about the project
An international retail company approached us to develop a system for automated recognition and verification of DataMatrix codes from photographs.
project goals
To create an automated system that:
– recognizes and verifies DataMatrix codes from photos;
– reduces the amount of manual work;
– increases transparency in the processing workflow;
– ensures reliable storage of data and source images.
solution
The team developed a computer vision service for DataMatrix code recognition and integrated it with the Sherpa RPA robot.
We also implemented a pipeline for asynchronous processing and storage of all data, including photographs.
In addition, we set up automatic generation of CSV reports and their delivery by email.
technology stack
result
We achieved recognition accuracy of over 95% on the client’s test dataset.
