Back to all work
Case study · Banking
Computer vision for cheque clearing
Detection, OCR/ICR and signature verification for cheque clearing, trained on 20,000 annotated cheques.
Headline result
~70%
less labelling time
- Industry
- Banking
- Client
- TBD
- Result
- ~70%
The problem
Cheque clearing depended on reading handwritten and printed fields, and verifying signatures, reliably and at volume.
What we built
A computer vision system trained on 20,000 annotated cheques, using Faster R-CNN, Mask R-CNN and YOLOv3 with custom OCR/ICR and signature verification.
An auto-annotation workflow cut labelling time by around 70%.
Outcome
Automated field extraction and signature verification for cheque clearing.
Next case study
Get in touch
Have a similar problem to solve?
Talk directly to a senior engineer about what you're building.