises the following steps:
determining scene partitions in which the second expert person speaks from the silicon slice sorting experience sharing scene;
determining whether browsing user comments exist in the scene partition;
if not, acquiring the browsing progress of the browsing scene partition of the browsing user, and if the browsing progress is greater than or equal to a preset browsing progress threshold, determining whether the corresponding browsing user has comment permission, if so, counting by a preset first weight;
if yes, carrying out semantic analysis on the browsed user comments based on a semantic analysis technology to obtain comment semantics;
matching comment semantics with objection semantics in a preset objection semantics library;
when the matching coincidence is achieved, a preset second weight and a preset third weight corresponding to the matched coincidence objection semantics are obtained, and browsing user comments corresponding to the matched coincidence comment semantics are used as target browsing user comments;
counting with a second weight when the target browsing user comments meet the first comment condition;
counting with a third weight when the target browsing user comments meet the second comment condition;
accumulating the sum of all the counts to obtain the credibility;
When the credibility is larger than or equal to a preset credibility threshold, passing verification;
wherein the first comment condition includes:
the target browsing user comments are replied by the second expert personnel;
the final reply of the browsing user who publishes the target browsing user comment to the second expert personnel is replied;
the second expert person browses the final reply of the browsing user who publishes the target browsing user comment but does not reply;
wherein the second comment condition includes:
the target browsing user comments are replied by the second expert personnel;
the final reply of the browsing user who publishes the target browsing user comment to the second expert person is browsed but not replied;
wherein the second weight is less than the first weight and less than the third weight.
Additional features and advantages of the invention will be set forth in the description which follows, and in part will be obvious from the description, or may be learned by practice of the invention. The objectives and other advantages of the invention will be realized and attained by the structure particularly pointed out in the written description and claims thereof as well as the appended drawings.
The technical scheme of the invention is further described in detail through the drawings and the embodiments.
Drawings
The accompanying drawings are included to provide a further understanding of the invention and are incorporated in and constitute a part of this specification, illustrate the invention and together with the embodiments of the invention, serve to explain the invention. In the drawings:
FIG. 1 is a schematic diagram of a machine vision-based single wafer rapid sorting system in an embodiment of the invention;
fig. 2 is a schematic diagram of a machine vision-based method for rapidly sorting single silicon wafers in an embodiment of the invention.
Detailed Description
The preferred embodiments of the present invention will be described below with reference to the accompanying drawings, it being understood that the preferred embodiments described herein are for illustration and explanation of the present invention only, and are not intended to limit the present invention.
The embodiment of the invention provides a machine vision-based single-wafer rapid sorting system, which is shown in fig. 1 and comprises the following steps:
the silicon wafer sorting experience library construction module 1 is used for constructing a silicon wafer sorting experience library and continuously updating the silicon wafer sorting experience library; a great amount of silicon wafer sorting experience exists in the silicon wafer sorting experience library;
the silicon wafer image acquisition module 2 is used for acquiring a first silicon wafer image of a silicon wafer to be sorted;
the sorting type determining module 3 is used for determining a first sorting type of the silicon wafer based on the first silicon wafer image and a silicon wafer sorting experience library; the first sort type includes: whether hidden cracks exist, whether hard particles exist, whether dirt exists, whether broken edges exist, whether holes exist or not and the like;
And the sorting module 4 is used for sorting the silicon wafers based on the first sorting type.
The working principle and the beneficial effects of the technical scheme are as follows:
the silicon wafers to be sorted enter a shooting position, and a camera shoots the silicon wafers located at the shooting position to obtain a first silicon wafer image. The system of the application rapidly determines the first sorting type of the silicon wafer according to the first silicon wafer image based on the silicon wafer sorting experience in the silicon wafer sorting experience library which is built in advance, and sorts the silicon wafer based on the first sorting type. The method has the advantages that the sorting personnel are not required to sort the silicon wafers based on the shot silicon wafer images, so that the labor cost is greatly reduced, and in addition, the problem that the sorting personnel can sort the silicon wafers by mistake due to fatigue, boring and the like is avoided.
