CN113312525A - Method for reversely calibrating steel seal code through java - Google Patents
Method for reversely calibrating steel seal code through java Download PDFInfo
- Publication number
- CN113312525A CN113312525A CN202110631594.6A CN202110631594A CN113312525A CN 113312525 A CN113312525 A CN 113312525A CN 202110631594 A CN202110631594 A CN 202110631594A CN 113312525 A CN113312525 A CN 113312525A
- Authority
- CN
- China
- Prior art keywords
- code
- similarity
- python
- matching
- algorithm
- 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.)
- Granted
Links
- 229910000831 Steel Inorganic materials 0.000 title claims abstract description 29
- 239000010959 steel Substances 0.000 title claims abstract description 29
- 238000000034 method Methods 0.000 title claims abstract description 20
- 230000009286 beneficial effect Effects 0.000 abstract description 2
- 238000013473 artificial intelligence Methods 0.000 description 10
- 238000004364 calculation method Methods 0.000 description 5
- 230000006870 function Effects 0.000 description 2
- 238000004519 manufacturing process Methods 0.000 description 2
- 238000010586 diagram Methods 0.000 description 1
- 238000005516 engineering process Methods 0.000 description 1
- 230000007613 environmental effect Effects 0.000 description 1
- 238000009472 formulation Methods 0.000 description 1
- 230000003993 interaction Effects 0.000 description 1
- 239000000463 material Substances 0.000 description 1
- 239000000203 mixture Substances 0.000 description 1
- 230000017105 transposition Effects 0.000 description 1
- 238000012795 verification Methods 0.000 description 1
- 230000000007 visual effect Effects 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/90—Details of database functions independent of the retrieved data types
- G06F16/903—Querying
- G06F16/90335—Query processing
- G06F16/90344—Query processing by using string matching techniques
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/22—Matching criteria, e.g. proximity measures
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V30/00—Character recognition; Recognising digital ink; Document-oriented image-based pattern recognition
- G06V30/10—Character recognition
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Data Mining & Analysis (AREA)
- Databases & Information Systems (AREA)
- Physics & Mathematics (AREA)
- General Engineering & Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Life Sciences & Earth Sciences (AREA)
- Computational Linguistics (AREA)
- Artificial Intelligence (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Bioinformatics & Computational Biology (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Evolutionary Biology (AREA)
- Evolutionary Computation (AREA)
- Devices For Executing Special Programs (AREA)
- Image Analysis (AREA)
Abstract
A method for carrying out reverse calibration steel seal coding through java comprises the following steps: the java server receives the coded data and the identification times identified by the Python identification program, compares the character strings in the database with the identified coded data by a contiins method in the ArrayList of java, and returns false to the Python identification program if the coded data is not completely matched with the character strings in the database; the Python identification program identifies the second photo, then sends the code to the java server, and performs the operation of the step 2), until the photographed photos are completely matched and the java server compares the database and still does not find the completely same code, the PC terminal calls a similarity matching algorithm to perform similarity comparison matching on the identified code of the last photo and the database code; the invention has the beneficial effects that: the steel seal code is identified and calibrated through multiple times of identification of multiple photos and combination of four similarity matching algorithms, so that the accuracy rate of identifying the steel seal code is improved.
Description
Technical Field
The invention relates to the technical field of reverse algorithms, in particular to a method for reversely calibrating steel seal codes through java.
Background
With the advent of the 5G era, intelligent factories become more and more popular, and the digitization of factory personnel management has developed more maturely, but is still in the development stage in the aspect of the digitization of industrial workpiece management.
Aiming at the management of the workpiece, AI visual identification can be carried out on the code of the workpiece through a camera at a fixed position or a mobile terminal to read the steel seal code, and the obtained steel seal code is identified for further processing. The work management technology is mastered, so that staff management and work management digitization can be combined, and industrial high-efficiency production management modes in multiple aspects such as work production, work sorting, staff attendance and piece counting can be created.
However, in the process of workpiece identification, due to environmental factors, for example: light, temperature, humidity and the environment of mill, the workman gives the locating place of work piece, shoots the angle, and the steel seal is beaten the mark and is lightly different also the discernment rate of accuracy problem of AI itself. Under the influence of the factors, the AI identification accuracy is greatly reduced, and the error rate of the identified code is high. This is a great challenge to the first step of making the work piece management system robust. It is considered that the accuracy is improved by controlling the above variables, but a great deal of manpower and material cost is increased, and more unknown variables are also generated.
