Specifications

COPYRIGHT © 1999 CANON INC. CANON SUPER G3 FAX BOARD-F1 REV.0 MAY 1999 PRINTED IN JAPAN (IMPRIME AU JAPON)
2-45
CHAPTER 2 OPERATION AND TIMING
Example:
The following explains how the learning table works; in the diagrams, the pixels 1 through 10 of
the model template are all assumed to be white:
1. The pixel pattern of the model template will be 000h.
2. The level of prediction against the pixel in question “?” of the pixel pattern 000h is “white”;
however, it is actually black, making the prediction “wrong.”
3. Since the status ST is “0”, a check against the rate of probability table causes the status to
move to “1,” and at the same time the next prediction will be made “black” by reversing the
present prediction.
Figure 2-405 Learning Table Operation (example 1)
Original image
? is actually black.
Pixel
pattern
Level of
prediction
Actual
level
Correct/
wrong
Status
ST
000h
001h
002h
ST
0
1
2
3
123
45678
910?
LSZ
5A1Dh
2586h
1114h
080Bh
NLPS
1
14
16
18
NMPS
1
2
3
4
SWITCH
1
0
0
0
0
0
0
0 (white)
0 (white)
0 (white)
1 (Black) Wrong
Rate of probability table
Learning table of present pixel
Pixel
pattern
Level of
prediction
Actual
level
Correct/
wrong
Status
ST
000h
001h
002h
1
0
0
1 (black)
0 (white)
0 (white)
Learning table of present pixel