Optical Character Recognition based KYC System
0 likes37 views
This document proposes an optical character recognition (OCR) based Know Your Customer (KYC) system to automate the KYC process. The system extracts data from identity documents like Aadhar and PAN cards using OCR, verifies the documents, and auto-fills KYC forms. This saves customers time by reducing manual form filling and errors. The system was built using Python libraries for OCR, data extraction and storage, and connecting to external KYC forms. Test results showed the system successfully mapped extracted user details to pre-defined forms to complete the KYC process in a few clicks.
![International Research Journal of Engineering and Technology (IRJET) e-ISSN: 2395-0056
Volume: 09 Issue: 02 | Feb 2022 www.irjet.net p-ISSN: 2395-0072
© 2022, IRJET | Impact Factor value: 7.529 | ISO 9001:2008 Certified Journal | Page 877
6. RESULTS
Figure 6.1
Figure 6.2
7. CONCLUSIONS
The keyboard is the most common way to enter data into a
computer. However, this isn't always the best or most
efficient solution. The goal of this system is to develop a new
KYC system that allows users to complete KYC ina fewclicks
and includes verification and auto-formfillingwithuserdata
that is precisely mapped to the form. Although many
processes are still carried out on paper, it is clear that
automatic data recognition technologies are gaining
popularity. During subsequent processing steps, the
document is repeatedly copied and changed, resulting in a
large number of copies. In somecases,theycanhelphumans,
but in others, they are useless. The aim of this system is to
successfully enter data into the form by accurately mapping
every user detail from the database to the form field using
OCR technology without compromising data quality.
8. ACKNOWLEDGEMENT
We would like to express our heartfelt gratitude to our
guide, Professor Sejal D'mello, for her motivation and
guidance throughout the process. We would also like to
thank the faculty of the InformationTechnologyDepartment
for their valuable assistance with our project.
9. REFERENCES
[1] Rishabh Mittal and Anchal Garg, “Text extraction using
OCR: A Systematic Review.” IEEE (2020)
[2] S. Tomovic, K. Pavlovic, M. Bajceta, “Aligning document
layouts extracted with different OCR engines with
clustering approach.” Science Direct (2020)
[3] Yash Kumar, Gaurav Sharma, Komal, Prof. Audumbar
Umbare “E-KYC Mobile Application using Optical
Character Recognition”. IRJET (2020)
[4] Hejing Wu, Fang Liu, Long Zhao Yabin Shao, “Data
Analysis andCrawlerApplicationImplementationBased
on Python.” IEEE 2020
[5] S. D. Bandari, Ankita Jagtap, Namrata Mane, Swarali
Garud “Intelligent Framework for Auto filling Web form
using Scanned Documents” IJISRT
[6] A Al Mamun, SR Hasan, MS Bhuiyan, M Shamim Kaisar,
Mohammad Abu Yousuf SecureandTransparentKYCfor
Banking System Using IPFS and Blockchain Technology.
IEEE (2020)](https://image.slidesharecdn.com/irjet-v9i2149-220913100916-838a98e0/85/Optical-Character-Recognition-based-KYC-System-3-320.jpg)
Optical Character Recognition based KYC System
- 1. International Research Journal of Engineering and Technology (IRJET) e-ISSN: 2395-0056 Volume: 09 Issue: 02 | Feb 2022 www.irjet.net p-ISSN: 2395-0072 © 2022, IRJET | Impact Factor value: 7.529 | ISO 9001:2008 Certified Journal | Page 875 Optical Character Recognition based KYC System Manasi Bhabal1, Dhruv Desai2, Harsh Desai3, Prof. Sejal D’mello4 1,2,3B.E. Student, Information Technology Engineering, Atharva College of Engineering, Mumbai, India 4Professor, Information Technology Engineering, Atharva College of Engineering, Mumbai, India -----------------------------------------------------------------------***-------------------------------------------------------------------- Abstract - The aim of this paper is to build a platform using Machine Learning and Optical CharacterRecognitiontomake it easier for customers to complete their KYC (Know Your Customer) for banking purposes such as opening a bank account, applying for a policy, and various other business purposes, for the sake of convenience and time savings. This will not only significantly speed up the KYC process, but it will also eliminate errors. Customers can update their KYC in our system by scanning and uploading their AADHAR and PAN cards. The app will use OCR (OpticalCharacterRecognition) to reduce typing errors, verify documents, and thus auto-fill the form without errors, saving time. Key Words: KYC (Know your customer), OCR (Optical Character Recognition), Tesseract, Authentication, Data Extraction. 1.INTRODUCTION Know Your Customer (KYC) is one of the most important processes for any businesses and digitalization of this process is need of the hour. Almost everything in the digital era is automated, and information is stored and communicated in digital formats. KYC can also refer to regulated banking practices that are used to verify clients' identities. The traditional method of working for KYC is a bit tedious and takes a long time to complete. The proposed system is a software-basedanddependableKYCmethodthat uses the concept of OCR (Optical Character Recognition) to extract text from images, verifies it, and then stores the data in a database. Additionally, auto form filling is implemented in the project. 1.1 PURPOSE Customers typically fill out forms on paper sheets by hand. Because of human errors, unclear handwriting,anddefective writing materials, this might lead to a lot of discrepancy.This could result in a significant amount of paper waste (considering this is a process adopted worldwide). Next, there could be errorsproducedbytheauthoritiesresponsible for data sentry, while referring the handwritten form. Even the current digitalKYCsystemrequiresmanualeffortsindata entry. Furthermore, because of the manner customers enter their data, there may be inconsistenciesinthecustomerdata, generating serious validation issues for the customer whose primary goal was to authenticate their identity in the first place. Customer identification also aids in the control of financial fraud, the detection of money laundering and suspicious activities, and the scrutiny and monitoring of large cash transactions. To avoid these problems, the Reserve Bank of India (RBI) directed all banks and financial institutions in India to implement a policy framework that requires themto know their customers before opening any accounts. This entails verifying customers' identities and addresses by requesting documents that are accepted as relevant proof. KYC norms require proof of identityandproofofresidenceas mandatory details. Proof of identitycanbeapassport,voter's ID card, Permanent Account Number (PAN) card, or driving license, and proof of residence can be a ration card, an electricity or telephone bill, or a letter from the employer or any recognized public authority certifying the address, in addition to proof of identity being used as residence proof if they carry address. 1.2 OBJECTIVES • To ease the process of KYC for various organizations. • To create an E-KYC system and make it convenient for users to complete their KYC. • To extract data from the uploaded documents of users. • To verify and authenticate the uploadeddocuments • To auto fill the user form and complete the process. • Making a database of the information gathered for each customer for further use. • Driving down cost of operations by automating the whole process. • By increasing customer satisfaction in a number of different functional operations across the organization. 2. FEASIBILITY 2.1. Technical Feasibility Thetechnical resources(hardwareandsoftware) requiredto build the project are the focus of technical feasibility. It also investigates the specifics of how you intend to deliver the product and whether the technical team is capable of
- 2. International Research Journal of Engineering and Technology (IRJET) e-ISSN: 2395-0056 Volume: 09 Issue: 02 | Feb 2022 www.irjet.net p-ISSN: 2395-0072 © 2022, IRJET | Impact Factor value: 7.529 | ISO 9001:2008 Certified Journal | Page 876 translating the concept into working systems. In our project, weare creating a website with HTML and CSS and storing the extracted data in an SQLite database. In addition, we use Python libraries to implement the algorithm. Other than a laptop, no additional hardware is required for this project. 2.2. Economic Feasibility The economic feasibility of a project is determined by the costs incurred in the project's development. HTML and CSS were used to build this site. It is a costless language. This project does not necessitate the use of any additional hardware. Asit ispurelya software project,wedidnothaveto spend any money on it. 2.3. Operational Feasibility It evaluateshow well a proposedsystemsolves problemsand meets the requirements identified during the requirement analysis phase. The primary goal of this project is to make the KYC process as simple as possible for employees while reducing the risks associated with the manual process. This project implements KYC in a single click with form automation, which is superior to traditional KYC methods. 3. PROPOSED SYSTEM The flow chart in figure 2.1. illustrates how the system works. First, the user registers in the system, after whichthe home page is displayed. Next, the user crops and uploads images of an Aadhar card and a Pan Card, and the data is extracted using OCR. Following data extraction, each uploaded document is authenticated by a verification process, and a KYC form is automatically filled out and the process is terminated. Fig-2.1. Flowchart 4. METHODOLOGY Data extraction, authentication of ID proofs, and auto-filling KYC forms are the three primary phases in the system. Data from the customer's documents is extracted and mapped, then saved in a database for future auto-form filling. As soon as the customerpresentshisdocuments,thesystemidentifies the relevant elements in the image. To auto-fill the form, the extracted data from the photos is checked first, then mapped to the form fields. After the form has been correctly filled,the customer is notified through e-mail about successful completion of KYC. Data is extracted from images using the Tesseract Python library. The algorithm mapsfieldsonimageswheredata isto be fetched from. Once this data is fetched, the extracted data is displayed to the user to ensure its accuracy. Furthermore, the retrieveddata ischeckedtoconfirmthatthe provided documents are authentic. To obtain the API for verifyingIndianAadharcards,anofficialapplicationshouldbe submitted to UIDAI at https://www.uidai.gov.in/. The user clicks the "Fill Form" button after the obtained data has been verified. The KYC form is auto-filled, and a Selenium web- driver is utilized to connect an external form to our system in order to accomplish this. Tesseract is a text recognition (OCR) engine that is open source and licensed under the Apache 2.0 license. It can be used directly or through an API (for programmers) to extract written text from photos. It helps us with a wide range of languages. Tesseract doesn't come with a graphical user interface, however there are a few on the 3rdParty website. Tesseract can be used with a variety of computer languages and frameworks thanks to wrappers available. It can be used in conjunction with the existing layout analysis to recognize text insidea hugedocument, or withan external textdetector to recognize text from a single text line image. 5. SCOPE KYC is performed using a few required government documents. Thissystemis notintendedforalldocuments.The KYC is performed by extracting relevant details from the predefineddocuments. Furthermore, theformis pre-defined, and the extracted details are eventually mapped to the form. Thesystem can beimprovedfurtherto be moresecure and to perform KYC with more documents. Basic data is gathered, and software generally compares it to lists of individuals known for corruption, sanctioned, suspected of committing a crime, or at high risk of bribery or money laundering.
- 3. International Research Journal of Engineering and Technology (IRJET) e-ISSN: 2395-0056 Volume: 09 Issue: 02 | Feb 2022 www.irjet.net p-ISSN: 2395-0072 © 2022, IRJET | Impact Factor value: 7.529 | ISO 9001:2008 Certified Journal | Page 877 6. RESULTS Figure 6.1 Figure 6.2 7. CONCLUSIONS The keyboard is the most common way to enter data into a computer. However, this isn't always the best or most efficient solution. The goal of this system is to develop a new KYC system that allows users to complete KYC ina fewclicks and includes verification and auto-formfillingwithuserdata that is precisely mapped to the form. Although many processes are still carried out on paper, it is clear that automatic data recognition technologies are gaining popularity. During subsequent processing steps, the document is repeatedly copied and changed, resulting in a large number of copies. In somecases,theycanhelphumans, but in others, they are useless. The aim of this system is to successfully enter data into the form by accurately mapping every user detail from the database to the form field using OCR technology without compromising data quality. 8. ACKNOWLEDGEMENT We would like to express our heartfelt gratitude to our guide, Professor Sejal D'mello, for her motivation and guidance throughout the process. We would also like to thank the faculty of the InformationTechnologyDepartment for their valuable assistance with our project. 9. REFERENCES [1] Rishabh Mittal and Anchal Garg, “Text extraction using OCR: A Systematic Review.” IEEE (2020) [2] S. Tomovic, K. Pavlovic, M. Bajceta, “Aligning document layouts extracted with different OCR engines with clustering approach.” Science Direct (2020) [3] Yash Kumar, Gaurav Sharma, Komal, Prof. Audumbar Umbare “E-KYC Mobile Application using Optical Character Recognition”. IRJET (2020) [4] Hejing Wu, Fang Liu, Long Zhao Yabin Shao, “Data Analysis andCrawlerApplicationImplementationBased on Python.” IEEE 2020 [5] S. D. Bandari, Ankita Jagtap, Namrata Mane, Swarali Garud “Intelligent Framework for Auto filling Web form using Scanned Documents” IJISRT [6] A Al Mamun, SR Hasan, MS Bhuiyan, M Shamim Kaisar, Mohammad Abu Yousuf SecureandTransparentKYCfor Banking System Using IPFS and Blockchain Technology. IEEE (2020)