home *** CD-ROM | disk | FTP | other *** search
- Path: sparky!uunet!ulowell!m2c!nic.umass.edu!noc.near.net!saturn.caps.maine.edu!maine.maine.edu!io10047
- From: IO10047@MAINE.MAINE.EDU
- Newsgroups: comp.ai.neural-nets
- Subject: WANTED: FEATURE EXTRACTION METHOD
- Message-ID: <93026.124724IO10047@MAINE.MAINE.EDU>
- Date: 26 Jan 93 17:47:24 GMT
- Organization: University of Maine System
- Lines: 52
-
- Orono
- Jan 26 1993
-
-
- Dear Sir/Madam,
-
- I am a graduate student in the Master's program at the University of
-
- Maine in Orono.
-
- My thesis mainly focuses on recognition of handwritten numbers using
-
- Neural Networks. The numbers that we are trying to use are "United States
-
- Postal Service Office of Advanced Technology Handwritten ZIP Code Database
-
- (1987)" supplied by the US Postal Service in Washington DC. This database
-
- contains images of handwritten ZIP Codes.
-
- We are trying to use a set of features for the identification task.
-
- We would be grateful if you could supply us with a database of features that
-
- you might have used in your research.
-
- Even if you don't have the database available, any software that can
-
- provide us with the following functions would be highly appreciated:
-
- - Binarization
-
- - Normalization
-
- - Separation
-
- - Feature Extraction
-
- - Identification using Neural Networks
-
-
- You can call me at: (207) 581-2025 / (207) 866-3606 or
-
- email at: sharma@watson.eece.maine.edu
-
-
- Thanking You.
-
-
- Sincerely
-
- Lalit Sharma
-