In one embodiment, the silicon wafer sorting experience library construction module 1 constructs a silicon wafer sorting experience library, including:
acquiring a second silicon wafer image and a second sorting type uploaded by a first expert person; the first expert personnel can be personnel, silicon wafer researchers and the like who are historically engaged in silicon wafer sorting, and can upload second silicon wafer images and corresponding second sorting types which are historically used for silicon wafer sorting;
Acquiring a preset first image feature searching template corresponding to the second sorting type; the first image feature search template is an execution template that searches for image features from the image that are related to the second classification type, such as: the second sorting type is dirt, and the first image feature searching template is an executing template for extracting dirt image features from the image;
searching for first image features representing a second classification type from the second silicon wafer image based on the first image feature search template;
pairing the first image features with the second sorting types and then using the paired first image features and the second sorting types as silicon chip sorting experience for warehousing;
and/or the number of the groups of groups,
extracting silicon wafer sorting experience description shared by a second expert from a preset silicon wafer sorting experience sharing scene; the silicon chip sorting experience sharing scene can be a forum, a bar, and the like for communicating among personnel, silicon chip researchers, and the like who are historically engaged in silicon chip sorting;
determining that a second expert person shares a third sort type involved and a second image feature representing the third sort type based on the silicon wafer sort experience description;
and pairing the second image characteristic with the third sorting type, and then warehousing as silicon chip sorting experience.
The working principle and the beneficial effects of the technical scheme are as follows:
And two ways are introduced to build a silicon wafer sorting experience library, so that the comprehensiveness of the silicon wafer sorting experience library is improved. Generally, the staff related to the silicon wafer have experience sharing scenes such as internal forums, so that the second mode has high applicability. In addition, the silicon wafer sorting experience library in the application can select the silicon wafer model and the like suitable for the silicon wafer sorting experience according to actual needs.
In one embodiment, the silicon wafer sorting experience library construction module 1 extracts a silicon wafer sorting experience description shared by a second expert from a preset silicon wafer sorting experience sharing scene, including:
verifying the credibility of the silicon chip sorting experience sharing scene;
when the verification is passed, acquiring speaking content and speaking time of a second expert in a silicon chip sorting experience sharing scene;
distributing the speaking content on a preset time axis based on speaking time; the speaking time corresponds to a time point on a time axis when the setting is performed;
determining a speech content cluster from the time axis; the speaking content clusters are formed by arranging a plurality of speaking contents, and the time distance between every two adjacent speaking contents in the speaking content clusters is smaller than or equal to a preset time distance threshold;
combining the speaking content clusters which meet the speaking content cluster combining conditions on the time axis and using the speaking content clusters as silicon chip sorting experience description;
The rest uncombined speaking content clusters are used as silicon chip sorting experience description;
the speech content cluster combining conditions include:
the minimum distance between the two speaking content clusters is within a preset first distance interval; the speaking content cluster has left and right endpoints, and the minimum distance is the distance between the right endpoint of the previous speaking content cluster and the left endpoint of the next speaking content cluster on the time axis;
the maximum distance between the two speaking content clusters falls within a preset second distance interval; the minimum distance is the distance between the left end point of the previous speaking content cluster and the right end point of the latter speaking content cluster on the time axis;
the difference of the number of the speaking contents in the previous speaking content cluster and the speaking contents in the next speaking content cluster in the two speaking content clusters is larger than or equal to a preset number difference threshold; the previous talk content cluster is the talk content cluster to the left on the time axis;
the forefront one of the two speaking content clusters is matched and matched with one standard speaking content in a preset sharing preamble library. The standard speaking content is the preamble content shared by silicon chip sorting experience, for example: "give your a broken silicon wafer" etc.
The working principle and the beneficial effects of the technical scheme are as follows:
generally, when a second expert person shares a silicon wafer sharing experience in a silicon wafer sorting experience sharing scenario, speaking (such as introduction, graph transfer, feature introduction, etc.) is continuously performed for a certain period of time, so as to form a speaking content cluster, and speaking content in this period of time is only a case of a silicon wafer sorting type. However, a special situation occurs when the second expert personnel share the silicon chip sharing experience, and after the sharing of one stage is finished, the second expert personnel supplements the silicon chip sharing experience after a short time, and the supplement content is also related to the silicon chip sorting type case corresponding to the previous sharing and needs to be combined with the content shared before. Normally, the new supplementary speaking content is continuous in a certain time, speaking content clusters are formed, the distance between the front speaking content clusters corresponding to the previous shared silicon chip sorting type cases is not too far or too close, the number of the speaking content clusters formed by the new supplementary speaking content is smaller than that of the front speaking content clusters corresponding to the previous shared silicon chip sorting type cases, and the beginning speaking content of the front speaking content clusters corresponding to the previous shared silicon chip sorting type cases is an introduction of silicon chip sorting experience sharing. In the conventional way, whether the utterances in the two utterances content clusters are related to each other or not needs to be analyzed based on a semantic understanding technology, and if yes, the utterances are combined.
