OCR Jobs

Needed right away, with long-term possibility as we have constant and non-stop projects for the next 12 months. We are a team working with the best freelance developers creating and maintaining long-term profitable relationships (not necessary if not interested). We are reliable, trustworthy, and friendly recruiters; we've been hiring freelancers since the conception of the start-up. What we're looking for: a. Top expert Python experience mandatory (Python Request and/or Selenium), other languages welcome but not necessary b. Strong web scraping and pdf/images data extraction (OCR) c. Expert on bypassing website security protection: fakeuser, proxies, captchas, etc. d. Excellent pay of minimum $500 per scraper; up to $1,000 depending on complexity e. Maximum confidentiality, no ...

Hey! We are building a search engine for international law. You can find a description at We are converting PDFs in varying qualities, to structured HTML in very high quality. We need help as we have a lot of PDFs to convert. We are using FineReader 15 for this operation. You need to have it. The process is: 1) Open the PDF file with FineReader OCR Editor 2) Replace the zone of recognition by an adjusted one 3) Control the zones in few pages and identify tables 4) Launch the OCR 5) Verify the recognition and correct the text according to the PDF (there is a very efficient tool in FineReader to do that). We will give you few documents and see how you work. We intend to build a long-term relationship with one or several serious freelancer(s) as we have a lot of documents to convert. Pl...

What I am trying to achieve is a denoised image ready for use in Android Tesseract OCR. The solution I came up with is to use the AbsDiff implementation of opencv to make the text colour within the images uniform. This is achieved with assistance of MedianBlur. After implementing this solution, i now experience noise in the images that can be as a result of the AbsDiff. What i now need is help in denoising the resulting image from AbsDiff or a custom implementation to achieve the solution I have described above. AbsDiff - Absolute Difference MedianBlur - Blur effect by the use of the median value within a range of pixels in an image.

Create a web application for the Correction and Supervision for OCR results 1) Admin panel to create projects, configure project requirements, upload data [Text Image/PDF/Tiff and corresponding OCR result in json], add team, distribute tasks to team members, User management as per roles 2) Team access: Start project, work on the task, and submit the output 3) Project Manager access: Team task progress, project completion status, download completed work 4) Deployed on AWS Project Configuration: Take Project details, names of key to be displayed in tasks for manual correction, confidence score threshold under which the images+json fields to be displayed for manual correction Task: In a single pane, view the receipt/invoice image and OCR result as key - value as a form. Make corrections man...