I’m trying to make a mobile app for extracting text from an image taken with a camera, and saving it in a pdf. I’m having a problem with extracting logos, tables and signatures from the image (expect a basic document you want to save as a pdf, similar to Adobe Scan), and keeping the layout of the document (paragraphs and such). My api is in asp.net core, and for the mobile app I’m using flutter. In the flutter app I’m taking the pictures with my phone and sending the bytes of the image to the api, where the whole logic with extracting the text from those images is.

For now I’m using tesseract for extracting the text, and saving it as a pdf file (but the image has to be really clear for the whole text to be extracted properly). Is there another library I should use in addition to tesseract so that I can keep the layout? I don’t want to save it as an image in the pdf because I want the text to be selectable.