In one embodiment, the silicon wafer sorting experience library construction module 1 verifies the credibility of the silicon wafer sorting experience sharing scenario, including:
determining scene partitions in which the second expert person speaks from the silicon slice sorting experience sharing scene; the scene partition can be a sub-patch in the forum;
determining whether browsing user comments exist in the scene partition;
if not, acquiring the browsing progress of the browsing scene partition of the browsing user, and if the browsing progress is greater than or equal to a preset browsing progress threshold, determining whether the corresponding browsing user has comment permission, if so, counting by a preset first weight; the browsing progress is the page turning and viewing ratio of the split paste when browsing by a browsing user, for example: turning pages from top to bottom, wherein the browsing progress is 100%; the browsing progress is larger, the browsing user is shown to read the sharing of the second expert personnel, the browsing user has comment permission but does not comment, and the large probability of sharing the second expert personnel is shown to be agreed;
if yes, carrying out semantic analysis on the browsed user comments based on a semantic analysis technology to obtain comment semantics;
matching comment semantics with objection semantics in a preset objection semantics library; the objection semantics are semantics of generating objection for sharing silicon wafer sorting experience of a second expert, for example: "this is not a broken edge bar? ";
When the matching coincidence is achieved, a preset second weight and a preset third weight corresponding to the matched coincidence objection semantics are obtained, and browsing user comments corresponding to the matched coincidence comment semantics are used as target browsing user comments;
counting with a second weight when the target browsing user comments meet the first comment condition;
counting with a third weight when the target browsing user comments meet the second comment condition;
accumulating the sum of all the counts to obtain the credibility;
when the credibility is larger than or equal to a preset credibility threshold, passing verification;
wherein the first comment condition includes:
the target browsing user comments are replied by the second expert personnel;
the final reply of the browsing user who publishes the target browsing user comment to the second expert personnel is replied;
the second expert person browses the final reply of the browsing user who publishes the target browsing user comment but does not reply; the browsing user generates objection, the second expert gives a reply, the browsing user also replies, the second expert does not reply to the browsing user finally but browses the final reply of the browsing user, and the silicon chip sorting experience of the second expert shares problems with high probability;
Wherein the second comment condition includes:
the target browsing user comments are replied by the second expert personnel;
the final reply of the browsing user who publishes the target browsing user comment to the second expert person is browsed but not replied; the browse user generates objection, the second expert gives a reply, the browse user does not reply finally but browses the final reply of the second expert, and the silicon slice sorting experience of the second expert is proved to have high probability and have no problem and can withstand the push;
wherein the second weight is less than the first weight and less than the third weight.
The working principle and the beneficial effects of the technical scheme are as follows:
the reliability of the silicon chip sorting experience sharing scene is verified, and when the silicon chip sorting experience description is verified, the silicon chip sorting experience description is extracted from the silicon chip sorting experience description, so that the accuracy and the reliability of the silicon chip sorting experience description are improved. In addition, when the credibility of the silicon chip sorting experience sharing scene is obtained, the first weight, the second weight and the third weight are respectively introduced from three different practical situations, so that the credibility obtaining accuracy is improved.
In one embodiment, the silicon wafer sort experience library construction module 1 determines, based on the silicon wafer sort experience description, that a second expert person shares a third sort type involved and a second image feature representative of the third sort type, comprising:
Based on semantic analysis technology, performing semantic analysis on silicon chip sorting experience description to obtain description semantics;
acquiring a preset sorting type determining library; the sort type determination library includes: a plurality of groups of first standard description semantics and type items which are in one-to-one correspondence; both are specific, for example: the first standard description semantics are that the silicon wafer is broken, and the type item is broken;
matching the description semantics with any first standard description semantics;
when the matching is met, the type item corresponding to the first standard description semantics of the matching is used as a third sorting type;
determining whether a third silicon wafer image exists in the silicon wafer sorting experience description;
if yes, acquiring a preset second image feature searching template corresponding to the third sorting type, and searching for second image features from the third silicon wafer image based on the second image feature searching template; the second image feature search template is the same as the first image feature search template, and will not be described in detail;
if not, acquiring a preset image feature determination library; the image feature determination library includes: a plurality of groups of second standard description semantics and characteristic items which are in one-to-one correspondence; both are specific, for example: the second standard description semantics is that the feature item is the image feature of the edge collapse of the silicon wafer when the edge collapse occurs on the silicon wafer;
Matching the description semantics which are not matched with the first standard description semantics with any second standard description semantics;
and when the matching is met, the feature item corresponding to the second standard description semantics of the matching is taken as the second image feature.
