Case Study: Best Practices Improves ICR Accuracy with Parascript

Best Practices - Customer Case Study

Best Practices worked with Parascript to improve the accuracy of intelligent character recognition (ICR) for handwritten data. Their challenge was that OCR and basic text-recognition methods are often accurate at the page or character level, but handwriting fields such as names, dates, and business information require much richer context to read reliably.

Parascript addressed this by using its FormXtra software and applying context through field types, properties, character style, vocabularies, and value constraints to each document field. This approach helped narrow possible interpretations and improve recognition of handwritten fields, with the case study citing typical accuracy ranges of 75%–90% for numeric fields, 70%–85% for dates, 40%–75% for alpha fields, and 30%–70% for alphanumeric fields depending on context and image quality.

Best Practices