Disclosure of Invention
Aiming at the problems in the prior art, the invention provides a reasonably designed method for reversely calibrating the steel seal code through java, so as to improve the accuracy of identifying the steel seal code.
The technical scheme of the invention is as follows:
a method for carrying out reverse calibration steel seal coding through java comprises the following steps:
1) the APP obtains a plurality of photos through photographing, the collected photos are transmitted to a Python identification program through an http protocol, and after the identification program finishes identification, the identified codes are transmitted to a java server;
2) the java server receives the coded data and the identification times identified by the Python identification program, compares the character strings in the database with the identified coded data by a contiins method in the ArrayList of java, and returns false to the Python identification program if the coded data is not completely matched with the character strings in the database;
3) the Python identification program identifies the second photo, then sends the code to the java server, and performs the operation of the step 2), until the photographed photos are completely matched and the java server compares the database and still does not find the completely same code, the PC terminal calls a similarity matching algorithm to perform similarity comparison matching on the identified code of the last photo and the database code;
4) after the similarity matching algorithm is compared and matched, if the matching is successful, the java server returns a Python identification program, and the personnel station information and the processing identification code are stored in a database; and if the matching fails, returning the three codes with the highest similarity of the APP for manual selection, performing manual input, and storing the codes in a database after the input is finished.
Further, the similarity matching algorithm in the step 4) comprises the following steps:
4.1) putting the Jaro-Winkler similarity algorithm at the first position of the algorithm, selecting a plurality of codes with higher similarity values, and if only one of the codes exceeds a set threshold, returning a Python identification program to the java server side, wherein the matching is successful; otherwise, starting to use a cosine similarity algorithm;
4.2) selecting a plurality of codes with higher similarity values through a cosine similarity algorithm, and if only one of the codes exceeds a set threshold, returning a Python identification program to the java server side, wherein the matching is successful; otherwise, starting to use a Levenshtein Distance algorithm;
4.3) selecting a plurality of codes with higher similarity values through a Levenshtein Distance algorithm, and if only one code exceeds a set threshold, returning a Python identification program to the java server side, wherein the matching is successful; otherwise, starting to use LCS algorithm;
4.4) selecting a plurality of codes with higher similarity values through an LCS algorithm, and if only one code exceeds a set threshold, returning the Java service end to a Python identification program, and successfully matching.
The invention has the beneficial effects that: the steel seal code is identified and calibrated through multiple times of identification of multiple photos and combination of four similarity matching algorithms, so that the accuracy rate of identifying the steel seal code is improved.
Drawings
FIG. 1 is an overall flow chart of the present invention;
FIG. 2 is an interaction flow diagram of a java server and a Python program according to the present invention;
FIG. 3 is a flow chart of similarity matching according to the present invention.
Detailed Description
The invention is further described below with reference to the accompanying drawings.
As shown in fig. 1-3, a method for reverse calibration of steel seal codes through java is based on a steel seal code recognition system, and the steel seal code recognition system comprises a shooting module, a photo code AI recognition module, a code matching calibration service module and a database module. The reverse workpiece steel seal calibration firstly needs to realize the functions of photographing, sending and receiving feedback of the workpiece steel seal on a mobile phone through the mobile phone with a camera. And transmitting the shot picture to an AI identification program through an HTTP (hyper text transport protocol), and transmitting the identified code to the server by the AI program. The method mainly teaches how a java server processes and matches codes from an AI program segment to obtain an optimal solution so as to assist in improving the identification rate of the steel seal codes.
The invention mainly provides a method for carrying out reverse verification on a steel seal through a java matching algorithm. The similarity value obtained by using a similarity algorithm to perform character string comparison calculation can also be matched with one or more coding values with high similarity, and although the correct steel seal code can be in the matching values, the time cost and the working efficiency of selecting the correct code for workers are undoubtedly increased. Meanwhile, different similarity algorithms can generate great similarity difference for different types of similar characters due to different principles, for example, the Jaro-Winkler similarity algorithm more highlights the character similarity calculation for prefixes, and if prefixes are similar or identical, a larger similarity value can be obtained, but the weight of the similarity of the following characters in the calculation is smaller. Therefore, the method uses four similarity matching algorithms, and performs reverse matching calibration by matching the Jaro-Winkler similarity algorithm, the LCS algorithm, the cosine similarity algorithm and the Levenshtein Distance algorithm.