The working principle and the beneficial effects of the technical scheme are as follows:
according to the embodiment of the invention, based on the silicon chip sorting experience description, a second expert person is determined to share the related third sorting type and the second image characteristic representing the third sorting type, firstly, the third sorting type is determined, a sorting type determining library is introduced, the accuracy and the determining efficiency of sorting type determination are improved, secondly, if the third silicon chip image exists in the description, the second image characteristic is directly extracted from the image, otherwise, the second image characteristic is determined based on the image characteristic determining library, and the applicability of the system is improved.
In one embodiment, the sorting type determining module 3 determines a first sorting type of the silicon wafer based on the first silicon wafer image and the silicon wafer sorting experience library, comprising:
determining a template based on a preset experience application rule, and generating an experience application rule according to silicon wafer sorting experience in a silicon wafer sorting experience library; specifically, the empirically applied rule determination template is, for example: determining whether the image characteristics of the silicon wafer image are matched with the image characteristics in the silicon wafer sorting experience, if so, outputting the sorting type in the silicon wafer sorting experience;
Performing feature extraction on the first silicon wafer image based on a preset feature extraction template to obtain a third image feature; the third image feature is an image feature which can be used as a silicon chip sorting basis in the first silicon chip image;
executing an empirical application rule on the third image feature to determine a first sort type;
and/or the number of the groups of groups,
training the neural network model by taking the silicon chip sorting experience library as a training sample until convergence; after the neural network model is trained to be converged, the silicon chip can be directly subjected to sorting type determination based on the silicon chip image;
inputting the first silicon chip image into a neural network model, and determining a first sorting type.
The working principle and the beneficial effects of the technical scheme are as follows:
the two ways are introduced to determine the first sorting type of the silicon wafer based on the first silicon wafer image and the silicon wafer sorting experience library, so that the applicability of the system is improved.
The embodiment of the invention provides a machine vision-based single-wafer rapid sorting method, which is shown in fig. 2 and comprises the following steps:
step S1: building a silicon wafer sorting experience library and continuously updating;
step S2: acquiring a first silicon wafer image of a silicon wafer to be sorted;
step S3: determining a first sorting type of the silicon wafer based on the first silicon wafer image and a silicon wafer sorting experience library;
Step S4: the silicon wafers are sorted based on the first sort type.
In one embodiment, step S1: building a silicon wafer sorting experience library, which comprises the following steps:
acquiring a second silicon wafer image and a second sorting type uploaded by a first expert person;
acquiring a preset first image feature searching template corresponding to the second sorting type;
searching for first image features representing a second classification type from the second silicon wafer image based on the first image feature search template;
pairing the first image features with the second sorting types and then using the paired first image features and the second sorting types as silicon chip sorting experience for warehousing;
and/or the number of the groups of groups,
extracting silicon wafer sorting experience description shared by a second expert from a preset silicon wafer sorting experience sharing scene;
determining that a second expert person shares a third sort type involved and a second image feature representing the third sort type based on the silicon wafer sort experience description;
and pairing the second image characteristic with the third sorting type, and then warehousing as silicon chip sorting experience.
In one embodiment, extracting the silicon wafer sorting experience description shared by the second expert from the preset silicon wafer sorting experience sharing scene comprises the following steps:
verifying the credibility of the silicon chip sorting experience sharing scene;
When the verification is passed, acquiring speaking content and speaking time of a second expert in a silicon chip sorting experience sharing scene;
distributing the speaking content on a preset time axis based on speaking time;
determining a speech content cluster from the time axis; the speaking content clusters are formed by arranging a plurality of speaking contents, and the time distance between every two adjacent speaking contents in the speaking content clusters is smaller than or equal to a preset time distance threshold;
combining the speaking content clusters which meet the speaking content cluster combining conditions on the time axis and using the speaking content clusters as silicon chip sorting experience description;
the rest uncombined speaking content clusters are used as silicon chip sorting experience description;
the speech content cluster combining conditions include:
the minimum distance between the two speaking content clusters is within a preset first distance interval;
the maximum distance between the two speaking content clusters falls within a preset second distance interval;
the difference of the number of the speaking contents in the previous speaking content cluster and the speaking contents in the next speaking content cluster in the two speaking content clusters is larger than or equal to a preset number difference threshold;
the forefront one of the two speaking content clusters is matched and matched with one standard speaking content in a preset sharing preamble library.
