Page 209 - DCAP303_MULTIMEDIA_SYSTEMS

Unit 12: Image Format

The YCbCr colour format is ideal for JPEG compression as it enables brightness and colour notes
information from an image to be separated. This is desirable as, described previously; the human
eye tends to notice brightness variation more than colour variation. By separating these two
properties we can apply different tolerances and compression ratios to each so as to get optimal
compression at the best image quality possible.
The RGB colour scheme is simply a method of storing the Red, Green and Blue colour components
of a pixel. It uses a byte of data per colour component for full colour (or 24 bit) values. In other
words, each colour component has a range of 0 to 65535 for the value of the colour being described.
Note that this is not the same as 16 bit colour which allows only a subset of the 24 bit colour
field.
Taking the RGB values and inputting them into a set of equations as follows derives the YCbCr
colour format coefficients,
Y = 0.3 × R + 0.6 × G + 0.1 × B – 128
Cb = –0.15 × R – 0.3 × G + 0.45 × B

Cr = 0.4375 × R –0.375 × G – 0.0625 × B
Where R, G, B and Y are unsigned 8 bit numbers and Cb and Cr are signed 8 bit numbers.
This gives 8 bit values for the Brightness and the two Chrominance coefficients. This now gives
more meaningful variables for the JPEG algorithm to work with.
The YCbCr or YCbCr, sometimes written YC C or Y′C C , is a family of
B
R
R
B
colour spaces used as a part of the colour image pipeline in video and digital
photography systems. Y′ is the luma component and C and C are the blue-
B
R
difference and red-difference chrome components.
12.4.4 JpeG file format
The File format that was used for this project was JFIF (JPEG File Interchange Format). This is not
strictly a part of the JPEG standard, but was developed to attempt to standardize the JPEG formats
so that greater compatibility between platforms and applications could be obtained. Currently
JFIF is the main file format that is used, especially for JPEG Baseline compressed images.

The JPEG standard defines a number of header fields and sections to allow the JPEG encoded
image to be decoded. It also left a number of sections open so that applications could make use
of them as they saw fit. The result was that a large amount of incompatible versions of JPEG were
released, as the header format of the file was not compatible. The JFIF solved this problem by
standardizing the application specific sections so that compatibility could be introduced.
There are quite a number of header sections in the JPEG that defined, but are not applicable in
JPEG baseline. For simplicity, this section will only document the sections that this project actually
used, but the reader should be aware that there a quite a few more.

table 12.4: Complete List of available segments in JpeG

Description symbol Marker
Baseline DCT, Huffman SOF0 FFC0H
Extended sequential DCT, Huffman SOF1 FFC1H
Progressive DCT, Huffman SOF2 FFC2H
Spatial (sequential) lossless, Huffman SOF3 FFC3H
Differential Sequential DCT, Huffman SOF5 FFC5H
Differential progressive DCT, Huffman SOF6 FFC6H
Contd...

LoveLy professionaL University 203