What the OCR result from Microsoft looks like. How can I extract some specific text?
The result of OCR from Microsoft is an array of ‘key-value’ objects. To get more information about “Recognize text (Microsoft)” function you can go to our knowledge base.
To select specific values from scanned documents:
- Use exact coordinates or enter an inaccuracy, if we know that the coordinates of the element can change within certain limits, let’s assume that the necessary text is exactly on the coordinates from X to X+50 and from Y to Y+50. Then we can choose all elements from a file at which parameters correspond to the given requirement.
- Use keywords. For example, we know that the necessary value in the table can go after the words of the name of the last title and before the “Total amount” and so on…
Is that clear from my answer? Maybe there are still questions? Please, feel free to ask them.