Height Detection Dataset

v4

dataset5

Generated on Jul 4, 2024

Dataset Split

Train Set 99%
7464Images
Valid Set 1%
68Images
Test Set 0%
32Images

Preprocessing

Auto-Orient: Applied
Static Crop: 4-100% Horizontal Region, 16-79% Vertical Region
Resize: Stretch to 640x640
Tile: 2 rows x 2 columns

Augmentations

Outputs per training example: 3
Flip: Horizontal
90° Rotate: Clockwise, Counter-Clockwise, Upside Down
Rotation: Between -12° and +12°
Saturation: Between -25% and +25%
Brightness: Between -22% and +22%
Noise: Up to 0.18% of pixels