In one embodiment, verifying the trustworthiness of a silicon wafer sorting experience sharing scenario comprises:
determining scene partitions in which the second expert person speaks from the silicon slice sorting experience sharing scene;
determining whether browsing user comments exist in the scene partition;
if not, acquiring the browsing progress of the browsing scene partition of the browsing user, and if the browsing progress is greater than or equal to a preset browsing progress threshold, determining whether the corresponding browsing user has comment permission, if so, counting by a preset first weight;
if yes, carrying out semantic analysis on the browsed user comments based on a semantic analysis technology to obtain comment semantics;
matching comment semantics with objection semantics in a preset objection semantics library;
when the matching coincidence is achieved, a preset second weight and a preset third weight corresponding to the matched coincidence objection semantics are obtained, and browsing user comments corresponding to the matched coincidence comment semantics are used as target browsing user comments;
counting with a second weight when the target browsing user comments meet the first comment condition;
counting with a third weight when the target browsing user comments meet the second comment condition;
Accumulating the sum of all the counts to obtain the credibility;
when the credibility is larger than or equal to a preset credibility threshold, passing verification;
wherein the first comment condition includes:
the target browsing user comments are replied by the second expert personnel;
the final reply of the browsing user who publishes the target browsing user comment to the second expert personnel is replied;
the second expert person browses the final reply of the browsing user who publishes the target browsing user comment but does not reply;
wherein the second comment condition includes:
the target browsing user comments are replied by the second expert personnel;
the final reply of the browsing user who publishes the target browsing user comment to the second expert person is browsed but not replied;
wherein the second weight is less than the first weight and less than the third weight.
It will be apparent to those skilled in the art that various modifications and variations can be made to the present invention without departing from the spirit or scope of the invention. Thus, it is intended that the present invention also include such modifications and alterations insofar as they come within the scope of the appended claims or the equivalents thereof.
Claims (10)
1. Quick sorting system of single-product silicon chip based on machine vision, its characterized in that includes:
The silicon wafer sorting experience library construction module is used for constructing a silicon wafer sorting experience library and continuously updating the silicon wafer sorting experience library;
the silicon wafer image acquisition module is used for acquiring a first silicon wafer image of the silicon wafer to be sorted;
the sorting type determining module is used for determining a first sorting type of the silicon wafer based on the first silicon wafer image and the silicon wafer sorting experience library;
and the sorting module is used for sorting the silicon wafers based on the first sorting type.
2. The machine vision-based single-wafer rapid sorting system according to claim 1, wherein the wafer sorting experience library construction module constructs a wafer sorting experience library, comprising:
acquiring a second silicon wafer image and a second sorting type uploaded by a first expert person;
acquiring a preset first image feature search template corresponding to the second sorting type;
searching for first image features representing the second classification type from the second silicon wafer image based on the first image feature search template;
pairing the first image features with the second sorting types and then using the paired first image features and the second sorting types as silicon chip sorting experience for warehousing;
and/or the number of the groups of groups,
extracting silicon wafer sorting experience description shared by a second expert from a preset silicon wafer sorting experience sharing scene;
Determining, based on the silicon wafer sort experience description, that the second expert person shares a third sort type involved and a second image feature representative of the third sort type;
and pairing the second image characteristic with the third sorting type, and then warehousing as silicon chip sorting experience.
3. The machine vision-based single-wafer rapid sorting system according to claim 2, wherein the wafer sorting experience library construction module extracts a wafer sorting experience description shared by a second expert from a preset wafer sorting experience sharing scene, and comprises:
verifying the credibility of the silicon chip sorting experience sharing scene;
when the verification is passed, acquiring speaking contents and speaking time of the second expert in the silicon chip sorting experience sharing scene;
distributing the speaking content on a preset time axis based on the speaking time;
determining a speech content cluster from the time axis; the speaking content clusters are formed by arranging a plurality of speaking contents, and the time distance between every two adjacent speaking contents in the speaking content clusters is smaller than or equal to a preset time distance threshold;
Combining the speaking content clusters which meet the speaking content cluster combining conditions on the time axis as the silicon wafer sorting experience description;
the residual uncombined speaking content clusters are used as the silicon chip sorting experience description;
wherein, the speaking content cluster combining condition comprises:
the minimum distance between the two speaking content clusters is within a preset first distance interval;
the maximum distance between the two speaking content clusters falls within a preset second distance interval;
the difference of the number of the speaking contents in the previous speaking content cluster and the speaking contents in the next speaking content cluster in the two speaking content clusters is larger than or equal to a preset number difference threshold;
the forefront one of the two speaking content clusters is matched and matched with one standard speaking content in a preset sharing preamble library.
