My client has an issue with screenshots which are included in their handbooks. For legal reasons, they want to blur out text areas in the images.

The screenshots are of a machine operator interface. The interface has many icons, each with a localised label. The screenshots are normally made in the manufacturer's language and the handbooks are translated in up to 28 languages. For this reason, the screenshots have kept the original text and are described in the target language.

The company now want to blur out the text areas. In order to help speed up the process, I am wondering how I can partially automate the process. My current idea is to load an image, detect the text areas, highlight these areas and allow the user to modify the areas and then finally to apply a blur in the text areas.

Can anyone recommend a library to support the detection of text areas and return a list of rectangles? Java or C# would be ideal. (Note that I don't need OCR)

  • what's problem with OCR ?
    – Arsenic
    Jun 7, 2021 at 12:00
  • @Arsenic as I mentioned, OCR is not required. I need to know where on the image the text is so that I can blur it. I have found many solutions which extract text but none which simply tell me where the text is.
    – paul
    Jun 7, 2021 at 12:17
  • you mean something like this ? fahmisalman.medium.com/…
    – Arsenic
    Jun 7, 2021 at 16:39


Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Browse other questions tagged or ask your own question.