| javax.media.jai.UntiledOpImage
 ca.forklabs.media.jai.opimage.KMeansOpImage
| KMeansOpImage | | public class KMeansOpImage extends UntiledOpImage (Code) | | Class
KMeansOpImage is the
OpImage for operator
kmeans.
author:
Daniel Léonard
version:
$Revision: 1.2 $ |
| Method Summary | |
| protected double[] | calculateSquareDistances(double[] pixel, double[][] centers, double[] distances)
Calculate the square of the distance of the pixel from each center pixel.
Parameters:
pixel - the pixel.
Parameters:
centers - the center pixels.
Parameters:
distances - the arrays to store the square distances. | | protected void | computeImage(Raster[] sources, WritableRaster sink, Rectangle bounds)
Computes the k-means image. | | protected int | findClosestCenter(double[] distances)
Gets the position of the smallest distance.
Parameters:
distances - the distances. | | protected int | getClusters()
Gets the desired number of clusters. | | protected int[][] | getColorMap()
Gets the color map. | | protected KMeansDescriptor.EvaluationFunction | getEvaluationFunction()
Gets the evaluation function. | | protected int | getIterations()
Gets the desired number of iterations. | | protected String | getUnknownDataTypeErrorMessage(int type)
Gets the error message telling that the raster data type is unknown.
Parameters:
type - the bad type. | | protected double[][] | initializeCenters(Raster source, Rectangle bounds, KMeansDescriptor.EvaluationFunction function)
Initialize the new centers. | | protected void | resetNewCenters(double[][] positions, double[] population)
Resets the new center data structures. | | protected void | setClusters(int clusters)
Changes the number of clusters. | | protected void | setColorMap(int[][] color_map)
Changes the color map. | | protected void | setEvaluationFunction(KMeansDescriptor.EvaluationFunction function)
Changes the evaluation function. | | protected void | setIterations(int iterations)
Changes the number of iterations. | | protected void | setup(int clusters, KMeansDescriptor.EvaluationFunction function, int iterations, int[][] color_map)
Sets up this image. | | protected void | sortCenters(double[][] centers)
Performs a stable in-place sort of the center from smallest to greatest. |
| clusters | | protected int clusters(Code) | | The number of clusters to make.
|
| color_map | | protected int[][] color_map(Code) | | The color map of the segmented image.
|
| iterations | | protected int iterations(Code) | | The number of iteration to perform.
|
| KMeansOpImage | | public KMeansOpImage(RenderedImage source, int clusters, KMeansDescriptor.EvaluationFunction function, int iterations, int[][] color_map, Map hints, ImageLayout layout)(Code) | | Constructor.
Parameters:
source - the image to segment.
Parameters:
clusters - the number of desired clusters.
Parameters:
function - the evaluation function.
Parameters:
iterations - the maximum number of iterations.
Parameters:
color_map - the color_map of the segmented image.
Parameters:
hints - rendering hints.
Parameters:
layout - image layout. |
| calculateSquareDistances | | protected double[] calculateSquareDistances(double[] pixel, double[][] centers, double[] distances)(Code) | | Calculate the square of the distance of the pixel from each center pixel.
Parameters:
pixel - the pixel.
Parameters:
centers - the center pixels.
Parameters:
distances - the arrays to store the square distances. distances . |
| computeImage | | protected void computeImage(Raster[] sources, WritableRaster sink, Rectangle bounds)(Code) | | Computes the k-means image.
Parameters:
sources - the source rasters.
Parameters:
sink - the sink raster.
Parameters:
the - bounds to work within. |
| findClosestCenter | | protected int findClosestCenter(double[] distances)(Code) | | Gets the position of the smallest distance.
Parameters:
distances - the distances. the position. |
| getClusters | | protected int getClusters()(Code) | | Gets the desired number of clusters.
the number of clusters. |
| getColorMap | | protected int[][] getColorMap()(Code) | | Gets the color map.
the color map. |
| getIterations | | protected int getIterations()(Code) | | Gets the desired number of iterations.
the number of iterations. |
| getUnknownDataTypeErrorMessage | | protected String getUnknownDataTypeErrorMessage(int type)(Code) | | Gets the error message telling that the raster data type is unknown.
Parameters:
type - the bad type. the formatted error message. |
| initializeCenters | | protected double[][] initializeCenters(Raster source, Rectangle bounds, KMeansDescriptor.EvaluationFunction function)(Code) | | Initialize the new centers. It tries to have different values for centers,
but is not always capable of doing so, especially if there are less
segments than clusters.
Parameters:
source - the source.
Parameters:
bounds - the source boundaries.
Parameters:
function - the evaluation function. the initial centers. |
| resetNewCenters | | protected void resetNewCenters(double[][] positions, double[] population)(Code) | | Resets the new center data structures.
Parameters:
positions - the position of the new centers.
Parameters:
population - the population of each new center. |
| setClusters | | protected void setClusters(int clusters)(Code) | | Changes the number of clusters.
Parameters:
clusters - the new number of clusters. |
| setColorMap | | protected void setColorMap(int[][] color_map)(Code) | | Changes the color map.
Parameters:
color_map - the new color map. |
| setIterations | | protected void setIterations(int iterations)(Code) | | Changes the number of iterations.
Parameters:
iterations - the new number of iterations. |
| setup | | protected void setup(int clusters, KMeansDescriptor.EvaluationFunction function, int iterations, int[][] color_map)(Code) | | Sets up this image.
Parameters:
clusters - the number of desired clusters.
Parameters:
function - the evaluation function.
Parameters:
iterations - the number of desired iterations.
Parameters:
color_map - the color map. |
| sortCenters | | protected void sortCenters(double[][] centers)(Code) | | Performs a stable in-place sort of the center from smallest to greatest. A
center is smaller if its square distance from the origin is smaller then
another square distance.
Parameters:
centers - the center to sort. |
|
|