AWS Recognition can detect upto 50 words in a single image only .
You should use an OCR software for large amount of data , such as free OCR
Now AWS textract service overcomes the limitations of rekognition, and provides OCR.
Trying to use the AWS Go SDK to do DetectText, basically trying it for OCR, and am seeing only about 30% of text being included in the response, and the rest of the text being ignored. Anyone aware of a limit on total number of text blobs being returned in the response? Any way to get around that?