Upload a photo of a single tile to test recognition
1
Upload
2
Verify
3
Score
Recognition Engine
Loading
●
OpenCV.js
Loading...
●
Recognition
Waiting...
●
Tile Database
0 tiles loaded
Algorithms
Fewer algorithms = faster. All 3 features = most accurate.
OpenCV Build & Cache
Forces recomputation of all tile features
Check each tile is correct. Tap to change any tile.
0
Identified
0
Tile Backs
0
Empty
0s
Time
๐
Orientation
๐
Detect Board
๐ง
Perspective
โ๏ธ
Extract Tiles
๐
Identify
Preparing analysis...
Check the tiles above are correct.
๐ง Debug Output
๐ท Board Detection Overlay:
๐ Board Detection:
Waiting for analysis...
๐ฏ Extracted Tiles:
Corner hexes marked with โญ: Col 1 Pos 1 (TopLeft), Col 1 Pos 3 (BottomLeft), Col 7 Pos 1 (TopRight), Col 7 Pos 3 (BottomRight)
๐ Recognition Details:
Breakdown of each tile's identification by method (ORB, AKAZE, BRISK features ยท OCR text ยท Color histogram)
Final Score
0
pts
Score Breakdown
Total
0 pts
Tiebreaker: most Animal tiles wins. If still tied, share victory.
Help improve the app by reporting incorrectly identified tiles