GrayscaleToRgb<T>
Models & Types · AiDotNet.Augmentation.Image
Converts a grayscale image to RGB by replicating the single channel across all three channels.
Inheritance: AugmentationBase<T, ImageTensor<T>> → ImageAugmenterBase<T> → GrayscaleToRgb<T>
For Beginners
Some models require 3-channel (RGB) input. If you have a grayscale image with only 1 channel, this copies that single channel three times to create a compatible 3-channel image. The image still looks gray, but now has the right number of channels.
How It Works
This creates a 3-channel image from a single-channel grayscale image by duplicating the luminance value into the R, G, and B channels. The resulting image will look identical to the grayscale original but is compatible with models that expect 3-channel input.
When to use:
- When feeding grayscale images to models trained on RGB data
- When using pretrained models that expect 3-channel input
- Mixing grayscale and color datasets in the same pipeline
Constructors
| Constructor | Summary |
|---|---|
GrayscaleToRgb(double probability) |
Creates a new grayscale to RGB conversion. |
Methods
| Method | Summary |
|---|---|
IDictionary<string, object> GetParameters() |