コミュニティ

Vastly different responses from very similar images

I am uploading 2 images:

source 1: http://diffusion.loto-quebec.com/Res_prod/Quebec49/2013_08/Quebec49_2013_08_17_det_AN.gif source 2: http://diffusion.loto-quebec.com/Res_prod/Quebec49/2013_08/Quebec49_2013_08_07_det_AN.gif

They are almost identical... but the structure of the response given is vastly different and we can't write a reusable script to get the values out of the response!

Why are the responses so different?

この記事は役に立ちましたか?

0人中0人がこの記事が役に立ったと言っています

コメント

1件のコメント

  • Avatar
    Permanently deleted user

    The issue occurs because ABBYY Cloud OCR SDK is not suitable well for reproducing the layout of such images, it is mostly suitable for the documents. We recommend either to use the field-level recognition (specify the exact coordinates of the text fragments you want to recognize), or export the result to an XML-file and process the recognized text with its coordinates as you need.

    0

サインインしてコメントを残してください。