MATLAB Interactive Pattern Analysis and Classification System with Application to Handwritten OCR

MATLAB Interactive Pattern Analysis and Classification System with Application to Handwritten OCR

4.11 - 1251 ratings - Source



Pattern Recognition is not a new subject. Its principles and methodologies have for many years influenced the course of technological development in almost every knowledge-based field. No single model exists for all pattern recognition problems, and no single technique is applicable to all problems. Rather, what we have in pattern recognition is a bag of tools and a bag of problems. Interactive Pattern Analysis and Classification System (IPACS) provides a flexible way of analyzing sample patterns and trying out various tools to determine which algorithm or approach should be selected for a given application. Many interactive pattern analyses and classification systems exist in the world. Few of them are general interactive pattern recognition systems. Most of the general systems are developed for special institutes and need special hardware. In this thesis, we present a new general Interactive Pattern Analysis and Classification System (IPACS) developed using MATLAB. The system consists of the major modules necessary for solving most of the practical pattern recognition application problems. These modules include data analysis, data display, feature analysis, and classifier design. Data analysis consists of two non linear mapping algorithms and two clustering algorithms. They are used to determine and analyze the general structure of high dimensional data. Data display includes many linear and non linear mapping techniques which help the user to view the data in one, two, and three dimensional displays. Linear mappings include coordinate, eigenvector, optimal discriminant, and least error planes. Non linear mappings include Sammon algorithm, relaxation algorithm, quadratic plane, and data histogram. Feature analysis anclude three functions: feature evaluation, feature rank, and feature subset selection. Classifier design includes three parametric classifier types and four error estimation methods. Parametric classifiers include nearest mean classifier, linear classifier, and piecewise classifier. Error estimations include Bhattacharayya upper bound, nearest neighbor, resubstitution, and holdout methods. Finally, IPACS is used to develop an optical character recognition system (OCR). 75 lines of code were needed to segment the characters and extract their features which were analyzed to design the classifier. In conclusion, a new general Interactive Pattern Analysis and Classifiction System is developed in MATLAB. The purpose of this system is to provide the user with tools necessary for solving practical pattern recognition problems. IPACS should be of interest to almost every researcher in the field of pattern recognition.Pattern Recognition is not a new subject.


Title:MATLAB Interactive Pattern Analysis and Classification System with Application to Handwritten OCR
Author: Jalal Al Ahmad
Publisher: - 1995
ISBN-13:

You must register with us as either a Registered User before you can Download this Book. You'll be greeted by a simple sign-up page.

Once you have finished the sign-up process, you will be redirected to your download Book page.

How it works:
  • 1. Register a free 1 month Trial Account.
  • 2. Download as many books as you like (Personal use)
  • 3. Cancel the membership at any time if not satisfied.


Click button below to register and download Ebook
Privacy Policy | Contact | DMCA