Image Processing Reference
In-Depth Information
Composite sample rates
CCIR Rec 601
4:2:2 sample rate
768 x 483
720 x 480
Composite sample rates
CCIR Rec 601
4:2:2 sample rate
948 x 576
720 x 576
Square pixels
Non-square pixels
Figure 5-28 Working areas.
lower resolution than the luma, so the red color seems to spill out of the area of the
player's shirt.
Luma and Chroma Component Signals
Reducing the video picture to component parts helps to process the video for compres-
sion. It certainly helps the transmission process at the expense of some of the information.
Deciding which way to componentize the picture is very challenging. The output of
the video circuits will ultimately drive a red, green, and blue additive-light emission sys-
tem in the display. Displays requiring reflective or transmissive illumination use subtrac-
tive color mixing that requires a conversion to cyan, yellow, and magenta (CMY) color
The TV industry decided a long time ago that the most versatile technique was to
describe images with one lightness signal and a pair of color-difference signals that work
together. The word luminance was used to describe the brightness of a video signal rela-
tive to a peak white value.
The Commission Internationale de l'Éclairage (CIE) defined the term luminance and
chose the symbol Y to represent it in mathematical equations that describe color values.
TV engineers and color scientists use the word luminance to mean quite different things.
Be mindful of your usage of the word luminance as it has been widely misused and
its meaning has been somewhat ambiguous. Also be wary of the notation if you see it in a
mathematical formula. The relative (to peak white) luminance value is indicated in math-
ematical formulas by the value Y but is rarely used in digital imaging systems. Luma,
which is the most interesting quantity, is a weighted value derived from the gamma-
Search WWH ::

Custom Search