# How to: Use a Color Matrix to Transform a Single Color

**.NET Framework (current version)**

GDI+ provides the Image and Bitmap classes for storing and manipulating images. Image and Bitmap objects store the color of each pixel as a 32-bit number: 8 bits each for red, green, blue, and alpha. Each of the four components is a number from 0 through 255, with 0 representing no intensity and 255 representing full intensity. The alpha component specifies the transparency of the color: 0 is fully transparent, and 255 is fully opaque.

A color vector is a 4-tuple of the form (red, green, blue, alpha). For example, the color vector (0, 255, 0, 255) represents an opaque color that has no red or blue, but has green at full intensity.

Another convention for representing colors uses the number 1 for full intensity. Using that convention, the color described in the preceding paragraph would be represented by the vector (0, 1, 0, 1). GDI+ uses the convention of 1 as full intensity when it performs color transformations.

You can apply linear transformations (rotation, scaling, and the like) to color vectors by multiplying the color vectors by a 4×4 matrix. However, you cannot use a 4×4 matrix to perform a translation (nonlinear). If you add a dummy fifth coordinate (for example, the number 1) to each of the color vectors, you can use a 5×5 matrix to apply any combination of linear transformations and translations. A transformation consisting of a linear transformation followed by a translation is called an affine transformation.

For example, suppose you want to start with the color (0.2, 0.0, 0.4, 1.0) and apply the following transformations:

Double the red component

Add 0.2 to the red, green, and blue components

The following matrix multiplication will perform the pair of transformations in the order listed.

The elements of a color matrix are indexed (zero-based) by row and then column. For example, the entry in the fifth row and third column of matrix M is denoted by M[4][2].

The 5×5 identity matrix (shown in the following illustration) has 1s on the diagonal and 0s everywhere else. If you multiply a color vector by the identity matrix, the color vector does not change. A convenient way to form the matrix of a color transformation is to start with the identity matrix and make a small change that produces the desired transformation.

For a more detailed discussion of matrices and transformations, see Coordinate Systems and Transformations.

## Example

The following example takes an image that is all one color (0.2, 0.0, 0.4, 1.0) and applies the transformation described in the preceding paragraphs.

The following illustration shows the original image on the left and the transformed image on the right.

The code in the following example uses the following steps to perform the recoloring:

Initialize a ColorMatrix object.

Create an ImageAttributes object and pass the ColorMatrix object to the SetColorMatrix method of the ImageAttributes object.

Pass the ImageAttributes object to the DrawImage method of a Graphics object.

Image image = new Bitmap("InputColor.bmp"); ImageAttributes imageAttributes = new ImageAttributes(); int width = image.Width; int height = image.Height; float[][] colorMatrixElements = { new float[] {2, 0, 0, 0, 0}, // red scaling factor of 2 new float[] {0, 1, 0, 0, 0}, // green scaling factor of 1 new float[] {0, 0, 1, 0, 0}, // blue scaling factor of 1 new float[] {0, 0, 0, 1, 0}, // alpha scaling factor of 1 new float[] {.2f, .2f, .2f, 0, 1}}; // three translations of 0.2 ColorMatrix colorMatrix = new ColorMatrix(colorMatrixElements); imageAttributes.SetColorMatrix( colorMatrix, ColorMatrixFlag.Default, ColorAdjustType.Bitmap); e.Graphics.DrawImage(image, 10, 10); e.Graphics.DrawImage( image, new Rectangle(120, 10, width, height), // destination rectangle 0, 0, // upper-left corner of source rectangle width, // width of source rectangle height, // height of source rectangle GraphicsUnit.Pixel, imageAttributes);

## Compiling the Code

The preceding example is designed for use with Windows Forms, and it requires PaintEventArgs* e*, which is a parameter of the Paint event handler.