CN116403203A - Label generation method, system, electronic equipment and storage medium - Google Patents
Label generation method, system, electronic equipment and storage medium Download PDFInfo
- Publication number
- CN116403203A CN116403203A CN202310660056.9A CN202310660056A CN116403203A CN 116403203 A CN116403203 A CN 116403203A CN 202310660056 A CN202310660056 A CN 202310660056A CN 116403203 A CN116403203 A CN 116403203A
- Authority
- CN
- China
- Prior art keywords
- tag
- template
- label
- filled
- article
- 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
- 238000000034 method Methods 0.000 title claims abstract description 34
- 238000003860 storage Methods 0.000 title claims abstract description 16
- 238000013136 deep learning model Methods 0.000 claims abstract description 38
- 238000013135 deep learning Methods 0.000 claims abstract description 7
- 238000004590 computer program Methods 0.000 claims description 9
- 230000004927 fusion Effects 0.000 claims description 8
- 238000000605 extraction Methods 0.000 abstract description 3
- 230000000694 effects Effects 0.000 description 2
- 230000006872 improvement Effects 0.000 description 2
- 230000003287 optical effect Effects 0.000 description 2
- 230000008569 process Effects 0.000 description 2
- 238000004458 analytical method Methods 0.000 description 1
- 230000002457 bidirectional effect Effects 0.000 description 1
- 230000008859 change Effects 0.000 description 1
- 230000007547 defect Effects 0.000 description 1
- 238000010586 diagram Methods 0.000 description 1
- 238000009776 industrial production Methods 0.000 description 1
- 238000007726 management method Methods 0.000 description 1
- 239000000463 material Substances 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 238000005406 washing Methods 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V20/00—Scenes; Scene-specific elements
- G06V20/60—Type of objects
- G06V20/62—Text, e.g. of license plates, overlay texts or captions on TV images
- G06V20/63—Scene text, e.g. street names
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/08—Learning methods
-
- 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
- G06V30/19—Recognition using electronic means
- G06V30/19007—Matching; Proximity measures
- G06V30/19093—Proximity measures, i.e. similarity or distance 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
- G06V30/19—Recognition using electronic means
- G06V30/191—Design or setup of recognition systems or techniques; Extraction of features in feature space; Clustering techniques; Blind source separation
- G06V30/1918—Fusion techniques, i.e. combining data from various sources, e.g. sensor fusion
-
- 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)
- Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Multimedia (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Biophysics (AREA)
- Evolutionary Computation (AREA)
- Artificial Intelligence (AREA)
- Biomedical Technology (AREA)
- Health & Medical Sciences (AREA)
- Computational Linguistics (AREA)
- Data Mining & Analysis (AREA)
- Life Sciences & Earth Sciences (AREA)
- General Health & Medical Sciences (AREA)
- Molecular Biology (AREA)
- Computing Systems (AREA)
- General Engineering & Computer Science (AREA)
- Mathematical Physics (AREA)
- Software Systems (AREA)
- Image Analysis (AREA)
Abstract
The invention discloses a label generation method, a label generation system, electronic equipment and a storage medium. The method comprises the following steps: acquiring an article tag image and a plurality of predefined tag templates, and searching a tag template with the highest matching degree for the article tag image to be used as a tag template to be filled; inputting the article tag image into a first deep learning model based on deep learning, wherein the first deep learning model is used for outputting fields and attribute values in the article tag image by adopting semantic entity identification and entity relation extraction technology; inputting the label template to be filled into the first deep learning model, and outputting the field of the label template to be filled; and matching and filling the fields and the attribute values in the article tag image with the fields of the tag template to be filled, and generating a tag based on the tag template to be filled. The method and the device can improve the user experience of label generation.
Description
Technical Field
The invention belongs to the technical field of computers, and in particular relates to a label generation method, a label generation system, electronic equipment and a storage medium.
Background
The label refers to the literal, graphic and symbol on the product and all the explanatory matters. The label is widely applied to the scenes of super-retail business, industrial production, express delivery, clothing, office management and the like, such as price labels, product description labels, shelf labels, bar code labels, clothing labels, document labels, file storage labels, various articles, stationery labels, express delivery face sheets and the like. The user can edit the label by using label printing software matched with the labeler, and the label is sent to the labeler for printing after the editing is finished.
