net.sourceforge.javaocr.cluster
Interface Metric

public interface Metric

metric computes "distance" between feature cluster extracted from candidate image and data stored inside match.

Author:

Konstantin Pribluda

Method Summary

double	distance(double[] features) compute distance between metric and feature vectors.
int	getDimensions() amount of dimensions of feature vectors

Method Detail

distance

double distance(double[] features)

compute distance between metric and feature vectors. Individual plugins will define concrete metric implementation ( euclidian, mahalonoubis, battacharya etc. )

Parameters:

features - amoutn of fweatures shall correspond to amount dimenstions

Returns:

distance between features and vector

getDimensions

int getDimensions()

amount of dimensions of feature vectors

Returns:

amount of dimensions