| Title: | Mathematics at DEC |
| Moderator: | RUSURE::EDP |
| Created: | Mon Feb 03 1986 |
| Last Modified: | Fri Jun 06 1997 |
| Last Successful Update: | Fri Jun 06 1997 |
| Number of topics: | 2083 |
| Total number of notes: | 14613 |
I'm looking for anybody out there who has some experience in
shaperecognition (correct word??) via Fourier transformation.
There are to areas of interest: Finding a segment in a complex
signal (e.g. a rising edge with a rise time shorter than xyx)
or more komplex forms e.g. a sine wave in a segment of the
original curve etc. Another area of interest is in symbol
recognition in graphic displays as it is used to teach a computer
how to read letters. Maybe this is the wrong conference for this
question. Does somebody know anything about this subject......
Regards
Patrik
| T.R | Title | User | Personal Name | Date | Lines |
|---|---|---|---|---|---|
| 1208.1 | PAM | HERON::BUCHANAN | combinatorial bomb squad | Thu Mar 08 1990 07:20 | 9 |
The Parallel Associative Memory project has produced a character recognition system which is apparently pretty nifty for signatures, etc. Mat Jaffe is the contact (he's in ELF). This doesn't use Fourier techniques, though, but who cares about the technology. I don't know too much about it, though. Hope this is useful. Regards, Andrew. | |||||
| 1208.2 | Big Subject... | SDOGUS::DRAKE | Dave (Diskcrash) Drake DTN 534-2660 | Tue Mar 13 1990 14:23 | 11 |
Its a big subject. You might locate Bill Pratt's book on Digital
Image Processing for its material. There is some good theory in
Joe Goodman's book on Optical Signal Processing. Also Fukanaga on
Statistical Pattern Recognition. FTs are very good for this
application, but are compute intensive. You can also apply Hadamard
transforms that work almost as well, but only require fixed point
math. I have recently seen demonstrated a powerful two dimensional
associative memory that operates at 40 Billion pixels per second
for very fast character and vector recognition. This system is sold
by AUDRE in Rancho Bernardo, CA, USA, with some very nifty software
support for E-Size drawing recognition. Nice product.
| |||||