← Company News

Eliminating Character Confusion with Advanced Character Recognition Technology

Dynamsoft Label Recognizer SDK v3.2.0: Enhanced MRZ Code Scanning Experience with Spatial Text Line Grouping

VANCOUVER, BC - June 03, 2024

Dynamsoft updated its Label Recognizer software development kit designed to recognize meaningful zonal text or symbols in an image. The OCR SDK is ideal for scanning passport MRZ, ID cards, price tags in supermarkets, inventory labels in warehouses, VIN codes on car windshields, driver licenses, and more. The new version introduced two enhanced features to optimize the scanning experience.

In the past, when Label Recognizer was used for MRZ scanning, it primarily focused on single-row recognition, neglecting the spatial relationship between rows. However, given that MRZ-like scenes Dynamsoft Label Recognizer SDK v3.2.0: Enhanced MRZ Code Scanning Experience with Spatial Text Line Grouping typically encompass 2 to 3 lines of text positioned closely together, a paradigm shift was necessary. By treating this area as a cohesive TextLineGroup, we've streamlined MRZ configuration for enhanced clarity and efficiency.

Furthermore, our latest update introduces an advanced method for distinguishing commonly confused characters, such as discerning between the letter 'o' (as in 'olive') and the digit '0'. This advancement significantly improves accuracy and usability.

The SDK comes with built-in support for font and confused character sets. The currently supported built-in fonts are as follows:

  • ARIAL
  • COURIER_NEW
  • OCR_B
  • TIMES_NEW_ROMAN
  • VERDANA
  • HELVETICA

The currently supported built-in confusable characters sets are as follows:

  • ["0", "o", "O"]
  • ["l", "1", "I"]
  • ["5", "s", "S"]

In scenarios where standard fonts are not employed, our innovative solution ensures seamless character scanning. With our state-of-the-art supervised machine learning algorithm integrated into the SDK, users can supply sample images to teach the system to recognize unique fonts and configuration. This empowers the technology to better differentiate commonly misread characters, providing unparalleled accuracy and adaptability.

"Customers have long awaited Dynamsoft Label Recognizer, 3.2.0, we're unveiling our most significant investment yet in the SDK," commented Amy Gu, President of Dynamsoft. "This enhanced and versatile toolkit revolutionizes scanning capabilities, particularly in MRZ scanning, while delivering improved accuracy and efficiency across the board."

More About Dynamsoft Label Recognizer

Dynamsoft Label Recognizer SDK is based on optical character recognition (OCR) technology. OCR is widely used across industries to recognize text from scanned paper documents to enable conversion to digital formats. For this task, most OCR solutions are suitable in speed and accuracy. However, general OCR often falls short in specific applications requiring more precise and context-aware recognition.

Compared with general-purpose OCR SDKs, Dynamsoft Label Recognizer SDK is optimized for recognizing sparse and short text, including random numbers and characters for machine reading. It targets images such as pricing labels, ID cards, and tags. Its primary purpose is to extract structured and semi-structured data and return it as key-value pairs. The SDK supports specifying target zones, such as text neighbouring a barcode or on top of a red rectangle.

Pricing and Availability

Dynamsoft Label Recognizer works with content that includes text, numbers, alphanumeric, and common symbols. Version 3.2.0 is now available and can be purchased and downloaded from the Dynamsoft website. Pricing varies by licensing type. There are several editions to cover development for C/C++, .NET, Java, Android, and iOS.

Dynamsoft Corp. provides enterprise-class TWAIN™ software development kits (SDK), a Barcode Reader SDK, a Document Normalizer SDK and a Label Recognizer SDK to help developers meet document imaging, scanning, barcode reader applications requirements when developing web, desktop, or mobile document management and label reading applications. The imaging SDKs help today's businesses seeking to migrate from wasteful paper-based workflows to efficient paperless electronic workflows. Thousands of customers use Dynamsoft's solutions. Customers include 3M®; EMC®; FUJIFILM Medical Systems U.S.A., Inc.; Fujitsu®; GE®; H&R Block®; HP®; IBM®, Intel®; Infosys®; Lockheed Martin®; Olympus®; Philips®; PricewaterhouseCoopers®; Samsung®; Siemens®; Symantec®; Unisys®; Verizon®; and more. Dynamsoft is an ISO 27001-certified organization and an associate member of the TWAIN Working Group that develops TWAIN standards. The company was founded in 2003. More information is available at https://www.dynamsoft.com.

Note: Whether noted or not, references to certain words may be trademarks or registered trademarks of their respective owners.