4. The machine vision-based single-wafer rapid sorting system according to claim 3, wherein the silicon wafer sorting experience library construction module verifies the credibility of the silicon wafer sorting experience sharing scene, comprising:
determining scene partitions in which the second expert person speaks from the silicon slice sorting experience sharing scene;
Determining whether browsing user comments exist in the scene partition;
if not, acquiring the browsing progress of the browsing user for browsing the scene partition, and if the browsing progress is greater than or equal to a preset browsing progress threshold, determining whether the corresponding browsing user has comment permission, if so, counting by a preset first weight;
if yes, carrying out semantic analysis on the browsed user comments based on a semantic analysis technology to obtain comment semantics;
matching the comment semantics with the objection semantics in a preset objection semantics library;
when the matching coincidence is the matching coincidence, acquiring a preset second weight and a preset third weight corresponding to the objection semantics of the matching coincidence, and taking the browsing user comment corresponding to the comment meaning of the matching coincidence as a target browsing user comment;
counting with the second weight when the target browsing user comments meet a first comment condition;
counting with the third weight when the target browsing user comment meets a second comment condition;
accumulating the sum of all the counts to obtain the credibility;
when the credibility is larger than or equal to a preset credibility threshold, passing verification;
Wherein the first comment condition includes:
the target browsing user comment is replied by the second expert;
the final reply of the browsing user who publishes the target browsing user comment to the second expert personnel is replied;
the second expert person has browsed but not replied to the final reply of the browsing user who posted the target browsing user comment;
wherein the second comment condition includes:
the target browsing user comment is replied by the second expert;
the final reply of the browsing user who publishes the target browsing user comment to the second expert person is browsed but not replied;
wherein the second weight is less than the first weight and less than the third weight.
5. The machine vision based single wafer rapid sorting system of claim 2, wherein the wafer sorting experience library construction module determines that the second expert person shares a third sort type involved and a second image feature representative of the third sort type based on the wafer sorting experience description, comprising:
based on semantic analysis technology, carrying out semantic analysis on the silicon chip sorting experience description to obtain description semantics;
Acquiring a preset sorting type determining library; the sort type determination library includes: a plurality of groups of first standard description semantics and type items which are in one-to-one correspondence;
matching the description semantics with any one of the first standard description semantics;
when the matching is met, the type item corresponding to the first standard description semantics of the matching is taken as the third sorting type;
determining whether a third silicon wafer image exists in the silicon wafer sorting experience description;
if yes, acquiring a preset second image feature searching template corresponding to the third sorting type, and searching the second image feature from the third silicon wafer image based on the second image feature searching template;
if not, acquiring a preset image feature determination library; the image feature determination library includes: a plurality of groups of second standard description semantics and characteristic items which are in one-to-one correspondence;
matching the description semantics which are not matched with the first standard description semantics with any second standard description semantics;
and when the second standard description semantics are matched and met, the feature item corresponding to the second standard description semantics which are matched and met is used as the second image feature.
6. The machine vision based single wafer rapid sorting system of claim 1, wherein the sort type determination module determines a first sort type of the wafer based on the first wafer image and the library of wafer sort experiences, comprising:
determining a template based on a preset experience application rule, and generating an experience application rule according to silicon wafer sorting experiences in the silicon wafer sorting experience library;
performing feature extraction on the first silicon wafer image based on a preset feature extraction template to obtain a third image feature;
performing the empirical application rule on the third image feature, determining the first sort type;
and/or the number of the groups of groups,
training the neural network model by taking the silicon chip sorting experience library as a training sample until convergence;
and inputting the first silicon chip image into the neural network model, and determining the first sorting type.
7. The quick sorting method of the single-product silicon wafer based on the machine vision is characterized by comprising the following steps of:
step S1: building a silicon wafer sorting experience library and continuously updating;
step S2: acquiring a first silicon wafer image of a silicon wafer to be sorted;
step S3: determining a first sorting type of the silicon wafer based on the first silicon wafer image and the silicon wafer sorting experience library;
Step S4: and sorting the silicon wafers based on the first sorting type.
8. The machine vision-based rapid sorting method for single-wafer silicon according to claim 7, wherein the step S1: building a silicon wafer sorting experience library, which comprises the following steps:
acquiring a second silicon wafer image and a second sorting type uploaded by a first expert person;
acquiring a preset first image feature search template corresponding to the second sorting type;
searching for first image features representing the second classification type from the second silicon wafer image based on the first image feature search template;
pairing the first image features with the second sorting types and then using the paired first image features and the second sorting types as silicon chip sorting experience for warehousing;
and/or the number of the groups of groups,
extracting silicon wafer sorting experience description shared by a second expert from a preset silicon wafer sorting experience sharing scene;
determining, based on the silicon wafer sort experience description, that the second expert person shares a third sort type involved and a second image feature representative of the third sort type;
and pairing the second image characteristic with the third sorting type, and then warehousing as silicon chip sorting experience.