In order to facilitate user operation and improve user experience, the existing label printing software proposes to automatically generate an electronic label on the label printing software by scanning paper labels on articles such as clothing, goods and the like. However, by visiting a large number of commercial users, there are problems at present, for example, after scanning is completed, each content block in the generated electronic tag is easy to mix in one paragraph, and cannot be directly used, and further editing and beautifying operations are required for the users.
Disclosure of Invention
Aiming at the defects or improvement demands of the prior art, the invention provides a label generation method, a label generation system, electronic equipment and a storage medium, which can promote user experience of label generation.
To achieve the above object, according to one aspect of the present invention, there is provided a tag generation method comprising the steps of:
acquiring an article tag image and a plurality of predefined tag templates, and searching a tag template with the highest matching degree for the article tag image to be used as a tag template to be filled;
inputting the article tag image to a first deep learning model based on deep learning, wherein the first deep learning model is used for outputting fields and attribute values in the article tag image;
inputting the label template to be filled into the first deep learning model, and outputting the field of the label template to be filled;
and matching and filling the fields and the attribute values in the article tag image with the fields of the tag template to be filled, and generating a tag based on the tag template to be filled.
Further, searching a label template with highest matching degree for the article label image comprises the following steps:
identifying texts contained in the article tag images, inputting the texts contained in the article tag images into a second deep learning model, and obtaining embedded representations of the article tag images;
identifying texts contained in each label template, and respectively inputting the texts contained in each label template into the second deep learning model to obtain embedded representations of each label template;
and calculating the similarity between the embedded representation of the article tag image and the embedded representation of each tag template, and searching the tag template with the highest similarity for the article tag image.
Further, the calculating the similarity between the embedded representation of the item label image and the embedded representation of each label template includes the steps of:
clustering the embedded representations of all the tag templates, and classifying the embedded representations of all the tag templates into multiple classes;
and determining a class matched with the embedded representation of the article tag image, and calculating the similarity between the embedded representation of the article tag image and the embedded representation of each tag template in the class.
Further, the method further comprises the steps of:
identifying layout information and pictures contained in the article tag image, carrying out multi-modal fusion on texts contained in the article tag image and the layout information and pictures contained in the article tag image, and then inputting the texts into the second deep learning model;
and identifying the layout information and the pictures contained in each label template, carrying out multi-modal fusion on the text contained in each label template and the layout information and the pictures contained in each label template, and then inputting the text into the second deep learning model.
Further, the matching and filling steps include:
the fields in the article label image are marked as Ai, i is more than or equal to 1 and less than or equal to N, N is the total number of the fields in the article label image, the fields in the label template to be filled are marked as Bj, j is more than or equal to 1 and less than or equal to M, and M is the total number of the fields in the label template to be filled;
for each field Ai, searching a field Bj with the highest matching degree with the field Ai, and taking the attribute value of the field Ai as the attribute value of the field with the highest matching degree.
Further, if the field with the highest matching degree is different from the field Ai, replacing the field with the highest matching degree with the field Ai in the template to be filled.
Further, the generating the label based on the label template to be filled comprises the steps of:
and after the label template to be filled is filled, adjusting the character layout style of the attribute value based on the size of the attribute value filling frame in the label template to be filled.
According to a second aspect of the present invention, there is also provided a tag generation system comprising:
the matching module is used for acquiring an article tag image and a plurality of predefined tag templates, searching the tag template with the highest matching degree for the article tag image, and taking the tag template with the highest matching degree as a tag template to be filled;
the identification module is used for inputting the article tag image into a first deep learning model based on deep learning, wherein the first deep learning model is used for outputting fields and attribute values in the article tag image, inputting the tag template to be filled into the first deep learning model and outputting the fields of the tag template to be filled;
and the filling module is used for matching and filling the fields and the attribute values in the article label image with the fields of the label template to be filled, and generating a label based on the label template to be filled.
According to a third aspect of the present invention there is also provided an electronic device comprising at least one processing unit and at least one storage unit, wherein the storage unit stores a computer program which, when executed by the processing unit, causes the processing unit to perform the steps of any of the methods described above.
