Optical characterization of an object
A dual optical detection system with diffused and collimated light sources and a multi-view CNN algorithm effectively addresses the challenge of detecting inclusions and cracks in gemstones, enhancing classification accuracy.
Patent Information
- Authority / Receiving Office
- GB · GB
- Patent Type
- Applications
- Current Assignee / Owner
- DE BEERS UK LTD
- Filing Date
- 2022-12-20
- Publication Date
- 2026-04-22
AI Technical Summary
Existing methods for classifying gemstones, such as diamonds, struggle to accurately detect inclusions and cracks, which affect their quality, due to interference from reflections and diffractions at outer surfaces.
A dual optical detection system using diffused and collimated light sources, combined with circular polarizers, to distinguish between reflections from outer surfaces and internal inclusions or cracks, and a multi-view CNN algorithm for enhanced classification.
Improves the detection of inclusions and cracks in gemstones by enhancing signal-to-noise ratio and enabling precise classification based on optical imaging and machine learning.
Smart Images

Figure 00000000_0001_ABST 
Figure 00000000_0000_ABST
Abstract
Citation Information
Patent Citations
Process and system for diamond clarity measurement
US20210033540A1
Gemstone sorting apparatus and methods
US5193685A
Inclusion detection in polished gemstones
WO2011054822A1