textDetection() method for 'complex' text in an image

[Димитър Клатуров]

I have this image (see the attachment) which in this case there a horizontal line that is separating the text in two columns, but Google Vision API its reading the file from left to right and ignoring the line and it is useless when there is such lines in the file and all my images are scanned old papers and books that i want to convert it to text using Google API, but i cant. Is there a way i can do this without using external library for cropping the image or do you plan to add this functionality in the future?

[Monica (Google Cloud Platform)]

Hello,

Are you using DOCUMENT_TEXT_DETECTION? Using blocks should address this issue. Please find here some tips & tricks for using Google Vision API for text detection.