WO2023022404A1 - Procédé d'identification d'image de code-barres et dispositif associé - Google Patents

Procédé d'identification d'image de code-barres et dispositif associé Download PDF

Info

Publication number
WO2023022404A1
WO2023022404A1 PCT/KR2022/011386 KR2022011386W WO2023022404A1 WO 2023022404 A1 WO2023022404 A1 WO 2023022404A1 KR 2022011386 W KR2022011386 W KR 2022011386W WO 2023022404 A1 WO2023022404 A1 WO 2023022404A1
Authority
WO
WIPO (PCT)
Prior art keywords
barcode
barcode image
image
decoding
model
Prior art date
Application number
PCT/KR2022/011386
Other languages
English (en)
Korean (ko)
Inventor
강창범
왕신조
Original Assignee
주식회사 에너자이
Priority date (The priority date 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 date listed.)
Filing date
Publication date
Priority claimed from KR1020210108491A external-priority patent/KR102393764B1/ko
Application filed by 주식회사 에너자이 filed Critical 주식회사 에너자이
Publication of WO2023022404A1 publication Critical patent/WO2023022404A1/fr

Links

Images

Classifications

    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06K—GRAPHICAL DATA READING; PRESENTATION OF DATA; RECORD CARRIERS; HANDLING RECORD CARRIERS
    • G06K7/00—Methods or arrangements for sensing record carriers, e.g. for reading patterns
    • G06K7/10—Methods or arrangements for sensing record carriers, e.g. for reading patterns by electromagnetic radiation, e.g. optical sensing; by corpuscular radiation
    • G06K7/14—Methods or arrangements for sensing record carriers, e.g. for reading patterns by electromagnetic radiation, e.g. optical sensing; by corpuscular radiation using light without selection of wavelength, e.g. sensing reflected white light
    • G06K7/1404—Methods for optical code recognition
    • G06K7/1408—Methods for optical code recognition the method being specifically adapted for the type of code
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F18/00—Pattern recognition
    • G06F18/10—Pre-processing; Data cleansing
    • 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
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T5/00—Image enhancement or restoration

