Difference between online OCR and OCR API?


#1

Hi, I used your OCR API and extracted text from an image. The response received read the alphabet “I” as “l” in small case. The response is given below:

•COME TAX \r\n ADHISH VERMA \r\n VIRENDRA KUMAR VERMA \r\n 21/01/1986 \r\n Permanent Account Number \r\n ' AlQPV7867A \r\n Signature \r\n GOVT. OF INDIA \r\n

However on uploading the same image to your online OCR form we got the correct response as given below:

***** Result for Image/Page 1 ****** INCOME TAX DEPARTMENT ADHISH VERMA VIRENDRA KUMAR VERMA 21/01/1986 Permanent Account Number AlQPV7867A Signature GOVT. OF INDIA

Why is there a difference?


#2

In the online OCR demo the parameter scale is set to true, but in the OCR API this parameter is set to false by default.

If scale=true, the api does some internal upscaling. This can improve the OCR result significantly, especially for low-resolution PDF scans or smaller images.

See also https://ocr.space/ocrapi#scale


OCR API and C#, extract text from images