LINGSHU OS · ASSESSMENT METHOD
AI tongue analysis
Image recognition extracts tongue features from captured photographs, with annotated data used to train feature models. The resulting structured observations can be included in a wider wellness report.
How it works
Multiple forms of tongue information are considered: shape, texture, colour, sublingual veins and visible material. It uses entity relationships, graph databases and semantic links to organise heterogeneous data, followed by neural-network feature learning.
Acquisition and analysis workflow
- 01
Capture the tongue image using the product procedure
- 02
Check image quality before analysis
- 03
Extract visual features with the trained model
- 04
Combine the observations with other assessment information