According to a fourth aspect of the present invention there is also provided a storage medium storing a computer program executable by an access authentication device, the computer program, when run on the access authentication device, causing the access authentication device to perform the steps of any of the methods described above.
In general, compared with the prior art, the technical scheme of the invention can select the optimal label template after scanning the object label, and identify and fill the fields and the attribute values based on the optimal label template, so that the risk of overlapping a plurality of contents in the generated label can be greatly reduced, the user is not required to carry out subsequent further editing and beautifying, and the user experience is greatly improved.
Drawings
FIG. 1 is a flow chart of a label generation method according to an embodiment of the present invention;
FIG. 2 is a schematic illustration of an image of an item tag according to an embodiment of the present invention;
fig. 3 is a schematic diagram of a tag generation system of an embodiment of the present invention.
Detailed Description
The present invention will be described in further detail with reference to the drawings and examples, in order to make the objects, technical solutions and advantages of the present invention more apparent. It should be understood that the specific embodiments described herein are for purposes of illustration only and are not intended to limit the scope of the invention. In addition, the technical features of the embodiments of the present invention described below may be combined with each other as long as they do not collide with each other.
In the description of the embodiments of the present application, the terms "first," "second," and the like are used for descriptive purposes only and are not to be construed as indicating or implying relative importance or implicitly indicating the number of technical features indicated. Thus, a feature defining "a first" or "a second" may explicitly or implicitly include at least one such feature. The meaning of "a plurality of" means at least two, e.g., two, three, etc., unless explicitly defined otherwise.
The terms "comprises," "comprising," and "having," and any variations thereof, are intended to cover a non-exclusive inclusion, such that a process, method, apparatus, article, or device that comprises a list of steps or modules is not necessarily limited to those steps or modules that are expressly listed or inherent to such process, method, article, or device.
The naming or numbering of the steps in the embodiments of the present invention does not mean that the steps in the method flow must be executed according to the time/logic sequence indicated by the naming or numbering, and the named or numbered flow steps may change the execution order according to the technical purpose to be achieved, so long as the same or similar technical effects can be achieved.
Reference herein to "an embodiment" means that a particular feature, structure, or characteristic described in connection with the embodiment may be included in at least one embodiment of the invention. The appearances of such phrases in various places in the specification are not necessarily all referring to the same embodiment, nor are separate or alternative embodiments mutually exclusive of other embodiments. Those of skill in the art will explicitly and implicitly appreciate that the embodiments described herein may be combined with other embodiments.
As shown in fig. 1, an embodiment of the present invention provides a tag generation method, which includes steps S101 to S104.
S101, acquiring an article label image and a plurality of predefined label templates, and searching the label template with the highest matching degree for the article label image to be used as a label template to be filled.
The item label may be scanned by a mobile device camera to obtain an item label image, such as a scanned price label, a product description label, a shelf label, a barcode label, a clothing label.
A plurality of different label templates are predefined, and the fields and styles of each label template are different. For example, according to the characteristics of industry labels, label templates of different industries may be preset, fields of label templates of clothing industries include names, manufacturers, materials, washing modes, and the like, and fields of label templates of food industries may include names, manufacturers, validity periods, and the like. Attribute values may also be predefined for the fields. For the same industry, tag templates with different layout can be designed.
Further, a picture may also be provided on the label template.
Further, searching a label template with highest matching degree for the label image of the article comprises the following steps: identifying texts contained in the object tag images, inputting the texts contained in the object tag images into a second deep learning model, and obtaining embedded representations of the object tag images; identifying texts contained in each label template, and respectively inputting the texts contained in each label template into a second deep learning model to obtain embedded representations of each label template; and calculating the similarity between the embedded representation of the object tag image and the embedded representation of each tag template, and searching the tag template with the highest similarity for the object tag image.
The second deep learning model may convert the input information into an embedded representation that is back-end processable. The second deep learning model may be any existing model for semantic analysis, for example, an ERNIE 3.0 model may be used, and the ERNIE 3.0 model may learn knowledge of various aspects of language, such as morphology, syntax, and semantic information, by setting a pre-training task.
