Geoffrey J. Woolfe

Color image processing using an image state architecture

Most digital images can be broadly classified into scene- referred and output-referred image states. This image state concept is the basis for an image processing architecture in which captured images are converted to a standard scene- referred color encoding where appropriate scene-state algorithms are applied. A rendering transform can then be used to convert images to a standard output-referred image state for further processing before the final printing/display of the image. This architecture is facilitated by the definition of standard image color encodings known as RIMM/ROMM RGB.