sujaynarumanchi / lire Goto Github PK
View Code? Open in Web Editor NEWAutomatically exported from code.google.com/p/lire
License: Other
Automatically exported from code.google.com/p/lire
License: Other
LIRE - Lucene Image REtrieval (A Homage to Lots of Great Holidays in Italy) ========================================================== The LIRE library allows the creation of a Lucene Index for content based image retrieval (CBIR). Furthermore methods for searching the index are provided. This library is part of the Caliph & Emir project and aims to provide the CBIR features to other Java projects in an easy and light weight way. For more information on the usage of LIRE see the online docs: http://www.semanticmetadata.net/wiki/doku.php?id=lire:lire Acknowledgements I'd like to thank the numerous people having contributed code to Lire or having published their work as open source allowing integration into LIRe: Savvas A. Chatzichristofis, Manuel Oraze, Lukas Esterle, Roman Kern, Roman Divotkey, Katharina Tomanec, Fabrizio Falchi, Bastian Hösch, Janine Lachner, Marko Keuschnig, Christian Penz and Benjamin Sznajder. - Mathias Lux, Klagenfurt, 2008-06-08 http://www.SemanticMetadata.net
Hi all
Maven's <em>interesting</em> xml parser seems to dislike the lire copyright
header, attached is a very small patch that hopefully makes this work a little
better.
Original issue reported on code.google.com by [email protected]
on 13 Aug 2013 at 5:29
Attachments:
net.semanticmetadata.lire.imageanalysis.LocalBinaryPatterns cannot be used (it
looks like it has been deactivated)
What steps will reproduce the problem?
1. add lire.jar v0.9.5 to a project
2. use the ParallelExtractor to extract the Local Binary Patterns feature from
a set of images
What is the expected output? What do you see instead?
The extraction process raises a NullPointerException
What version of the product are you using? On what operating system?
Code running under Intellij Idea 13.0.1 on Windows 7 SP1 64 bit
Please provide any additional information below.
The problem lies in net.semanticmetadata.lire.indexing.tools.ParallelExtractor
where the LocalBinaryPatterns is not added in the member called "features":
public static final String[] features = new String[]{
"net.semanticmetadata.lire.imageanalysis.CEDD", // 0
"net.semanticmetadata.lire.imageanalysis.FCTH", // 1
"net.semanticmetadata.lire.imageanalysis.OpponentHistogram", // 2
"net.semanticmetadata.lire.imageanalysis.joint.JointHistogram", // 3
"net.semanticmetadata.lire.imageanalysis.AutoColorCorrelogram", // 4
"net.semanticmetadata.lire.imageanalysis.ColorLayout", // 5
"net.semanticmetadata.lire.imageanalysis.EdgeHistogram", // 6
"net.semanticmetadata.lire.imageanalysis.Gabor", // 7
"net.semanticmetadata.lire.imageanalysis.JCD", // 8
"net.semanticmetadata.lire.imageanalysis.JpegCoefficientHistogram",
"net.semanticmetadata.lire.imageanalysis.ScalableColor", // 10
"net.semanticmetadata.lire.imageanalysis.SimpleColorHistogram", // 11
"net.semanticmetadata.lire.imageanalysis.Tamura", // 12
"net.semanticmetadata.lire.imageanalysis.LuminanceLayout", // 13
"net.semanticmetadata.lire.imageanalysis.PHOG", // 14
};
I could fix by just adding
"net.semanticmetadata.lire.imageanalysis.LocalBinaryPatterns" as additional
entry.
Original issue reported on code.google.com by [email protected]
on 20 Apr 2015 at 8:14
FuzzyColorHistogram#getDoubleHistogram()
throws an UnsupportedOperationException.
I'd expect to return the int array.
The patch includes an additional toDoubleArray method in the serialization
class + an implementation of getDoubleHistogram()
Original issue reported on code.google.com by franz.graf
on 8 Jul 2013 at 7:01
Attachments:
hi,I find that this project support lucene4.0 and the svn commit logs says
that: sift test successfully.
so,I have a try.
I copy the src/test/java/...../lire/indexing/IndexVisualWordsTest.java and run
it.
but i found when the number of images which will be indexed are very small
(for example , only 5 images should be indexed), then when run the code:
-----------------
System.out.println("-< Creating SURF based histograms >--------------");
SurfFeatureHistogramBuilder surfFeatureHistogramBuilder = new
SurfFeatureHistogramBuilder(IndexReader.open(FSDirectory.open(new
File(index))), numSamples, clusters);
surfFeatureHistogramBuilder.index();
---------------------
there will be a bug: when run k-means algorithm.
the console print: k.getFeatureCount() = 0
then there will be a NullPoint at:KMeans.init():
------------------------------
Set<Integer> medians = selectInitialMedians(numClusters);
Iterator<Integer> mediansIterator = medians.iterator();
------------------------------
if I debug this code, I can find that medians=null.
But when I add images to be indexed, for example 20 images, the issue will
disappear.
