AlphaTrack AI Synthetic Demo
Synthetic demonstration. This detects track-like features in
labelled synthetic CR-39 imagery. It is not real-world CR-39 accuracy and
makes no particle species, energy or yield claim. Detections and overlays below are precomputed with the real
alphatrack_core kernel; the viewer lets you explore them.1 · Choose an example
Loading…
Viewer
2 · Image metadata
3 · Run analysis
- Image validation
- Image-quality checks
- Background correction
- Candidate detection
- Morphology extraction
- Duplicate reconciliation
- Confidence calibration
- Result generation
4 · Results
Safe output (per detection)
5 · Human review
Every edit creates a new annotation version rather than modifying the original result. Compare the algorithm count with your reviewed count.
Benchmark card
See the full AlphaTrack model card and limitations. The OpenCV baseline can't run in the browser Python runtime, so results here are precomputed with the real kernel and shipped as static data.