org.openimaj.demos

Class BasicDuplicateImageDatabase<T>

java.lang.Object

org.openimaj.demos.BasicDuplicateImageDatabase<T>

Type Parameters:

T - type of metadata being stored

@Reference(author={"Wei Dong","Zhe Wang","Kai Li"},
           title="High-Confidence Near-Duplicate Image Detection",
           type=Inproceedings,
           year="2012",
           booktitle="ACM International Conference on Multimedia Retrieval",
           customData={"location","Hong Kong, China"})
public class BasicDuplicateImageDatabase<T>
extends Object

An in-memory image duplicate database. Stores hashes of features as the basis of determining duplicates.

This is a basic in-memory implementation of the approach described by Dong, Wang and Li for high-confidence duplicate detection. It does not currently implement the query expansion phase based on a graph cut of the duplicity graph, but instead relies directly on a simple thresholded count of collisions.

Author:

Jonathon Hare (jsh2@ecs.soton.ac.uk), Jonathon Hare (jsh2@ecs.soton.ac.uk)

Field Summary

Fields

Modifier and Type	Field and Description
protected List<gnu.trove.map.hash.TIntObjectHashMap<Set<T>>>	database

Constructor Summary

Constructors

Constructor and Description
BasicDuplicateImageDatabase()

Method Summary

Modifier and Type	Method and Description
int[][]	extractFeatures(FImage image) Extract hashed features from an image
int[][]	extractFeatures(MBFImage image) Extract hashed features from an image
void	indexImage(int[][] features, T metadata) Index a new image based on its features
static void	main(String[] args)
List<ObjectIntPair<T>>	search(int[][] features) Search for a given image (represented by its features) in the database

Methods inherited from class java.lang.Object

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

Field Detail

database

protected List<gnu.trove.map.hash.TIntObjectHashMap<Set<T>>> database

Constructor Detail

BasicDuplicateImageDatabase

public BasicDuplicateImageDatabase()

Method Detail

extractFeatures

public int[][] extractFeatures(FImage image)

Extract hashed features from an image

Parameters:

image - the image to index

Returns:

the features

extractFeatures

public int[][] extractFeatures(MBFImage image)

Extract hashed features from an image

Parameters:

image - the image to index

Returns:

the features

indexImage

public void indexImage(int[][] features,
                       T metadata)

Index a new image based on its features

Parameters:

features - the image to index

metadata - the metadata for the image

search

public List<ObjectIntPair<T>> search(int[][] features)

Search for a given image (represented by its features) in the database

Parameters:

features - the features

Returns:

the list of matching images and their scores (bigger == more chance of match)

main

public static void main(String[] args)
                 throws MalformedURLException,
                        IOException

Throws:

MalformedURLException

IOException