Definitions

  • the present invention relates to barcode image identification and an apparatus therefor, and more particularly, to a method and apparatus capable of identifying a damaged barcode image as well as a normal barcode image.
  • a barcode is a combination of vertical bars with different thicknesses to represent alphanumeric characters or letters in a machine-readable form, which can be read optically. It is widely used for various purposes such as product management. Barcodes are composed of various types, such as one-dimensional barcodes such as EAN, CODE39, OCDE93, CODE128, and ISBN, and two-dimensional barcodes such as PDF-417, QR CODE, and 2D Data Matrix, and their recognition methods are all similar.
  • a camera and an infrared scanner are widely used in a conventional barcode recognition device, and the infrared scanner can read a printed barcode, decode data included in the barcode, and send the data to a computer.
  • infrared scanners require accurate alignment of barcodes in the recognition area, and in the case of cameras, performance is degraded due to uneven lighting conditions, barcode distortion, camera performance, and barcode damage due to dust or graffiti.
  • machine learning models have been introduced to barcode recognition devices, but even in the case of machine learning models, they have the same disadvantages as conventional infrared scanners, and furthermore, only those formats that can be supported by the model among various barcode formats can be used, limiting their functions.
  • An object of the present invention is to solve the above problems, and to facilitate interpretation of a barcode image by applying a model for correcting a damaged barcode image.
  • the present invention provides a method for identifying a barcode image by an electronic device, comprising step a of receiving an image including at least one barcode, step b of extracting a first barcode image from the image, and step c of decoding the barcode image; step d of generating a second barcode image by inputting the first barcode image to a preprocessing model if step c fails; and step e of decoding the second barcode image.
  • the present invention provides a barcode input module that receives an image including at least one barcode, a barcode extraction module that extracts a first barcode image from the image, and decodes the first barcode image. It is characterized in that it includes a barcode decoding module for generating a second barcode image by inputting a barcode image and decoding the second barcode image.
  • the present invention as described above, it is possible to facilitate interpretation of a barcode image by applying a model for correcting a damaged barcode image.
  • the damaged barcode image can be automatically restored to the original barcode image without the user's help, and data on the barcode image can be identified and provided to the user.
  • FIG. 1 is a diagram showing the configuration of a barcode image identification device according to an embodiment of the present invention
  • FIG. 2 is a flowchart illustrating a barcode image identification method according to an embodiment of the present invention
  • FIG. 3 is a flowchart illustrating a method of learning a barcode correction model according to an embodiment of the present invention.
  • a barcode image identification method includes a step of receiving an image including at least one barcode; step b of extracting a first barcode image from the image; step c of decoding the first barcode image; step d of generating a second barcode image by inputting the first barcode image to a preprocessing model if step c fails; and step e of decoding the second barcode image.
  • a first barcode image may be extracted from the image using an object recognition model.
  • the first barcode image may include a pattern in which foreground color and background color areas are formed at specific intervals.
  • the first barcode image may be converted into binary numbers according to a preset rule, and data corresponding to the first barcode image may be extracted based on the converted binary numbers.
  • the preprocessing model may include loading a third barcode image stored in a database; generating a fourth barcode image by adding random noise to the third barcode image through an image augmentation technique; and learning to output a third barcode image when a fourth barcode image is input based on the third barcode image and the fourth barcode image.
  • a second barcode image may be generated by inputting the first barcode image to the preprocessing model.
  • a barcode image identification device includes a barcode input module that receives an image including at least one barcode; a barcode extraction module extracting a first barcode image from the image; and a barcode decoding module that decodes the first barcode image and, if decoding fails, inputs the first barcode image to a preprocessing model to generate a second barcode image and decodes the second barcode image.
  • each component may be implemented as a hardware processor, and the above components may be integrated and implemented as one hardware processor, or the above components may be combined with each other and implemented as a plurality of hardware processors.
  • FIG. 1 is a perspective view showing the configuration of a barcode recognition device according to an embodiment of the present invention.
  • a barcode recognition device 1 includes a barcode input module 10, a barcode output module 20, a barcode decoding module 30, a model generation module 40, storage module 50.
  • the barcode input module 10 may receive an image including at least one barcode from the user terminal 2 or any external device. For example, a user may photograph a barcode printed on a product, paper, leaflet, etc. through the user terminal 2, and the user terminal 2 may transmit the photographed image to the barcode recognition device 1.
  • the barcode extraction module 20 may extract a first barcode image from an input image.
  • the barcode extraction module 20 may recognize a region including a barcode included in an image through a conventionally used object recognition model, and extract a first barcode image from the region.
  • the first barcode image means that a pattern in which the foreground color area and the background color area are formed at specific intervals is included, and examples thereof include EAN, UPC, CODABAR, and QR CODE.
  • the foreground color may be black and the background color may be white.
  • the barcode extraction module 20 does not recognize barcodes through a conventional rule-based algorithm, but rather applies a machine learning-based object recognition model to recognize barcodes even in situations where barcode recognition is difficult to obtain a first barcode image. can be extracted.
