imagingbook.pub.color.quantize

Class KMeansClusteringQuantizer

java.lang.Object

imagingbook.pub.color.quantize.ColorQuantizer

imagingbook.pub.color.quantize.KMeansClusteringQuantizer

public class KMeansClusteringQuantizer
extends ColorQuantizer

This class implements color quantization using k-means clustering of image pixels in RGB color space. Two modes of selecting the colors for the initial clusters are provided: (a) random sampling of the input colors, (b) using the most frequent colors.

Version:

2017/01/04

Nested Class Summary

Nested Classes

Modifier and Type	Class and Description
static class	KMeansClusteringQuantizer.Parameters
static class	KMeansClusteringQuantizer.SamplingMethod

Field Summary

Fields inherited from class imagingbook.pub.color.quantize.ColorQuantizer

MAX_RGB

Constructor Summary

Constructors

Constructor and Description
KMeansClusteringQuantizer(int[] pixels)
KMeansClusteringQuantizer(int[] pixels, KMeansClusteringQuantizer.Parameters params) Creates a new quantizer instance from the supplied sequence of color values (assumed to be ARGB-encoded integers).

Method Summary

Modifier and Type	Method and Description
int[][]	getColorMap() Retrieves the color map produced by this color quantizer.
double	getTotalError() Returns the total error of this clustering, calculated as the sum of the squared distances of the color samples to the associated cluster center.
void	listClusters() Lists the color clusters to System.out (intended for debugging only).

Methods inherited from class imagingbook.pub.color.quantize.ColorQuantizer

findColorIndex, intToRgb, listColorMap, log2, quantize, quantize, quantize, rgbToInt, sqr

Methods inherited from class java.lang.Object

clone, equals, finalize, getClass, hashCode, notify, notifyAll, toString, wait, wait, wait

Constructor Detail

KMeansClusteringQuantizer

public KMeansClusteringQuantizer(int[] pixels,
                                 KMeansClusteringQuantizer.Parameters params)

Creates a new quantizer instance from the supplied sequence of color values (assumed to be ARGB-encoded integers).

Parameters:

pixels - Sequence of input color values.

params - Parameter object.

KMeansClusteringQuantizer

public KMeansClusteringQuantizer(int[] pixels)

Method Detail

listClusters

public void listClusters()

Lists the color clusters to System.out (intended for debugging only).

getTotalError

public double getTotalError()

Returns the total error of this clustering, calculated as the sum of the squared distances of the color samples to the associated cluster center. This calculation is performed during the final iteration.

Returns:

The sum of the squared distances between samples and cluster centers.

getColorMap

public int[][] getColorMap()

Description copied from class: ColorQuantizer

Retrieves the color map produced by this color quantizer. The returned array is in the format int[idx][rgb], where rgb = 0 (red), 1 (green), 2 (blue) and 0 ≤ idx < nColors. This method must be implemented by any derived concrete class.

Specified by:

getColorMap in class ColorQuantizer

Returns:

The table of quantization colors.