Core Image - Pixel Accuracy

Pixel Accuracy

All pixel processing provided by an Image Unit is performed in a pre-multiplied alpha (RGBA) color space, storing four color channels: red, green, blue, and transparency (alpha). Each color channel is represented by a 32-bit, floating point number. This provides exceptional color depth, far greater than can be perceived by the human eye, as each pixel is represented by a 128-bit vector (four 32-bit color channels). For color spaces of lower bit-depth, the floating-point calculation model employed by Core Image provides exceptional performance, which is useful when processing multiple images or video frames.