  • the barcode extraction module 20 may improve recognition accuracy of a barcode image through an object recognition model.
  • the barcode decryption module 30 may decode the first barcode image. In decoding the first barcode image, the barcode decoding module 30 converts the first barcode image into binary numbers according to a preset rule, and then converts data corresponding to the first barcode image based on the converted binary number value. can be extracted. At this time, data corresponding to the barcode image may be stored in the storage module 50 .
  • the barcode decoding module 30 can also apply a machine learning-based model, but when using the model, the barcode decoding module 30 is limited to barcodes in a format that can be identified by the model. It can be based on the base algorithm.
  • the barcode decoding module 30 may determine that decoding of the first barcode image has failed when data corresponding to the first barcode image is not stored in the storage module 50 .
  • the barcode decoding module 30 may generate a second barcode image by applying a barcode correction model to the first barcode image.
  • the second barcode image means that the first barcode image has been preprocessed and converted to be readable.
  • the barcode decryption module 30 may decode the second barcode image.
  • the process of decoding the second barcode image by the barcode decoding module 30 is the same as the process of decoding the first barcode image.
  • the model generation module 40 may generate a barcode correction model for preprocessing the undecipherable first barcode image.
  • the model generating module 40 may generate a training data set prior to generating a barcode correction model.
  • the model generation module 40 may generate a fourth barcode image by generating random noise in a plurality of third barcode images stored in the storage module 50 or received by a manager terminal (not shown).
  • the model generation module 40 may utilize image augmentation techniques in generating noise.
  • the model generation module 40 may generate a training data set based on the third barcode image and the fourth barcode image, and may train a barcode correction model with the generated training data set.
  • the model generating module 40 may generate a contaminated fourth barcode image by setting a barcode image successfully decoded among the first barcode images received by the barcode input module 10 as a third barcode image.
  • the model generation module 40 may continuously strengthen the barcode correction model by adding the barcode image successfully decrypted as described above to the training data set.
  • the storage module 50 may store various data used in the barcode recognition device 1 according to an embodiment of the present invention.
  • the storage module 50 may temporarily or semi-permanently store various types of data. Examples of the storage module 50 include a hard disk drive (HDD), a solid state drive (SSD), a flash memory, a read-only memory (ROM), and a random access memory (RAM). ), etc. may be present.
  • the storage module 50 may be provided in a form embedded in the barcode recognition device 1 or in a detachable form.
  • the storage module 50 includes an operating system (OS) for driving the barcode recognition device 1 or a program for operating each component of the barcode recognition device 1, as well as information about the operation of the barcode recognition device 1.
  • OS operating system
  • Various types of data such as a required barcode correction model and a training data set for a preprocessing model, may be stored.
  • FIG. 2 is a flowchart illustrating a barcode image identification method according to an embodiment of the present invention.
  • a barcode image identification method will be described with reference to FIG. 2 .
  • detailed embodiments overlapping with the barcode image identification device described above may be omitted.
  • the electronic device may receive an image including at least one barcode from the user terminal 2 or any external device.
  • the electronic device may extract a first barcode image from the input image.
  • the electronic device may recognize an area including a barcode included in an image through a conventionally used object recognition model, and extract a first barcode image from the area.
  • the electronic device may decode the first barcode image.
  • the electronic device converts the first barcode image into binary numbers according to a preset rule, and then extracts data corresponding to the first barcode image based on the converted binary number value. .
  • data corresponding to the barcode image may be stored in the storage module 50 .
  • step 130 when data corresponding to the first barcode image is not stored in the storage module 50, the electronic device may determine that decoding of the first barcode image has failed. If decoding of the first barcode image fails, the electronic device may generate a second barcode image by applying a barcode correction model to the first barcode image.
  • the second barcode image means that the first barcode image has been preprocessed and converted to be readable.
  • the present invention can correct the first barcode image that is damaged and cannot be decoded to generate a second barcode image, thereby enabling decryption of the damaged first barcode image.
  • FIG. 3 is a flowchart illustrating a method of generating a barcode correction model according to an embodiment of the present invention, and a barcode pretreatment model used in step 140 may be generated through FIG. 3 .
  • the electronic device may generate a learning data set to learn the barcode correction model.
  • the electronic device may generate a fourth barcode image by generating random noise in a plurality of third barcode images stored in the storage module 50 or received by a manager terminal (not shown).
  • the electronic device may utilize an image augmentation technique in generating the fourth barcode image.
  • the electronic device may generate a training data set based on the third barcode image and the fourth barcode image.
  • the electronic device may train a barcode correction model with the training data set generated in step 300. Furthermore, the electronic device may set the first barcode image successfully decoded in step 120 of FIG. 2 as a third barcode image, generate a fourth contaminated barcode image corresponding to the first barcode image, and further add it to the training data set. there is.
  • step 140 the electronic device may decode the second barcode image. This step may be performed in the same manner as step 120 in which the electronic device decodes the first barcode image.
  • step 150 the electronic device may provide data corresponding to the first barcode image to the user terminal 2.

