ken@turtlevax.UUCP (Ken Turkowski) (07/13/85)
It seems as if discussions about image processing will take place in net.graphics. Pattern recognition is related to image processing. Do discussions about this subject take place in net.graphics? Or is net.ai more appropriate? Net.ai seems to be more occupied with inference, lisp, and prolog, which bore me to tears. Something like net.ai.vision would be more appropriate. Until there is sufficient traffic to warrant the creation of a newsgroup, where should pattern recognition discussions be kept? net.graphics, net.ai, or net.misc? Is anyone else out there interested in pattern recognition? Does anyone have a rudimentary character recognition program for use with either binary or grey-scale images? -- Ken Turkowski @ CADLINC, Menlo Park, CA UUCP: {amd,decwrl,hplabs,nsc,seismo,spar}!turtlevax!ken ARPA: turtlevax!ken@DECWRL.ARPA
herbie@watdcsu.UUCP (Herb Chong [DCS]) (07/15/85)
In article <824@turtlevax.UUCP> ken@turtlevax.UUCP (Ken Turkowski) writes: >It seems as if discussions about image processing will take place in >net.graphics. Pattern recognition is related to image processing. Do >discussions about this subject take place in net.graphics? Or is >net.ai more appropriate? Net.ai seems to be more occupied with >inference, lisp, and prolog, which bore me to tears. Something like >net.ai.vision would be more appropriate. image processing is only a small application of pattern recognition. in my opinion, net.math.stat is a better place, though i suppose with much work still going on in syntactic PR, here is a good a choice as any. the discussions of LISP/PROLOG/C also have me bored to tears. the best AI language, or for that matter, the best programming language has yet to be designed, and will be obsolete by the time it's standardized. another pet peeve of mine is that almost all the AI work i've seen lately is completely ignores errors in input. unless all relevant parameters of a problem can be defined and measured or recreated in the analysis process, the inputs are a stochastic process, not neccesarily stationary. LISP/PROLOG deal with assertions, but i have seen little discussed here that deals with the reliablity of an assertion and its influence on any results. most of my work is in classical PR theory, so i may have just not read the right papers. if i am wrong and there is work going on in this area, then i'd like to hear about it. >Is anyone else out there interested in pattern recognition? Does >anyone have a rudimentary character recognition program for use with >either binary or grey-scale images? i have a program that uses a combination of filtering and relaxation to detect clouds in unenhanced LANDSAT images. i have no idea how good it is because i've only tested it on one image. it was able to segment the image as well as the eye could on the same image. does someone have a set of LANDSAT or related images that they would like to test it on? Herb Chong... I'm user-friendly -- I don't byte, I nybble.... UUCP: {decvax|utzoo|ihnp4|allegra|clyde}!watmath!water!watdcsu!herbie CSNET: herbie%watdcsu@waterloo.csnet ARPA: herbie%watdcsu%waterloo.csnet@csnet-relay.arpa NETNORTH, BITNET, EARN: herbie@watdcs, herbie@watdcsu