net.semanticmetadata.lire.imageanalysis.features.global.correlogram

类 NaiveAutoCorrelogramExtraction

java.lang.Object

net.semanticmetadata.lire.imageanalysis.features.global.correlogram.NaiveAutoCorrelogramExtraction

所有已实现的接口:

IAutoCorrelogramFeatureExtractor

public class NaiveAutoCorrelogramExtraction
extends java.lang.Object
implements IAutoCorrelogramFeatureExtractor

NaiveAutoCorrelogramExtraction is an implementation of the naice approach to extract auto-correlogram feature vector from images (Full Neighborhood is used). It is based on Huang et al paper, "Image Indexing Using Color Correlograms", CVPR1997. J Huang, S. R Kumar, M. Mitra, W. Zhu, R. Zabih

This method is very similar to MLux FullNeighbourhood algorithm, but doesn't acummulate the feature frequence over different distances and uses the standard normalization.

作者:

Rodrigo Carvalho Rezende http://www.rodrigorezende.net/

构造器概要

构造器

构造器和说明
NaiveAutoCorrelogramExtraction()

方法概要

限定符和类型	方法和说明
float[][]	extract(int maxFeatureValue, int[] distanceSet, int[][] img) extract extracts an auto-correlogram from an Image

从类继承的方法 java.lang.Object

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

构造器详细资料

NaiveAutoCorrelogramExtraction

public NaiveAutoCorrelogramExtraction()

方法详细资料

extract

public float[][] extract(int maxFeatureValue,
                         int[] distanceSet,
                         int[][] img)

extract extracts an auto-correlogram from an Image

指定者:

extract 在接口中 IAutoCorrelogramFeatureExtractor

参数:

maxFeatureValue - the maximum feature (color) value

distanceSet - the distance windows of auto-correlogram

img - the image

返回:

float[][] the auto-correlogram A[color][distance]