After the embedded representation of the object tag image and the embedded representation of each tag template are obtained, the similarity of the embedded representation of the object tag image and each tag template can be calculated one by one, and the tag template with the highest similarity is used as the tag template to be filled.
Further, the method further comprises the steps of:
identifying layout information and pictures contained in the article tag image, carrying out multi-modal fusion on texts contained in the article tag image and the layout information and pictures contained in the article tag image, and then inputting the texts into a second deep learning model;
and identifying the layout information and the pictures contained in each label template, carrying out multi-modal fusion on the text contained in each label template and the layout information and the pictures contained in each label template, and then inputting the text into the second deep learning model.
In this embodiment, the characteristics of the text, the layout information and the picture after the multimodal fusion can be used as the input of the second deep learning model. Compared with a method based on text only, the method can simultaneously utilize text information, layout information and pictures, so that matching precision can be improved. The layout information comprises the field and the position information of the picture; fields, relative position information between pictures, etc.
Further, calculating a similarity of the embedded representation of the item label image and the embedded representation of each label template, comprising the steps of:
clustering the embedded representations of all the tag templates, and classifying the embedded representations of all the tag templates into multiple classes;
determining the class to which the embedded representation of the item label image belongs, and calculating the similarity between the embedded representation of the item label image and the embedded representation of each label template in the class.
The clustering can adopt various existing clustering algorithms, the clustering algorithm is adopted to divide the embedded representations of all the label templates into several types, the embedded representations of different types are greatly different, and then the embedded representation of the label image of the article is determined to belong to which of the several types. In calculating the similarity, it is not necessary to calculate the similarity between the embedded representation of the item label image and the embedded representations of all the label templates, but only the similarity between the embedded representation of the item label image and the embedded representation of the label templates in the class to which it belongs. Thus, the data volume for calculating the similarity can be greatly reduced, and the efficiency for calculating the similarity is improved.
S102, inputting the object tag image into a first deep learning model based on deep learning, wherein the first deep learning model is used for outputting fields and attribute values in the object tag image.
The first deep learning model may employ a BERT (Bidirectional Encoder Representation from Transformers) model that outputs field and attribute values in the item-tag image using semantic entity identification and entity relationship extraction techniques.
Semantic entity identification, each detected text may be classified, such as into name, price, etc. Entity relationship extraction can identify relationships between text, take fields as keys, and find corresponding values for each key.
For example, as shown in fig. 2, it is recognized that the article tag image includes "article", "95 gasoline", "unit", "liter", and "article" and "unit" as fields, and "95 gasoline" as an attribute value of "article" and "liter" as an attribute value of "unit".
S103, inputting the label template to be filled into the first deep learning model, and outputting the field of the label template to be filled.
And similarly, inputting the label template with the highest matching degree screened in the step S101 into the first deep learning model. The fields of the tag template may have corresponding attribute values, or may be empty. If the fields of the label template have corresponding attribute values, the attribute values corresponding to each field can also be identified.
S104, matching and filling the fields and the attribute values in the object label image with the fields of the label template to be filled, and generating a label based on the label template to be filled.
Further, matching and filling are performed, comprising the steps of:
the fields in the article label image are marked as Ai, i is more than or equal to 1 and less than or equal to N, N is the total number of fields in the article label image, the fields in the label template to be filled are marked as Bj, j is more than or equal to 1 and less than or equal to M, and M is the total number of fields in the label template to be filled.
For each field Ai, searching a field Bj with the highest matching degree with the field Ai, and taking the attribute value of the field Ai as the attribute value of the field with the highest matching degree.
The field with the highest matching degree with A1 is determined, the field with the highest matching degree with A2 is determined, the field with the highest matching degree with A3 is determined, and the field with the highest matching degree with A2 is determined, and the field with the highest matching degree with A3 is determined, and the field with the highest matching degree with A2 is determined. And taking the attribute value of A1 as the attribute value of B2, directly filling the attribute value of A1 into B2 if the attribute value of B2 is null, and directly replacing the attribute value of A1 with the attribute value of B2 if the attribute value of B2 is not null. For example, if the label template to be filled contains a field "name", the article label image also contains a field "name", and the attribute value is "95 gasoline", then "95 gasoline" is taken as the attribute value of the field "name" in the label template to be filled.
