I used processImage method with textExtraction profile for extracting number from an Image. But this is not my required output.

Here is sample image(cropped) to be read by sdk. Image link --->InputImage

and here is the output

here is another image to be read sampleImage2.

and this is output.

Actually I need to capture an image of Energy meter and then only the meter reading will be recognized by ABBYY OCR-SDK.(This is my primary requirement).

this is the full image of the EnergyMeter.

If you have already deal with my type of requirement please guide me to the direct way that I will achieve my output easily. We have strict deadline.



1 comment

  • Avatar
    Anna Falevskaya


    We’re extremely sorry for such delay in response.

    Usually for detecting small amounts of text the processFields and processTextField methods are recommended, but unfortunately, there is no stage of analysis of the document when using these methods, so the Cloud OCR SDK cannot determine that the text is inverted and does not recognize it. The inverted text can be recognized only by the processImage method. So in your case (inverted text on the InputImage) you can use the processImage method or recognize the inverted images with the processFields or the processTextField methods.

    Comment actions Permalink

Please sign in to leave a comment.