The reverse matching object is a database table which stores all workpiece steel seal codes, and the steel seal code table introduced here is stored in a MySQL database. Firstly, the calculation logic of the algorithm, the matching logic among codes and the rule making requirement of the steel seal code are expressed through a java language, and an API (application programming interface) for an AI (artificial intelligence) recognition program running based on a Python language can receive the code obtained after the AI program is well recognized. And comparing and matching the code codes meeting the requirement of the steel seal code formulation rule with the code characters in the database table one by one through a preset similarity matching algorithm to respectively obtain a similarity value, wherein the range is between [0 and 1], if the similarity value is equal to 0, the AI is identified as empty, and if the similarity value is equal to 1, the AI is identified as successfully and completely matched. And taking the value with the maximum similarity, and if the value is returned to the server side, the matching is successful. Information returned to the Python program true. However, the number of codes in the database table can reach thousands or even tens of thousands, and a large number of codes which only differ by one or a few characters exist, so that a server is usually returned with a few codes with high matching degree for the next operation. If a plurality of character codes (up to three character strings) are returned, these codes are subjected to a second comparison matching with the code codes recognized by the AI, at which time the similarity calculation is performed using the second algorithm. And returning one or more values (at most three values) with the maximum similarity value after the comparison of the calculated values is finished, and so on until the fourth algorithm calculates and obtains the final similar code returned to the server computer and the maximum similarity value. And if a plurality of values still exist, the staff is enabled to select or manually input.
Example (b):
the method for carrying out reverse calibration steel seal coding through java comprises the following specific steps:
1) the APP obtains 5 photos through photographing, the collected photos are transmitted to a Python identification program through an http protocol, and after the identification program finishes identification, the identified codes are transmitted to a java server;
2) the java server receives the coded data and the identification times (tryTimes) identified by the Python identification program, compares the character strings in the database with the identified coded data by using the contiins method in the ArrayList of java, and returns false to the Python identification program if the matching is not complete;
3) a Python identification program identifies the 2 nd picture, then sends the code to a java server, and performs the step 2), until the 5 pictures are completely matched, and the java server compares the database and still does not find the completely same code, the PC terminal calls a similarity matching algorithm to perform similarity comparison matching on the identified code of the last 1 picture and the database code;
4) after the similarity is compared and matched, if the matching is successful, the java server returns to a Python identification program, and the personnel station information and the processing identification code are stored in a database; and if the matching fails, the java server returns 3 codes with the highest similarity of the APP at most for manual selection, manual input is carried out, and the codes are stored in the database after the input is finished.
The similarity contrast matching process is as follows:
4.1) putting the Jaro-Winkler similarity algorithm at the first position of the algorithm, selecting 3 codes with higher similarity values, and if only 1 code exceeds a set threshold value of 0.8, returning a Python identification program to the java server side, wherein the matching is successful; otherwise, starting to use a cosine similarity algorithm;
Jaro-Winkler similarity is modified on the basis of Jaro similarity, and the Jaro similarity is modified through the same prefixes of the two character strings, so that the importance of the prefixes is emphasized;
simJ-w＝simjaro+l*p(1-simjaro)
l is the same number of prefixes of the two character strings;
p is a scaling factor constant describing the contribution of the common prefix to the similarity; the larger p, the larger the common prefix weight; maximum not more than 0.25; p is taken to be 0.1 by default,
wherein-simjaroNamely Jaro similarity;
wherein | - [ s ]1|，|s2L represents the length of the character string;
m represents the number of matched characters of the two character strings;
t represents 1/2 for the index number transposition;
a match window is also needed, which can be understood as a threshold within which two characters equal, can be considered a match; beyond this threshold, even if there is another character equal to the character, the correlation is too low to consider them a match because they are too far apart. Matching window (mw) refers to the size of the boundary where the Jaro algorithm performs comparisons between characters, and if within this boundary two characters from two different strings are equal, the match is considered successful, m is incremented by 1; if no matching character can be found within this limit, the matching is considered to fail. In the Jaro algorithm, the size of the matching window is defined as follows:
4.2) selecting 3 codes with higher similarity values through a cosine similarity algorithm, and if only 1 code exceeds a set threshold value of 0.8, returning a Python identification program to the java server side, wherein the matching is successful; otherwise, starting to use a Levenshtein Distance algorithm;
the cosine similarity algorithm adopts the idea of space vector, all character strings are regarded as a space vector, and the character appearing in the character strings are regarded as a component in the vector.
