de.lmu.ifi.dbs.jfeaturelib.features

Class Sift

java.lang.Object

de.lmu.ifi.dbs.jfeaturelib.features.AbstractFeatureDescriptor

de.lmu.ifi.dbs.jfeaturelib.features.Sift

All Implemented Interfaces:

Descriptor, FeatureDescriptor

public class Sift
extends AbstractFeatureDescriptor

This descriptor is a wrapper class for the famous SIFT algorithm proposed by David Lowe: http://www.cs.ubc.ca/~lowe/keypoints/ Related papers:
The most complete and up-to-date reference for the SIFT feature detector is given in the following journal paper:
David G. Lowe, "Distinctive image features from scale-invariant keypoints," International Journal of Computer Vision, 60, 2 (2004), pp. 91-110.
The SIFT approach to invariant keypoint detection was first described in the following ICCV 1999 conference paper, which also gives some more information on the applications to object recognition:
David G. Lowe, "Object recognition from local scale-invariant features," International Conference on Computer Vision, Corfu, Greece (September 1999), pp. 1150-1157.
The following paper gives methods for performing 3D object recognition by interpolating between 2D views. It also provides a probabilistic model for verification of recognition.
David G. Lowe, "Local feature view clustering for 3D object recognition," IEEE Conference on Computer Vision and Pattern Recognition, Kauai, Hawaii (December 2001), pp. 682-688. [PDF]; Patents:
Method and apparatus for identifying scale invariant features in an image and use of same for locating an object in an image
David G. Lowe, US Patent 6,711,293 (March 23, 2004). Provisional application filed March 8, 1999. Asignee: The University of British Columbia. Keep in mind that the SIFT algorithm is patented. The SIFT binary is NOT included in this lib in order to avoid violating the patent.

Author:

graf

Nested Class Summary

Nested classes/interfaces inherited from interface de.lmu.ifi.dbs.jfeaturelib.Descriptor

Descriptor.Supports

Field Summary

Fields inherited from class de.lmu.ifi.dbs.jfeaturelib.features.AbstractFeatureDescriptor

pcs

Fields inherited from interface de.lmu.ifi.dbs.jfeaturelib.Descriptor

DOES_ALL

Constructor Summary

Constructors

Constructor and Description
Sift() Initialize the Sift wrapper without a binary being set!
Sift(File siftBinary) Initialize the Sift wrapper with the sift binary file.

Method Summary

Methods

Modifier and Type	Method and Description
String	getDescription() Returns a short plain Text description about the Descriptor and the semantics of the double array obtained from FeatureDescriptor.getFeatures().
List<double[]>	getFeatures() The output feature vectors are a parsed version of the SIFT-binary output.
void	run(ij.process.ImageProcessor ip) Start processing of this algorithm on the given image processor.
void	setProperties(LibProperties properties) Injects the properties class that can be used to retrieve several properties.
EnumSet<Descriptor.Supports>	supports() Determine the capabilities of this algorithm.

Methods inherited from class de.lmu.ifi.dbs.jfeaturelib.features.AbstractFeatureDescriptor

addData, addData, addData, addPropertyChangeListener, endProgress, firePropertyChange, getMask, inMask, setMask, startProgress

Methods inherited from class java.lang.Object

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

Constructor Detail

Sift

public Sift()

Initialize the Sift wrapper without a binary being set!

Sift

public Sift(File siftBinary)
     throws IOException

Initialize the Sift wrapper with the sift binary file. This file is called by a java process that wraps the call and parses the output.

Parameters:

siftBinary -

Throws:

IOException

Method Detail

setProperties

public void setProperties(LibProperties properties)
                   throws IOException

Description copied from interface: FeatureDescriptor

Injects the properties class that can be used to retrieve several properties.

Specified by:

setProperties in interface FeatureDescriptor

Overrides:

setProperties in class AbstractFeatureDescriptor

Throws:

IOException - if something went wrong

getFeatures

public List<double[]> getFeatures()

The output feature vectors are a parsed version of the SIFT-binary output. The features of each vector are assigned as follows:

1. y-coordinate
2. x-coordinate
3. scale factor
4. rotation angle in radians
5. 128 gradient values

Specified by:

getFeatures in interface FeatureDescriptor

Overrides:

getFeatures in class AbstractFeatureDescriptor

Returns:

list of sift features

supports

public EnumSet<Descriptor.Supports> supports()

Description copied from interface: Descriptor

Determine the capabilities of this algorithm.

Specified by:

supports in interface Descriptor

Overrides:

supports in class AbstractFeatureDescriptor

Returns:

the set of supported capabilities.

run

public void run(ij.process.ImageProcessor ip)

Description copied from interface: Descriptor

Start processing of this algorithm on the given image processor. Keep in mind that if the imageProcessor is changed, Supports.NoChanges must not be set. This is also the case even if only ImageProcessor.snapshot() is used.

getDescription

public String getDescription()

Description copied from interface: FeatureDescriptor

Returns a short plain Text description about the Descriptor and the semantics of the double array obtained from FeatureDescriptor.getFeatures(). May return null if no description is implemented - even though this should be avoided by any means.

Returns:

semantic description of getFeatures or null.