It should be noted here that for the same meaning, the descriptions in the item label image and the label template may be different, e.g. the field in the item label image is "good", while in the label template to be filled the field that best matches the field "good" is "name", both expressed in the same meaning. By adopting the method provided by the embodiment of the invention, the expansion of the paraphrasing and the synonyms can be supported, and the matching probability is increased.
Further, if the field with the highest matching degree is different from the field Ai, the field with the highest matching degree is replaced by the field Ai in the template to be filled.
In order to further enhance the user experience, the expression habit of the user is respected, so that the field 'name' of the label template to be filled is replaced by 'commodity', and the description of 'commodity' and 'name' is adopted in the finally generated label. But the layout etc. information is generated based on the label template to be populated. For example, the size and position of the field "commodity" are generated according to the preconfigured size and position information of the label template to be filled, and the characters are replaced by the "commodity".
Further, generating a label based on the label template to be filled comprises the steps of:
after the label template to be filled is filled, the character layout style of the attribute value is adjusted based on the size of the attribute value filling frame in the label template to be filled. For example, if the character of the attribute value is too large, the character needs to be scaled down to be displayed entirely within the filled-in frame. Therefore, related information such as related text fonts, sizes, line feed widths, styles and the like can be predefined, and only the image library related interface is required to be called for adjusting the predefined format.
As shown in fig. 3, according to a second aspect of the present invention, there is also provided a tag generation system including:
the matching module 201 is configured to obtain an article tag image and a predefined plurality of tag templates, search a tag template with the highest matching degree for the article tag image, and use the tag template as a tag template to be filled;
the identification module 202 is configured to input an article tag image to a first deep learning model based on deep learning, where the first deep learning model is configured to output fields and attribute values in the article tag image, and input a tag template to be filled to the first deep learning model, and output the fields and attribute values of the tag template to be filled;
and the filling module 203 is configured to match and fill the fields and the attribute values in the article tag image with the fields and the attribute values of the tag template to be filled, and generate a tag based on the tag template to be filled.
The principle and effect of the label generation system are the same as those of the label generation, and are not repeated here.
The embodiment of the invention also provides electronic equipment, which comprises at least one processing unit and at least one storage unit, wherein the storage unit stores a computer program, and the computer program, when executed by the processing unit, causes the processing unit to execute the steps of any one of the methods.
The electronic device may be a portable electronic device such as a mobile phone, a tablet computer, a wearable device, and the like.
The embodiment of the invention also provides a storage medium storing a computer program executable by an access authentication device, which when run on the access authentication device causes the access authentication device to perform the steps of any of the methods described above.
The storage medium may include, but is not limited to, any type of disk including floppy disks, optical disks, DVDs, CD-ROMs, micro-drives, and magneto-optical disks, ROM, RAM, EPROM, EEPROM, DRAM, VRAM, flash memory devices, magnetic or optical cards, nanosystems (including molecular memory ICs), or any type of media or device suitable for storing instructions and/or data.
It will be readily appreciated by those skilled in the art that the foregoing description is merely a preferred embodiment of the invention and is not intended to limit the invention, but any modifications, equivalents, improvements or alternatives falling within the spirit and principles of the invention are intended to be included within the scope of the invention.
Claims (10)
1. A label generation method, characterized by comprising the steps of:
acquiring an article tag image and a plurality of predefined tag templates, and searching a tag template with the highest matching degree for the article tag image to be used as a tag template to be filled;
inputting the article tag image to a first deep learning model based on deep learning, wherein the first deep learning model is used for outputting fields and attribute values in the article tag image;
inputting the label template to be filled into the first deep learning model, and outputting the field of the label template to be filled;
and matching and filling the fields and the attribute values in the article tag image with the fields of the tag template to be filled, and generating a tag based on the tag template to be filled.