9. The machine vision based single wafer rapid sorting method according to claim 8, wherein the extracting the silicon wafer sorting experience description shared by the second expert from the preset silicon wafer sorting experience sharing scene comprises:
Verifying the credibility of the silicon chip sorting experience sharing scene;
when the verification is passed, acquiring speaking contents and speaking time of the second expert in the silicon chip sorting experience sharing scene;
distributing the speaking content on a preset time axis based on the speaking time;
determining a speech content cluster from the time axis; the speaking content clusters are formed by arranging a plurality of speaking contents, and the time distance between every two adjacent speaking contents in the speaking content clusters is smaller than or equal to a preset time distance threshold;
combining the speaking content clusters which meet the speaking content cluster combining conditions on the time axis as the silicon wafer sorting experience description;
the residual uncombined speaking content clusters are used as the silicon chip sorting experience description;
wherein, the speaking content cluster combining condition comprises:
the minimum distance between the two speaking content clusters is within a preset first distance interval;
the maximum distance between the two speaking content clusters falls within a preset second distance interval;
the difference of the number of the speaking contents in the previous speaking content cluster and the speaking contents in the next speaking content cluster in the two speaking content clusters is larger than or equal to a preset number difference threshold;
The forefront one of the two speaking content clusters is matched and matched with one standard speaking content in a preset sharing preamble library.
10. The machine vision based single-wafer rapid sorting method according to claim 9, wherein verifying the credibility of the wafer sorting experience sharing scenario comprises:
determining scene partitions in which the second expert person speaks from the silicon slice sorting experience sharing scene;
determining whether browsing user comments exist in the scene partition;
if not, acquiring the browsing progress of the browsing user for browsing the scene partition, and if the browsing progress is greater than or equal to a preset browsing progress threshold, determining whether the corresponding browsing user has comment permission, if so, counting by a preset first weight;
if yes, carrying out semantic analysis on the browsed user comments based on a semantic analysis technology to obtain comment semantics;
matching the comment semantics with the objection semantics in a preset objection semantics library;
when the matching coincidence is the matching coincidence, acquiring a preset second weight and a preset third weight corresponding to the objection semantics of the matching coincidence, and taking the browsing user comment corresponding to the comment meaning of the matching coincidence as a target browsing user comment;
Counting with the second weight when the target browsing user comments meet a first comment condition;
counting with the third weight when the target browsing user comment meets a second comment condition;
accumulating the sum of all the counts to obtain the credibility;
when the credibility is larger than or equal to a preset credibility threshold, passing verification;
wherein the first comment condition includes:
the target browsing user comment is replied by the second expert;
the final reply of the browsing user who publishes the target browsing user comment to the second expert personnel is replied;
the second expert person has browsed but not replied to the final reply of the browsing user who posted the target browsing user comment;
wherein the second comment condition includes:
the target browsing user comment is replied by the second expert;
the final reply of the browsing user who publishes the target browsing user comment to the second expert person is browsed but not replied;
wherein the second weight is less than the first weight and less than the third weight.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202310382245.4A CN116493282A (en) | 2023-04-12 | 2023-04-12 | Rapid single-product silicon wafer sorting system and method based on machine vision |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202310382245.4A CN116493282A (en) | 2023-04-12 | 2023-04-12 | Rapid single-product silicon wafer sorting system and method based on machine vision |
Publications (1)
Publication Number | Publication Date |
---|---|
CN116493282A true CN116493282A (en) | 2023-07-28 |
Family
ID=87317607
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN202310382245.4A Pending CN116493282A (en) | 2023-04-12 | 2023-04-12 | Rapid single-product silicon wafer sorting system and method based on machine vision |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN116493282A (en) |
Citations (10)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN101234381A (en) * | 2008-03-07 | 2008-08-06 | 天津市华核科技有限公司 | Granular material sorting classifying method based on visual sense recognition |
