@kensanata Tesseract has really good accuracy and is open. I have some slides on how to use it: https://nyu-dataservices.github.io/ExtractingOCR/#/13