Reading Vehicle Licence Plate (Vehicle Number Plate) Using Document Information Extraction

[Singh_Rishiraj]

I am working on a project where I am using Document Information Extraction technology to extract text from vehicle number plates. This technology helps recognize and pull out the alphanumeric characters (text) displayed on number plates. However, I am facing challenges when it comes to extracting the color of the number plate. While the text extraction is not working properly and the color detection part isn't being detected, and the results are not accurate. I need help to improve this process so that I can reliably capture both the text (numbers and letters) and the color of the number plate from images.

I am looking for advice, suggestions, or best practices that could help resolve this issue and make the system work more efficiently.

Specifically, I want to ensure that both the text and color extraction happen smoothly and accurately in my project.
@tomasz_janasz @MariaGoepfert 

Document Information Extraction 

[tomasz_janasz]

Hi, when it comes to the OCR extraction the best practice is as follows: "The service extracts dark text on a light background more accurately than light text on a dark background." see: https://help.sap.com/docs/document-information-extraction/document-information-extraction/optical-ch.... When it comes to the color extraction, I believe the service cannot recognize the paint today.

[Singh_Rishiraj]

Does not work properly