Cosine similarity formula:
4.3) selecting 3 codes with higher similarity values through a Levenshtein Distance algorithm, and if only 1 code exceeds a set threshold value of 0.8, returning a Python identification program to the java server side, wherein the matching is successful; otherwise, starting to use LCS algorithm;
the Levebshot Distance algorithm is also called as an edit Distance algorithm, and the main idea of the algorithm is to calculate the edit Distance between two character strings. Editing distance: i.e., string a translates to the minimum number of operands required by string B. The character operations contained therein include: deleting a character; adding a character; a character is modified.
The similarity P (A, B) of the two character strings is as follows:
P(A，B)＝1-LD(A，B)/max(|A|，|B|)
LD (a, B) represents an edit distance;
4.4) selecting 3 codes with higher similarity values through an LCS algorithm, and if only 1 code exceeds a set threshold value of 0.8, returning the Python identification program to the java server side, wherein the matching is successful.
The LCS (long Common Sub-sequence) algorithm, the Longest subsequence algorithm, uses the LCS algorithm to solve the similarity between two strings, the idea of which is to calculate the similarity between two strings by calculating the Longest subsequence of the two strings; the subsequence is to delete any items in a character string, and the rest character sequence is the subsequence of the character string; the longest subsequence of two strings is the longest of the same subsequences of the two strings.
According to the calculated longest subsequence length, the similarity of the two character strings can be calculated by the following formula.
Wherein, A and B are two character strings to be calculated respectively; calculating the longest subsequence of the two character strings by an LCS () function; length and b.length refer to the length of string a and string B, respectively.
Claims (2)
1. A method for carrying out reverse calibration steel seal coding through java is characterized by comprising the following steps:
1) the APP obtains a plurality of photos through photographing, the collected photos are transmitted to a Python identification program through an http protocol, and after the identification program finishes identification, the identified codes are transmitted to a java server;
2) the java server receives the coded data and the identification times identified by the Python identification program, compares the character strings in the database with the identified coded data by a contiins method in the ArrayList of java, and returns false to the Python identification program if the coded data is not completely matched with the character strings in the database;
3) the Python identification program identifies the second picture, then sends the code to the java server, and performs the operation of the step 2), until the photographed pictures are completely matched and the java server compares the database and still does not find the completely same code, the java server calls a similarity matching algorithm to perform similarity comparison matching on the code of the identified last picture and the database code;
4) after the similarity matching algorithm is compared and matched, if the matching is successful, the java server returns a Python identification program, and the personnel station information and the processing identification code are stored in a database; and if the matching fails, returning the three codes with the highest similarity of the APP for manual selection, performing manual input, and storing the codes in a database after the input is finished.