CN103487147A (en) * | 2013-09-06 | 2014-01-01 | 闻泰通讯股份有限公司 | System and method for picking fruits through electronic equipment |
CN110163144A (en) * | 2019-05-17 | 2019-08-23 | 成都信息工程大学 | A method of the automatic identification human body behavior from video |
CN110333242A (en) * | 2019-07-15 | 2019-10-15 | 武汉楚锐视觉检测科技有限公司 | A kind of detection method and its system of PET bottle quality |
CN110802791A (en) * | 2019-11-09 | 2020-02-18 | 台州市黄岩达意隆模具有限公司 | Mould pressing system lid machine on-line early warning monitoring system based on meshing image processing technology |
CN110825768A (en) * | 2019-10-10 | 2020-02-21 | 安徽康佳电子有限公司 | Remote television exception handling method and system based on cloud analysis |
CN111318470A (en) * | 2019-07-02 | 2020-06-23 | 王多 | Method and device for identifying article type |
CN114385789A (en) * | 2022-01-11 | 2022-04-22 | 国网安徽省电力有限公司宣城供电公司 | Semantic analysis-based substation protection pressing plate operation regulation analysis method and system |
CN115017396A (en) * | 2022-04-18 | 2022-09-06 | 山东省计算中心(国家超级计算济南中心) | Intelligent city knowledge graph construction method and system |
CN115452841A (en) * | 2022-10-19 | 2022-12-09 | 深圳市力琪新材料科技有限公司 | Plastic product delivery quality detection equipment and system |
-
2023
- 2023-04-12 CN CN202310382245.4A patent/CN116493282A/en active Pending
Patent Citations (10)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN101234381A (en) * | 2008-03-07 | 2008-08-06 | 天津市华核科技有限公司 | Granular material sorting classifying method based on visual sense recognition |
CN103487147A (en) * | 2013-09-06 | 2014-01-01 | 闻泰通讯股份有限公司 | System and method for picking fruits through electronic equipment |
CN110163144A (en) * | 2019-05-17 | 2019-08-23 | 成都信息工程大学 | A method of the automatic identification human body behavior from video |
CN111318470A (en) * | 2019-07-02 | 2020-06-23 | 王多 | Method and device for identifying article type |
CN110333242A (en) * | 2019-07-15 | 2019-10-15 | 武汉楚锐视觉检测科技有限公司 | A kind of detection method and its system of PET bottle quality |
CN110825768A (en) * | 2019-10-10 | 2020-02-21 | 安徽康佳电子有限公司 | Remote television exception handling method and system based on cloud analysis |
CN110802791A (en) * | 2019-11-09 | 2020-02-18 | 台州市黄岩达意隆模具有限公司 | Mould pressing system lid machine on-line early warning monitoring system based on meshing image processing technology |
CN114385789A (en) * | 2022-01-11 | 2022-04-22 | 国网安徽省电力有限公司宣城供电公司 | Semantic analysis-based substation protection pressing plate operation regulation analysis method and system |
CN115017396A (en) * | 2022-04-18 | 2022-09-06 | 山东省计算中心(国家超级计算济南中心) | Intelligent city knowledge graph construction method and system |
CN115452841A (en) * | 2022-10-19 | 2022-12-09 | 深圳市力琪新材料科技有限公司 | Plastic product delivery quality detection equipment and system |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN101021850B (en) | Word search apparatus, word search method | |
EP2124159B1 (en) | Image learning, automatic annotation, retrieval method, and device | |
CN102388392B (en) | Pattern recognition device | |
CN103366367A (en) | Pixel number clustering-based fuzzy C-average value gray level image splitting method | |
CN105095434B (en) | The recognition methods of timeliness demand and device | |
CN103838754B (en) | Information retrieval device and method | |
CN110210294A (en) | Evaluation method, device, storage medium and the computer equipment of Optimized model | |
CN107153670A (en) | The video retrieval method and system merged based on multiple image | |
CN109711298B (en) | Method and system for efficient face characteristic value retrieval based on faiss | |
CN107609105B (en) | Construction method of big data acceleration structure | |
CN111797846A (en) | Feedback type target detection method based on characteristic pyramid network | |
CN108595411B (en) | Method for acquiring multiple text abstracts in same subject text set | |
CN105989001A (en) | Image searching method and device, and image searching system | |
CN109885651A (en) | A kind of question pushing method and device | |
CN110443290A (en) | A kind of product competition relationship quantization generation method and device based on big data | |
CN104462347B (en) | The sorting technique and device of keyword | |
CN113254572B (en) | Electronic document classification supervision system based on cloud platform | |
CN110362620A (en) | A kind of list data structure method based on machine learning | |
WO2024082756A1 (en) | Video search method and apparatus, server, and terminal device | |
CN116493282A (en) | Rapid single-product silicon wafer sorting system and method based on machine vision | |
CN113743443A (en) | Image evidence classification and identification method and device | |
CN111382604B (en) | Face recognition method and system | |
CN110555219B (en) | Three-dimensional CAD model similarity retrieval system and method based on image recognition | |
CN102841988A (en) | System and method for matching nucleotide sequence information | |
CN113553947B (en) | Method and device for generating and describing multi-mode pedestrian re-recognition and electronic equipment |
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 |