2. The tag generation method of claim 1, wherein searching for the tag template having the highest matching degree for the article tag image comprises the steps of:
identifying texts contained in the article tag images, inputting the texts contained in the article tag images into a second deep learning model, and obtaining embedded representations of the article tag images;
identifying texts contained in each label template, and respectively inputting the texts contained in each label template into the second deep learning model to obtain embedded representations of each label template;
and calculating the similarity between the embedded representation of the article tag image and the embedded representation of each tag template, and searching the tag template with the highest similarity for the article tag image.
3. The label generation method of claim 2, wherein said calculating the similarity of the embedded representation of the item label image and the embedded representation of each label template comprises the steps of:
clustering the embedded representations of all the tag templates, and classifying the embedded representations of all the tag templates into multiple classes;
and determining the class to which the embedded representation of the article tag image belongs, and calculating the similarity between the embedded representation of the article tag image and the embedded representation of each tag template in the class.
4. The tag generation method of claim 2, further comprising the step of:
identifying layout information and pictures contained in the article tag image, performing multi-mode fusion on texts contained in the article tag image and the layout information and pictures contained in the article tag image, and then inputting the multi-mode fusion into the second deep learning model;
and identifying the layout information and the pictures contained in each label template, carrying out multi-mode fusion on the text contained in each label template and the layout information and the pictures contained in each label template, and inputting the text and the layout information and the pictures into the second deep learning model.
5. The tag generation method of claim 1, wherein the matching and filling comprises the steps of:
the fields in the article label image are marked as Ai, i is more than or equal to 1 and less than or equal to N, N is the total number of the fields in the article label image, the fields in the label template to be filled are marked as Bj, j is more than or equal to 1 and less than or equal to M, and M is the total number of the fields in the label template to be filled;
for each field Ai, searching a field Bj with the highest matching degree with the field Ai, and taking the attribute value of the field Ai as the attribute value of the field with the highest matching degree.
6. The tag generation method according to claim 5, wherein if the field with the highest matching degree is different from the field Ai, the field with the highest matching degree is replaced with the field Ai in the template to be filled.
7. The tag generation method of claim 1, wherein the generating the tag based on the tag template to be filled comprises the steps of:
and after the label template to be filled is filled, adjusting the character layout style of the attribute value based on the size of the attribute value filling frame in the label template to be filled.
8. A label producing system, comprising:
the matching module is used for acquiring an article tag image and a plurality of predefined tag templates, searching the tag template with the highest matching degree for the article tag image, and taking the tag template with the highest matching degree as a tag template to be filled;
the identification module is used for inputting the article tag image into a first deep learning model based on deep learning, wherein the first deep learning model is used for outputting fields and attribute values in the article tag image, inputting the tag template to be filled into the first deep learning model and outputting the fields of the tag template to be filled;
and the filling module is used for matching and filling the fields and the attribute values in the article label image with the fields of the label template to be filled, and generating a label based on the label template to be filled.
9. An electronic device comprising at least one processing unit and at least one storage unit, wherein the storage unit stores a computer program that, when executed by the processing unit, causes the processing unit to perform the steps of the method of any of claims 1-7.
