hy there!
I am using gpt 4o in form of a costum-gpt to read text on old postcards and give the infos in form of a csv file. this works fine till i reach the file limit.
If i put some jpg files in a Zip-archive and upload this Zip-archive, there is an issue with the OCR and it does not work.
The failure is stated in German, as i work in German as : “Es scheint, dass die deutsche Sprachdatei für Tesseract fehlt oder nicht korrekt geladen werden kann”
How can i get gpt working not only to analyse 20 files and then stopping because the file upload limit. I need a few hundreds being processed.
thx
Unless I’m mistaken if you are zipping images then Code Interpreter is used to unzip and process the images, not GPT-Vision.
Tesseract is a common, open source OCR technology.
Besides convenience there’s no reason why’d you’d use a GPT like this. You can ask GPT-4 to convert this task into a Python script and run it locally on your computer.
Thx for your reply.
Yea, one thing is the OCR Recognition of the Text on the old photos/postcards. But the other is the compiling the information in an article text for my archive.
For example: The village seen in a postcard should then be written in a csv file and the district should be added. All works fine, when i up them as jpg .