Feature extraction and classification :a case study of classifying a simulated digital mammogram images using self-organizing maps (som)
Feature extraction is important in image processing and is a preliminary step to perform pattern classification. This project aims to propose a feature extraction technique. This feature extraction technique can be used to find five parameters which are the size, intensi...
Saved in:
Main Author: | |
---|---|
Format: | Final Year Project Report |
Language: | English English |
Published: |
Universiti Malaysia Sarawak (UNIMAS)
2007
|
Subjects: | |
Online Access: | http://ir.unimas.my/id/eprint/6728/1/Lau.pdf http://ir.unimas.my/id/eprint/6728/4/Lau%20Leh%20Teen.pdf http://ir.unimas.my/id/eprint/6728/ |
Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
Summary: | Feature extraction is important in image processing and is a preliminary step to perform pattern classification. This project aims to propose a feature extraction technique. This feature extraction technique can be used to find five parameters
which are the size, intensity, centroid X, centroid Y and region distribution of segmented regions
. Several experiments have been conducted to verify the proposed algorithm and feature extraction results obtained will be used for the training of Neural Network classifier, Self-Organizing Maps (SOM).
A set of training input data is used to train SOM. The accuracy of classification performance was acquired. A case study of breast cancer has been
demonstrated in this study by using a simulated digital mammogram images. In this study, the results show that this system is able to perform the classification of mass with low intensity, mass with high intensity, cluster microcalcification, separate microcalcification and special case
to detect abnormality of the digital
mammogram images. |
---|