Detection of vertical text, boxed text and info about OCR enigne

OCR.space Free OCR API
1

Hi,
I tried a PNG and a PDF image and I have the following questions. I would deeply appreciate if you can respond to them, at least a few.

  1. My PNG is 6530 X 3528 but it doesn’t allow showing parse error and saying that max size is 10K X 10K
  2. PDF worked and when I overlaid the image I can see that it doesn’t detect the vertical text. Can something be done about this?
  3. It certainly doesn’t detect any text bounded inside a circle or a square. Only the text which is free is detected. Can something be done about it?
  4. Can you reveal which OCR engine is used for this tool? Is it tesseract ?
  5. Since not all the text is detected can you suggest whether this tool can be configured OR some image processing should be done before feeding the image to OCR tool?
    Thanks a lot.
2

  1. Please attach the image so that we can test. The size should work.

  2. Vertical text is not detected. If everything is vertical, then it works again (for images), because we have autorotation

  3. It depends. An example image would be useful.

  4. We combine various OCR engines and methods, and also use some parts of Tesseract

  5. Again, it depends. Once I see sample images, I can give some more concrete advise.

3
  1. I have attached the image. I tried again and I got the following error message.
    Error: Parsing Error: Image dimensions are too large! Max image dimesion supported: 10000 x 10000.
    PID_test_image_2
    PID_test_image_2.png6530×3528 686 KB
  2. Less than 5% text is vertical and I think autorotation is conditional and not forcible.
  3. You can see text bounded inside a circle, bounded inside a circle which is inside a square. These are symbols and I need to recognize the text with their positions even if I have to remove the symbols through some pre-processing.
  4. I am using tesseract and getting almost everything but not the vertical and not the bounded text. When the online app recognizes text even then it combines various OCR engines? Can you tell which are popular ones apart from tesseract please?
  5. Can your tool be configured for my research project?
    Many thanks for your swift responses.
4

I hope you got my responses as in my earlier reply. Did you get time to look at them? Can I reliably rely on your OCR product to achieve fully or partially my objectives? I would greatly appreciate if you can spare a little time to have a quick glance on my questions so that I can make a decision. Thanks a lot.