2. The method for reverse calibration steel seal coding by java according to claim 1, wherein the similarity matching algorithm in step 4) comprises the following steps:
4.1) putting the Jaro-Winkler similarity algorithm at the first position of the algorithm, selecting a plurality of codes with higher similarity values, and if only one of the codes exceeds a set threshold, returning a Python identification program to the java server side, wherein the matching is successful; otherwise, starting to use a cosine similarity algorithm;
4.2) selecting a plurality of codes with higher similarity values through a cosine similarity algorithm, and if only one of the codes exceeds a set threshold, returning a Python identification program to the java server side, wherein the matching is successful; otherwise, starting to use a Levenshtein Distance algorithm;
4.3) selecting a plurality of codes with higher similarity values through a Levenshtein Distance algorithm, and if only one code exceeds a set threshold, returning a Python identification program to the java server side, wherein the matching is successful; otherwise, starting to use LCS algorithm;
4.4) selecting a plurality of codes with higher similarity values through an LCS algorithm, and if only one code exceeds a set threshold, returning the Java service end to a Python identification program, and successfully matching.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110631594.6A CN113312525B (en) | 2021-06-07 | 2021-06-07 | Method for reversely calibrating seal code through java |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110631594.6A CN113312525B (en) | 2021-06-07 | 2021-06-07 | Method for reversely calibrating seal code through java |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN113312525A true CN113312525A (en) | 2021-08-27 |
| CN113312525B CN113312525B (en) | 2024-02-09 |
Family
ID=77377882
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202110631594.6A Active CN113312525B (en) | 2021-06-07 | 2021-06-07 | Method for reversely calibrating seal code through java |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN113312525B (en) |
Citations (17)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JPH1082613A (en) * | 1996-09-09 | 1998-03-31 | Matsushita Electric Ind Co Ltd | Method and apparatus for visual recognition |
| JP2011065646A (en) * | 2009-09-18 | 2011-03-31 | Fujitsu Ltd | Apparatus and method for recognizing character string |
| CN103927352A (en) * | 2014-04-10 | 2014-07-16 | 江苏唯实科技有限公司 | Chinese business card OCR (optical character recognition) data correction system utilizing massive associated information of knowledge base |
| WO2015087359A1 (en) * | 2013-12-11 | 2015-06-18 | Unicredit S.P.A. | Method for biometric recognition of a user amongst a plurality of registered users to a service, employing user localization information |
| US20170032605A1 (en) * | 2015-07-31 | 2017-02-02 | Ncr Corporation | Identifying desynchronization between representations of data obtained from an ordered plurality of documents |
| CN108765970A (en) * | 2018-08-07 | 2018-11-06 | 普行智能停车（深圳）有限公司 | The method and apparatus of curb parking automatic identification automotive license plate |
| CN109767376A (en) * | 2018-12-13 | 2019-05-17 | 平安科技（深圳）有限公司 | Image authentication method, device, computer equipment and storage medium |
| JP2019117622A (en) * | 2017-12-27 | 2019-07-18 | セイコーエプソン株式会社 | Image processing system and image processing program |
| US10402490B1 (en) * | 2015-08-14 | 2019-09-03 | Shutterstock, Inc. | Edit distance based spellcheck |
| CN110781272A (en) * | 2019-09-10 | 2020-02-11 | 杭州云深科技有限公司 | Text matching method and device and storage medium |
| CN110795607A (en) * | 2019-10-29 | 2020-02-14 | 中国人民解放军32181部队 | Equipment guarantee data matching method and system based on multi-stage similarity calculation |
| CN111695349A (en) * | 2019-02-28 | 2020-09-22 | 北京京东尚科信息技术有限公司 | Text matching method and text matching system |
| CN111797214A (en) * | 2020-06-24 | 2020-10-20 | 深圳壹账通智能科技有限公司 | FAQ database-based problem screening method and device, computer equipment and medium |
| CN112149401A (en) * | 2020-08-10 | 2020-12-29 | 江苏群杰物联科技有限公司 | Document comparison identification method and system based on ocr |
| CN112215216A (en) * | 2020-09-10 | 2021-01-12 | 中国东方电气集团有限公司 | Character string fuzzy matching system and method for image recognition result |
| US20210049326A1 (en) * | 2016-06-03 | 2021-02-18 | Bottomline Technologies (De) Inc. | Identifying Non-Exactly Matching Text |
| CN112613522A (en) * | 2021-01-04 | 2021-04-06 | 重庆邮电大学 | Method for correcting recognition result of medicine taking order based on fusion font information |
-
2021
- 2021-06-07 CN CN202110631594.6A patent/CN113312525B/en active Active
Patent Citations (17)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JPH1082613A (en) * | 1996-09-09 | 1998-03-31 | Matsushita Electric Ind Co Ltd | Method and apparatus for visual recognition |
| JP2011065646A (en) * | 2009-09-18 | 2011-03-31 | Fujitsu Ltd | Apparatus and method for recognizing character string |
