Crop and deskew module for MemoraCar's OCR

In collaboration with MemoraCar @ https://www.memoracar.com/

cURL request:

Just for the demo, we use the cURL command on linux.

curl -X PUT -H "Authorization: <token>" -F "image=<path_to_image>" <adress> --output <image_output>

Example:

curl -X PUT -H "Authorization: 8bf54562-3000-11eb-adc1-0242ac120002" -F "image=@/home/axel/Pictures/DS_MemoraCar/greyCard-1.jpg" http://localhost:5000/scanDeskew --output scanned_image.jpg

Réponses de l'API

• Code 200

  • Image treated successfully, returning the treated image.

• Code 501

  • Couldn't treat the image, returning the untreated image.

TODO:

• potentially make the code more readable than it is (it is readable but can be improved)
  • Using classes?
  • Reformatting the project structure?