Next:
Introduction
Up:
No Title
Previous:
Discussion
A Ten Class Problem
Introduction
Materials and Methods
Fluorescence Microscopy
Image Processing
Zernike Features
Haralick texture features
ad hoc
features
Features Calculated Using Only the Protein Localization Image
Features Calculated Using Both the DNA and Protein Localization Images
Features Calculated Using the Convex Hull of the Protein Fluorescence
Feature Selection
Back-Propagation Neural Network
k-Nearest Neighbor Method
Results
Image Collection and Processing
Feature Extraction
Zernike and Haralick Features
ad hoc
Features
Classification with All Features
Classification with the
ad hoc
Features
Classification with the ``best'' Features
Classification Without a DNA Image
Classification of Images at Lower Resolution
Classification of
Sets
of Images
Discussion
Copyright ©1999 Michael V. Boland
1999-09-18