| WO2015087359A1 (en) * | 2013-12-11 | 2015-06-18 | Unicredit S.P.A. | Method for biometric recognition of a user amongst a plurality of registered users to a service, employing user localization information |
| CN103927352A (en) * | 2014-04-10 | 2014-07-16 | 江苏唯实科技有限公司 | Chinese business card OCR (optical character recognition) data correction system utilizing massive associated information of knowledge base |
| US20170032605A1 (en) * | 2015-07-31 | 2017-02-02 | Ncr Corporation | Identifying desynchronization between representations of data obtained from an ordered plurality of documents |
| US10402490B1 (en) * | 2015-08-14 | 2019-09-03 | Shutterstock, Inc. | Edit distance based spellcheck |
| US20210049326A1 (en) * | 2016-06-03 | 2021-02-18 | Bottomline Technologies (De) Inc. | Identifying Non-Exactly Matching Text |
| JP2019117622A (en) * | 2017-12-27 | 2019-07-18 | セイコーエプソン株式会社 | Image processing system and image processing program |
| CN108765970A (en) * | 2018-08-07 | 2018-11-06 | 普行智能停车（深圳）有限公司 | The method and apparatus of curb parking automatic identification automotive license plate |
| CN109767376A (en) * | 2018-12-13 | 2019-05-17 | 平安科技（深圳）有限公司 | Image authentication method, device, computer equipment and storage medium |
| CN111695349A (en) * | 2019-02-28 | 2020-09-22 | 北京京东尚科信息技术有限公司 | Text matching method and text matching system |
| CN110781272A (en) * | 2019-09-10 | 2020-02-11 | 杭州云深科技有限公司 | Text matching method and device and storage medium |
| CN110795607A (en) * | 2019-10-29 | 2020-02-14 | 中国人民解放军32181部队 | Equipment guarantee data matching method and system based on multi-stage similarity calculation |
| CN111797214A (en) * | 2020-06-24 | 2020-10-20 | 深圳壹账通智能科技有限公司 | FAQ database-based problem screening method and device, computer equipment and medium |
| CN112149401A (en) * | 2020-08-10 | 2020-12-29 | 江苏群杰物联科技有限公司 | Document comparison identification method and system based on ocr |
| CN112215216A (en) * | 2020-09-10 | 2021-01-12 | 中国东方电气集团有限公司 | Character string fuzzy matching system and method for image recognition result |
| CN112613522A (en) * | 2021-01-04 | 2021-04-06 | 重庆邮电大学 | Method for correcting recognition result of medicine taking order based on fusion font information |
Also Published As
| Publication number | Publication date |
|---|---|
| CN113312525B (en) | 2024-02-09 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| KR101323439B1 (en) | Method and apparatus for representing and identifying feature descriptors utilizing a compressed histogram of gradients | |
| CN109446885B (en) | Text-based component identification method, system, device and storage medium | |
| CN112100337B (en) | Emotion recognition method and device in interactive dialogue | |
| CN109002784B (en) | Street view identification method and system | |
| CN110348362B (en) | Label generation method, video processing method, device, electronic equipment and storage medium | |
| EP3553702A1 (en) | Image recognition-based communication method and device | |
| US10210602B2 (en) | System and method for normalized focal length profiling | |
| CN111401326A (en) | Target identity recognition method based on picture recognition, server and storage medium | |
| CN111368867A (en) | Archive classification method and system and computer readable storage medium | |
| US11568661B2 (en) | Methods and systems for data retrieval from an image | |
| CN115630236A (en) | Global fast retrieval positioning method of passive remote sensing image, storage medium and equipment | |
| CN110633733B (en) | Image intelligent matching method, device and computer readable storage medium | |
| CN113254687B (en) | Image retrieval and image quantification model training method, device and storage medium | |
| CN111460829A (en) | Intention identification method, device and equipment under multi-scene application and storage medium | |
| CN108415938A (en) | A kind of method and system of the data automatic marking based on intelligent mode identification | |
| CN113312525A (en) | Method for reversely calibrating steel seal code through java | |
| CN116843155B (en) | SAAS-based person post bidirectional matching method and system | |
| WO2021042527A1 (en) | Character recognition method and apparatus, and computer-readable storage medium | |
| CN114882334A (en) | Method for generating pre-training model, model training method and device | |
| KR101551122B1 (en) | Pattern recognition method and apparatus within Terminal | |
| CN111267096A (en) | Robot translation skill training method and device, electronic equipment and storage medium | |
| CN113920306A (en) | Target re-identification method and device and electronic equipment | |
| CN113591752A (en) | Equipment oil leakage monitoring method based on convolutional neural network and related equipment | |
| KR20210058157A (en) | Apparatus and method for identificating user based on face image | |
| CN109920406B (en) | Dynamic voice recognition method and system based on variable initial position |
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 | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |