ID Document OCR

Machine Eye ID Document Scanner

#1 OCR Scanner
Best-Adapted to Read Korean ID Documents.

Korean ID documents present unique challenges. What works well for scanning alphabet-based ID Documents often falls short when handling the Korean script and complex security features in Korean ID documents, increasing the difficulty of document recognition. But no worries - we’ve got it all covered. Simply integrate our solution—available as Web, Mobile, Server SDKs, or SaaS—into your application or platform to meet your needs seamlessly.

Talk to Sales

#1 OCR Scanner Chosen by 10 out of 15 Tier 1 Banks in Korea

Machine Eye ID Scanner has been adopted by over 50 financial institutions in Korea to enhance and accelerate their eKYC processes.

➊ Tested and Approved by the Majority of Tier 1 Banks in Korea

➋ Widest Range of Integration Options Available to support business of all sizes

➌ ID Document Liveness Detection available to Combat Identity Fraud

Machine Eye-trained to process Korean Characters with over 98% accuracy in under 1 sec

Top Solution to Improve First Time Pass Rate for Korean Users

A scanner optimized for the alphabet won’t suffice for Korean characters, which are far more complex, with over 48 times the combinations of the 52 letters in the English alphabet.

With 11,172 possible combinations—about 2,500 commonly used in modern Korean—and additional variations from foreign words, the complexity increases significantly.

Plus, Alphabet, which consists of 52 visually distinct letters (uppercase: 26, lowercase: 26) is relatively easier for OCR to identify even if impacted by shade, glare or other noise. In contrast, for Hangul(the Korean Character), shade, glare, or any other noise that goes through a stroke can be a challenging factor in OCR because that single stroke may result in extracting the wrong character, impacting overall accuracy.

[How a “single stroke” can change the OCR Result]

Unlike the alphabet letters shown below, 김, 감, and 검 look very similar but are distinct letters, making them far more complex for OCR to process.

As illustrated, training OCR to precisely extract Korean Characters demands advanced expertise, robust data, and precise technology. Rest assured- we’ve done the hard work for you. Instead of taking the long, data-intensive route, use our ready-to-go model to save time and resources.

Machine Eye ID Document OCR Performance and Adoption

Optimized to Handle OCR Challenges in Real-World Settings

Machine Eye is fully trained and optimized to handle the following challenges, ensuring precise and high-speed recognition.

Challenge 1

❶ Multiple Skew Detection Algorithm

Users often submit tilted or skewed ID card photos during eKYC process, but we resolve this by employing multiple advanced skew detection algorithms. By identifying document edges and vertices, we deskew the image to realign the text, ensuring accurate extraction.

Challenge 2

❷ Irregular Text Spacing

Inconsistent letter spacing complicates OCR, which relies on word boundaries. Our model, trained on diverse data, uses adaptive heuristics to split or merge text as needed, ensuring precise character-level processing.

Challenge 3

❸ Varying Font Styles for Text and Number

Varying fonts in a single image increase recognition complexity, but we handle this with separate models for text and numbers. Our approach improves handling font variations, reduces parameters, and enhances speed and efficiency.

Challenge 4

❹ Embossed Text and Hologram

Some documents incorporate built-in security features like embossed text or holograms, which can create glares and reduce contrast, complicating OCR. Our model is trained to handle such distortions, accurately extracting text as long as it’s visible to the human eye—and even in cases where text is barely discernible.

Supports all Types of IDs issued by the Korean Government

Machine Eye ID Scanner supports all types of ID Documents issued by the Korean Government, including National ID Card, Driver’s License, Passport, and Residence Card.

Tested and Proven by Tier 1 Financial Institutions

Discover areas where businesses are leveraging
Machine Eye ID Scanner to cut costs and improve efficiency.

1. Mobile App
Financial Institutions

In-app eKYC for Financial Institutions

2. Kiosk
Self-Service Hotels

ID verification via kiosks in Self-Service Hotels

3. Work Tablet
Field Insurance Agents

ID verification via work tablet for Insurance Agents

Developer-Friendly SDKs and SaaS Solution at Your Fingertips

Select the integration type that best fits your requirements. We provide comprehensive SDKs for mobile, server, and web platforms, ensuring seamless integration and flexibility for your development needs.

Easy to Integrate, Add New Options, and Scale

There’s no need for a complex system overhaul. You can effortlessly integrate our pre-built solution, already perfected to meet over 50 financial institutions demanding standards, or customize it to suit your preferences.

Need additional tweaks? Just let us know—we’re here to help with any customization you require.

➊ Simply Integrate ID Document Scanner to your application via our SDK

➋ Request customized features tailored precisely to meet your unique needs.

➌ Enable and Exploreother Scanning Options,including Barcode Scanner

MachineEye Scans it all

Looking for more than just scanning ID Documents? Check out our ID Liveness Detector to fight identity fraud. Or explore beyond ID recognition, as capturing barcodes, QR codes, and credit cards are available. Choose our bundled package or customize a single product to meet your exact needs.

➊ ID Liveness Check Scanner

➋ Barcode/QR Code Scanner

➌ Credit/Debit Card Scanner

Ready to Transform the Way You Scan?