10. A storage medium storing a computer program executable by an access authentication device, the computer program, when run on the access authentication device, causing the access authentication device to perform the steps of the method of any one of claims 1 to 7.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202310660056.9A CN116403203B (en) | 2023-06-06 | 2023-06-06 | Label generation method, system, electronic equipment and storage medium |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202310660056.9A CN116403203B (en) | 2023-06-06 | 2023-06-06 | Label generation method, system, electronic equipment and storage medium |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN116403203A true CN116403203A (en) | 2023-07-07 |
| CN116403203B CN116403203B (en) | 2023-08-29 |
Family
ID=87016429
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202310660056.9A Active CN116403203B (en) | 2023-06-06 | 2023-06-06 | Label generation method, system, electronic equipment and storage medium |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN116403203B (en) |
Citations (16)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20140281871A1 (en) * | 2013-03-15 | 2014-09-18 | Meditory Llc | Method for mapping form fields from an image containing text |
| CN108052408A (en) * | 2017-12-14 | 2018-05-18 | 杭州富春云科技有限公司 | Alarm work order based on warning information content quickly submits method and system |
| CN108229299A (en) * | 2017-10-31 | 2018-06-29 | 北京市商汤科技开发有限公司 | The recognition methods of certificate and device, electronic equipment, computer storage media |
| CN108388831A (en) * | 2018-01-10 | 2018-08-10 | 链家网（北京）科技有限公司 | A kind of identification of spare part and finish message method and device |
| US20200050845A1 (en) * | 2018-08-13 | 2020-02-13 | International Business Machines Corporation | Method and system for extracting information from an image of a filled form document |
| CN112800848A (en) * | 2020-12-31 | 2021-05-14 | 中电金信软件有限公司 | Structured extraction method, device and equipment of information after bill identification |
| CN113065557A (en) * | 2021-04-16 | 2021-07-02 | 潍坊工程职业学院 | Image matching method based on character extraction |
| CN113920335A (en) * | 2021-09-28 | 2022-01-11 | 苏州冷王网络科技有限公司 | Deep learning-based image and text embedded drug label identification method |
| US20220027611A1 (en) * | 2021-03-03 | 2022-01-27 | Beijing Baidu Netcom Science Technology Co., Ltd. | Image classification method, electronic device and storage medium |
| US20220044058A1 (en) * | 2020-08-07 | 2022-02-10 | Salesforce.Com, Inc. | Template-Based Key-Value Extraction for Inferring OCR Key Values Within Form Images |
| CN114117171A (en) * | 2021-11-16 | 2022-03-01 | 华中师范大学 | Intelligent project file collecting method and system based on energized thinking |
| US20220092276A1 (en) * | 2020-09-22 | 2022-03-24 | Samsung Electronics Co., Ltd. | Multimodal translation method, apparatus, electronic device and computer-readable storage medium |
| CN114821623A (en) * | 2022-05-09 | 2022-07-29 | 深圳市商汤科技有限公司 | Document processing method and device, electronic equipment and storage medium |
| CN115527220A (en) * | 2022-10-11 | 2022-12-27 | 太保科技有限公司 | Document image classification method and device |
| US20220415012A1 (en) * | 2021-06-24 | 2022-12-29 | Walmart Apollo, Llc | Image processing based methods and apparatus for planogram compliance |
| CN115810197A (en) * | 2022-12-19 | 2023-03-17 | 北京中电普华信息技术有限公司 | Multi-mode electric power form recognition method and device |
-
2023
- 2023-06-06 CN CN202310660056.9A patent/CN116403203B/en active Active
Patent Citations (16)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20140281871A1 (en) * | 2013-03-15 | 2014-09-18 | Meditory Llc | Method for mapping form fields from an image containing text |
| CN108229299A (en) * | 2017-10-31 | 2018-06-29 | 北京市商汤科技开发有限公司 | The recognition methods of certificate and device, electronic equipment, computer storage media |
| CN108052408A (en) * | 2017-12-14 | 2018-05-18 | 杭州富春云科技有限公司 | Alarm work order based on warning information content quickly submits method and system |
| CN108388831A (en) * | 2018-01-10 | 2018-08-10 | 链家网（北京）科技有限公司 | A kind of identification of spare part and finish message method and device |
| US20200050845A1 (en) * | 2018-08-13 | 2020-02-13 | International Business Machines Corporation | Method and system for extracting information from an image of a filled form document |
| US20220044058A1 (en) * | 2020-08-07 | 2022-02-10 | Salesforce.Com, Inc. | Template-Based Key-Value Extraction for Inferring OCR Key Values Within Form Images |
| US20220092276A1 (en) * | 2020-09-22 | 2022-03-24 | Samsung Electronics Co., Ltd. | Multimodal translation method, apparatus, electronic device and computer-readable storage medium |