Landscapes

  • Engineering & Computer Science (AREA)
  • Theoretical Computer Science (AREA)
  • Physics & Mathematics (AREA)
  • General Physics & Mathematics (AREA)
  • Artificial Intelligence (AREA)
  • Computer Vision & Pattern Recognition (AREA)
  • General Health & Medical Sciences (AREA)
  • General Engineering & Computer Science (AREA)
  • Health & Medical Sciences (AREA)
  • Life Sciences & Earth Sciences (AREA)
  • Data Mining & Analysis (AREA)
  • Evolutionary Computation (AREA)
  • Evolutionary Biology (AREA)
  • Bioinformatics & Cheminformatics (AREA)
  • Bioinformatics & Computational Biology (AREA)
  • Electromagnetism (AREA)
  • Toxicology (AREA)
  • Molecular Biology (AREA)
  • Computing Systems (AREA)
  • Biomedical Technology (AREA)
  • Mathematical Physics (AREA)
  • Software Systems (AREA)
  • Biophysics (AREA)
  • Computational Linguistics (AREA)
  • Image Analysis (AREA)

Abstract

La présente invention concerne une identification d'image de code-barres et un dispositif associé, et un but de la présente invention est de faciliter l'interprétation d'une image de code-barres en appliquant un modèle pour corriger une image de code-barres endommagée. Pour atteindre un tel but, la présente invention comprend : une étape a consistant à recevoir une entrée d'une image comprenant au moins un code-barres ; une étape b consistant à extraire une première image de code-barres de l'image ; une étape c consistant à déchiffrer la première image de code-barres ; une étape d consistant à générer une seconde image de code-barres en entrant la première image de code-barres dans un modèle de prétraitement, si l'étape c échoue ; et une étape e consistant à déchiffrer la seconde image de code-barres.
PCT/KR2022/011386 2021-08-18 2022-08-02 Procédé d'identification d'image de code-barres et dispositif associé WO2023022404A1 (fr)

Applications Claiming Priority (4)

Application Number Priority Date Filing Date Title
KR1020210108491A KR102393764B1 (ko) 2021-08-18 2021-08-18 바코드 이미지 식별 방법 및 그 장치
KR10-2022-0052714 2021-08-18
KR10-2021-0108491 2021-08-18
KR1020220052714A KR102682942B1 (ko) 2021-08-18 2022-04-28 바코드 이미지 식별 방법 및 그 장치

Publications (1)

Publication Number Publication Date
WO2023022404A1 true WO2023022404A1 (fr) 2023-02-23

Family

ID=85241153

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/KR2022/011386 WO2023022404A1 (fr) 2021-08-18 2022-08-02 Procédé d'identification d'image de code-barres et dispositif associé

Country Status (2)

Country Link
KR (1) KR102682942B1 (fr)
WO (1) WO2023022404A1 (fr)

Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JP2012103825A (ja) * 2010-11-09 2012-05-31 Denso Wave Inc 光学的情報読取装置
KR20200050091A (ko) * 2018-10-31 2020-05-11 삼성전자주식회사 바코드를 판독하는 방법 및 전자 장치
CN111767750A (zh) * 2019-05-27 2020-10-13 北京沃东天骏信息技术有限公司 图像处理方法和装置
KR102263184B1 (ko) * 2020-12-14 2021-06-09 주식회사 대곤코퍼레이션 저 품질 바코드 판독시스템
KR102393764B1 (ko) * 2021-08-18 2022-05-06 주식회사 에너자이(ENERZAi) 바코드 이미지 식별 방법 및 그 장치

Family Cites Families (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
KR101632380B1 (ko) * 2009-10-19 2016-06-21 삼성전자주식회사 바코드 인식 장치 및 방법
KR20190119548A (ko) * 2019-10-02 2019-10-22 엘지전자 주식회사 이미지 노이즈 처리방법 및 처리장치

