0 Replies Latest reply on Jun 5, 2014 7:05 AM by ucfadll

    Automatically renaming file based on OCR field

    ucfadll

      Each week we received hundreds of partly pre-populated and partly hand-written survey forms each week (hard copies). We then scan the forms and rename them according to some of the pre-populated parts of the forms.

       

      To help speed this up, I am looking for a way of automatically renaming the forms. This would involve OCRing the forms, and then using a few of the boxes (always in the same place of the form) to create a filename.

       

      If it helps, the forms are about the activities of vehicles. The sections we use to name the resulting PDF is based on:

      Registration mark (which appears in same place all the time)

      Year of the survey (which can be got from a field containing the survey date)

      Whether the form was sent to a British firm, or a firm based in Northern Ireland (which can be got from the title at the top of the page)

       

      If more context helps, the process we currently follows is:

      1. 1) Scan 10 double-side sheets at once
      2. 2) Rotate each even page (this is because the back of the form is in landscape, so we rotate it to make the words the same way around in the PDF)
      3. 3) Match each pair of sheets together and save them in their pairs to a new PDF. So pages 1 and 2 get put into one PDF, then 3 and 4, then 5 and 6, etc
      4. 4) Build a list of paired sheet filenames in Excel, open each file and put the registration mark, year and country in Excel
      5. 5) Create a batch file to rename and move the PDFs in Excel
      6. 6) Run batch file to rename and move files

       

      Stage 2 is carried out using the standard rotation tools in Acrobat. We have a Javascript action for stage 3. Stage 4 is all done manually – by opening each PDF and eyeballing the document, and then typing the relevant information into Excel. It is this stage that takes the longest (and gives opportunities for typos) so I would like to automate it.

       

      Any thoughts, hints or tips gratefully received.

       

      Thanks