| CN112800848A (en) * | 2020-12-31 | 2021-05-14 | 中电金信软件有限公司 | Structured extraction method, device and equipment of information after bill identification |
| US20220027611A1 (en) * | 2021-03-03 | 2022-01-27 | Beijing Baidu Netcom Science Technology Co., Ltd. | Image classification method, electronic device and storage medium |
| CN113065557A (en) * | 2021-04-16 | 2021-07-02 | 潍坊工程职业学院 | Image matching method based on character extraction |
| US20220415012A1 (en) * | 2021-06-24 | 2022-12-29 | Walmart Apollo, Llc | Image processing based methods and apparatus for planogram compliance |
| CN113920335A (en) * | 2021-09-28 | 2022-01-11 | 苏州冷王网络科技有限公司 | Deep learning-based image and text embedded drug label identification method |
| CN114117171A (en) * | 2021-11-16 | 2022-03-01 | 华中师范大学 | Intelligent project file collecting method and system based on energized thinking |
| CN114821623A (en) * | 2022-05-09 | 2022-07-29 | 深圳市商汤科技有限公司 | Document processing method and device, electronic equipment and storage medium |
| CN115527220A (en) * | 2022-10-11 | 2022-12-27 | 太保科技有限公司 | Document image classification method and device |
| CN115810197A (en) * | 2022-12-19 | 2023-03-17 | 北京中电普华信息技术有限公司 | Multi-mode electric power form recognition method and device |
Non-Patent Citations (2)
| Title |
|---|
| PRANJAL DHAKAL ET AL.: "One-Shot Template Matching for Automatic Document Data Capture", 《2019 ARTIFICIAL INTELLIGENCE FOR TRANSFORMING BUSINESS AND SOCIETY (AITB)》, pages 1 - 6 * |
| 曾紫薇: "基于OCR技术的发票验证小程序设计与实现", 《中国优秀硕博学位论文全文数据库 信息科技辑》, pages 138 - 603 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN116403203B (en) | 2023-08-29 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| US11715313B2 (en) | Apparatus and methods for extracting data from lineless table using delaunay triangulation and excess edge removal | |
| US10726198B2 (en) | Method, device, and system, for identifying data elements in data structures | |
| CN108391446B (en) | Automatic extraction of training corpus for data classifier based on machine learning algorithm | |
| CN109783635A (en) | Use machine learning and fuzzy matching AUTOMATIC ZONING classifying documents and identification metadata | |
| CN112800848A (en) | Structured extraction method, device and equipment of information after bill identification | |
| US20210209289A1 (en) | Method and apparatus for generating customized content based on user intent | |
| US10699112B1 (en) | Identification of key segments in document images | |
| US20230206670A1 (en) | Semantic representation of text in document | |
| CN107679208A (en) | A kind of searching method of picture, terminal device and storage medium | |
| CN110874534B (en) | Data processing method and data processing device | |
| Ha et al. | Recognition of OCR invoice metadata block types | |
| CN110147545A (en) | The structuring output method and system of text, storage medium and computer equipment | |
| Romero et al. | Modern vs diplomatic transcripts for historical handwritten text recognition | |
| CN111753535A (en) | Method and device for generating patent application text | |
| CN110738050A (en) | Text recombination method, device and medium based on word segmentation and named entity recognition | |
| CN114022891A (en) | Method, device and equipment for extracting key information of scanned text and storage medium | |
| CN111274799B (en) | Article title generation method, device and storage medium | |
| CN116403203B (en) | Label generation method, system, electronic equipment and storage medium | |
| CN114090630A (en) | Commodity data integration method based on distributed micro-service cluster | |
| JP5510221B2 (en) | Information processing apparatus and information processing program | |
| US20230308731A1 (en) | Method for providing service of producing multimedia conversion content by using image resource matching, and apparatus thereof | |
| US11315351B2 (en) | Information processing device, information processing method, and information processing program | |
| CN116798417B (en) | Voice intention recognition method, device, electronic equipment and storage medium | |
| CN110837740B (en) | Comment aspect opinion level mining method based on dictionary improvement LDA model | |
| EP4336379A1 (en) | Tracking concepts within content in content management systems and adaptive learning systems |
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 | ||
| PE01 | Entry into force of the registration of the contract for pledge of patent right | ||
| PE01 | Entry into force of the registration of the contract for pledge of patent right |
Denomination of invention: A label generation method, system, electronic device, and storage medium Granted publication date: 20230829 Pledgee: Bank of China Limited Wuhan Donghu New Technology Development Zone Branch Pledgor: WUHAN JINGCHEN WISDOM LOGO TECHNOLOGY Co.,Ltd. Registration number: Y2024980008335 |