Patent Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JP2012103825A (ja) * 2010-11-09 2012-05-31 Denso Wave Inc 光学的情報読取装置
KR20200050091A (ko) * 2018-10-31 2020-05-11 삼성전자주식회사 바코드를 판독하는 방법 및 전자 장치
CN111767750A (zh) * 2019-05-27 2020-10-13 北京沃东天骏信息技术有限公司 图像处理方法和装置
KR102263184B1 (ko) * 2020-12-14 2021-06-09 주식회사 대곤코퍼레이션 저 품질 바코드 판독시스템
KR102393764B1 (ko) * 2021-08-18 2022-05-06 주식회사 에너자이(ENERZAi) 바코드 이미지 식별 방법 및 그 장치

Non-Patent Citations (1)

* Cited by examiner, † Cited by third party
Title
JEONG YONG-BAE, JEONG DONG-SEUK, YANG, CHO HAN-JIN, LEE JUNE-HWAN: "The Study of Fast mail recognition and development of image error correction technology using 4-state barcode", PROCEEDINGS OF THE KOREA CONTENTS ASSOCIATION CONFERENCE, 1 May 2011 (2011-05-01), pages 573 - 574, XP093036916, Retrieved from the Internet <URL:https://www.dbpia.co.kr/journal/articleDetail?nodeId=NODE06096618> [retrieved on 20230403] *

Also Published As

Publication number Publication date
KR102682942B1 (ko) 2024-07-09
KR20230028126A (ko) 2023-02-28

Similar Documents

Publication Publication Date Title
US11210510B2 (en) Storing anonymized identifiers instead of personally identifiable information
US11349666B2 (en) Electronically signing and distributing identification data as a service that provides proof of identity, integrity, validity and origin of data for non-repudiation and ID validation methods
US20210042488A1 (en) Method and apparatus for processing encoded pattern, storage medium, and electronic apparatus
US8430301B2 (en) Document authentication using hierarchical barcode stamps to detect alterations of barcode
US7936901B2 (en) System and method for encoding high density geometric symbol set
US6895116B2 (en) Automatically extracting graphical bar codes
WO2014044159A1 (fr) Procédé et dispositif de lecture par lots de codes à barres 2d
JP2000511320A (ja) 光学的文字認識（ｏｃｒ）支援のバーコード解読システム及び方法
US7305131B2 (en) Extracting graphical bar codes from an input image
US20170200247A1 (en) Systems and methods for authentication of physical features on identification documents
CN112508145B (zh) 电子印章生成及验证方法、装置、电子设备及存储介质
CN105787403A (zh) 一种高像素图像的条码识读方法及设备
KR102393764B1 (ko) 바코드 이미지 식별 방법 및 그 장치
WO2023022404A1 (fr) Procédé d&#39;identification d&#39;image de code-barres et dispositif associé
CN113553876A (zh) 条形码识别方法、装置、设备及存储介质
KR102085561B1 (ko) Qr코드 위변조 방지장치 및 방법
KR101685161B1 (ko) 여권 정보 변환 장치 및 이를 이용한 변환 방법
Mantoro et al. Real-time printed document authentication using watermarked qr code
WO2024136304A1 (fr) Procédé d&#39;insertion de code de sécurité holographique dans un support d&#39;enregistrement holographique et procédé de lecture de code de sécurité holographique
KR102514817B1 (ko) 딥러닝 기반의 시각적 문서 이해 시스템 및 방법
CN110866578B (zh) 一种二维纹理码及其编码和解码方法
Noppakaew et al. Construction of multi-layered QR codes utilizing partitions of positive integers
JP2002216073A (ja) 可読文字等の読取り装置及び可読文字等の読取り方法
Gupta et al. Solving QR Code Distortions using a Recursive-based Backtracking Algorithm
JPH06111060A (ja) 光学的文字読取装置

Legal Events

Date Code Title Description
121 Ep: the epo has been informed by wipo that ep was designated in this application

Ref document number: 22858642

Country of ref document: EP

Kind code of ref document: A1

NENP Non-entry into the national phase

Ref country code: DE