🚀 go-pugleaf

In article <1993Jan24.020112.9370@netcom.com> pdh@netcom.com (Phil Howard ) writes:
>"danny hawrysio" <danny.hawrysio@canrem.com> writes:
>
>			A summary of JPEG
>				by
>			Philip D. Howard
>
>
>JPEG works with 24-bit color images.  It functions separately in each of the
>3 primary colors, that is, do the reds, then do the greens, then do the blues.
>The three colors are 8 bits each.  It also works with gray scale 8-bit.  The
>theory could also be applied in other ways.

For best results, JPEG work with the YUV colour model, where represents the
intensity of a colour, and U and V together describe the chrominance of
a colour.  The reason this is done is that the Y component has mostly
low spatial frequency (changes slowly) components and so can have its high
frequency components quatized by a larger amount.

>The image is chopped up into little squares of 8 pixels by 8 pixels in order
>to reduce the computation scale.  These image segments are then analyzed by
>a method called the Digital Cosine Transform, which is similar in certain
>ways to a Fourier Transform.  The result of this transform is that the various
>repetitious patterns are extracted with respect to their periodicy and their
>amplitude (the degree to which they contribute).

The transform used is called the Discrete Cosine Transform.

>These quantized numbers are then formed into variable length integers that
>occupy only a few bits.  The zeroes are treated as a special case since it
>is known that the quantizing results in a lot of them; they take the fewest
>bits, typically 1 bit and maybe fewer (I have not looked at this area of
>detail yet).  The more loss there is, the more zeroes and the smaller the
>data.  The non-zero numbers will also have less range and can be squeezed
>into few bits in many cases as well.

The coefficients from the DCT process are seperated into to types, the DC
(average colour of the 8x8 block) and AC (the remaining frequency components).
The DC coefficients are differentailly coded by themselves.  Coding them
differentially (storing the difference between two consecutive DC's) is
advantageous in that it produces many 0's since the average colour of
adjacent 8x8 blocks is usually the same.  

As pointed out the AC coefficients have many zero's in them and are coded
using run-lengths (store the number of consecutive 0's).

>The best way to compress a JPEG image is to start from the original 24-bit
>data.  Images with less noise work better, of course, since noise can appear
>to be high frequency spectral components.  Noise typically appears in the
>image as graininess.

Actaully I expect JPEG to improve the quality of noisy images since it chops
off high frequency components and noise is usually present in the high
frequencies.

Carl.