CLAHE<T>
Models & Types · AiDotNet.Augmentation.Image
Applies Contrast Limited Adaptive Histogram Equalization (CLAHE).
Inheritance: AugmentationBase<T, ImageTensor<T>> → ImageAugmenterBase<T> → CLAHE<T>
For Beginners
Regular histogram equalization can make noise very visible. CLAHE works on small regions and limits how much contrast it adds, so you get better detail without amplifying noise. It’s widely used in medical imaging.
How It Works
CLAHE improves upon standard histogram equalization by dividing the image into tiles and equalizing each tile independently, with a clip limit to prevent noise amplification. The tile boundaries are blended using bilinear interpolation for smooth results. Based on Zuiderveld (1994).
Constructors
| Constructor | Summary |
|---|---|
CLAHE(double clipLimit, int tileGridSize, double probability) |
Creates a new CLAHE augmentation. |
Properties
| Property | Summary |
|---|---|
double ClipLimit { get; } |
Gets the clip limit for contrast limiting. |
int TileGridSize { get; } |
Gets the number of tiles in each dimension. |
Methods
| Method | Summary |
|---|---|
IDictionary<string, object> GetParameters() |