Does this means surf need many images ? and sift need more?(even though I use
20 images to test sift algorithm, the NullPoint exception appears . )
Original issue reported on code.google.com by [email protected]
on 10 Dec 2012 at 6:20
int line 87:
System.arraycopy(descriptor, 0, clusters[i].mean, 0, descriptor.length);
clusters[i].mean is float[], but descriptor is double[], so it will cause
java.lang.ArrayStoreException
I am use version 0.9.3
Original issue reported on code.google.com by [email protected]
on 31 Mar 2014 at 2:28
This patch upgrades the Maven POM file for the following:
1. Remove a parsing error by removing '--' from the copyright header and
replace it with '=='
2. Use a Mavenized and upgraded 'jopensurf' library so as not to rely on system
dependency
3. Add the WEKA library dependency for preventing missing dependencies
Original issue reported on code.google.com by [email protected]
on 11 Dec 2013 at 6:19
Attachments:
in function createShape()
Line 193: int[] pixel = {0, 0, 0};
If the input image with alpha channel
Exception in thread "main" java.lang.ArrayIndexOutOfBoundsException: 3
at java.awt.image.ComponentSampleModel.getPixel(Unknown Source)
at java.awt.image.Raster.getPixel(Unknown Source)
at net.semanticmetadata.lire.imageanalysis.mpeg7.ColorLayoutImpl.createShape(ColorLayoutImpl.java:196)
Suggestion:
int[] pixel = {0, 0, 0, 0};
Original issue reported on code.google.com by [email protected]
on 1 Aug 2013 at 2:15
https://code.google.com/p/lire/source/browse/trunk/src/main/java/net/semanticmet
adata/lire/utils/SerializationUtils.java#245
In the 0.9.4-beta2 code, the function seems to shift the output array by
offset, and chop off the rest of the float array at the end.
public static float[] toFloatArray(byte[] in, int offset, int length) {
float[] result = new float[length / 4];
byte[] tmp = new byte[4];
for (int i = offset; i < length / 4; i++) {
System.arraycopy(in, (i - offset) * 4 + offset, tmp, 0, 4);
result[i] = toFloat(tmp);
}
return result;
}
For comparison, the code for converting to double array is correct:
public static double[] toDoubleArray(byte[] data, int offset, int length) {
double[] result = new double[length / 8];
byte[] tmp = new byte[8];
for (int i = 0; i < result.length; i++) {
System.arraycopy(data, i * 8 + offset, tmp, 0, 8);
result[i] = toDouble(tmp);
}
return result;
}
Original issue reported on code.google.com by [email protected]
on 12 Aug 2014 at 9:49
What steps will reproduce the problem?
1. add only one image to index with help of CEDD documentbuilder
2. search by this image (note, same image as you added to index)
3. in ImageSearchHits you will have only one record, calling hits.score(0) will
throw java.lang.AssertionError exception
What is the expected output? What do you see instead?
expected - 0; exception instead
What version of the product are you using? On what operating system?
0.9.3, Windows 7x64
Please provide any additional information below.
I attached source files for your convenience
thanks
Original issue reported on code.google.com by [email protected]
on 13 Oct 2014 at 4:54
Attachments:
Add generics type-safety to code.
Original issue reported on code.google.com by [email protected]
on 11 Dec 2013 at 6:48
Attachments:
1. StringField is not tokenized, no matter stored or not(lucene 4.2.1)
2. line 48 in FieldType.java :private boolean tokenized = true;
3. so, any StringField will be tokenized if display it's filedtype. would some
mistakes occurs?
4. until lucene 4.8.0 ,keep unchanged .
Original issue reported on code.google.com by [email protected]
on 13 May 2014 at 12:21
What steps will reproduce the problem?
1. Run the Lire-demo
2. Index a database
3. Search with SURF BOvW
What is the expected output? What do you see instead?
No results, even though the features were extracted and saved.
What version of the product are you using? On what operating system?
LireDemo-0.9.4-beta_2.zip on Ubuntu 12.04
Please provide any additional information below.
I tried to work out examples with Sift/Surf/MSER using VisualWordsImageSearcher
class. But most of the extraction methods are not implemented. I guess the
search mechanism is different for bang of words kind of features. Can someone
let me know the status of the aforementioned features and if they are in place,
how can I make use of them.
regards
Mohit
Original issue reported on code.google.com by [email protected]
on 20 Sep 2013 at 12:38
This patch fixes the compilation errors in
FloatArray2DScaleOctaveDoGDetector.java by using Commons Math 3 library for
ArrayMatrix handling, rather than relying on version 2 of library.
Original issue reported on code.google.com by [email protected]
on 11 Dec 2013 at 6:23
Attachments:
This patch includes the fixes to the following issues:
* fixes warnings on used code
* fixes dead code warning
* fixes an obvious null pointer exception in GenericDocumentBuilder
* adds type-safety by removing raw-type usage
* fixed resource leak by closing all streams
* fixed lucene IndexReader.open() deprecation warnings
Hope this helps.
Original issue reported on code.google.com by [email protected]
on 26 Jun 2013 at 8:38
Attachments:
The line 152 of Class LocalFeatureHistogramBuilder.java is
if (features.size() < numClusters)
but I think it should be
if (k.getFeatureCount() < numClusters)
because features only stores one images features.
Original issue reported on code.google.com by [email protected]
on 21 Jul 2013 at 9:21
The source codes of net.semanticmetadata.lire.clustering.Image.java and
net.semanticmetadata.lire.imageanalysis.sift.Found.java are missing in the zip
file
Original issue reported on code.google.com by [email protected]
on 26 Jul 2014 at 9:07
Hi, I am using lire.jar as a library in one of my applications. I need to
serialise some classes that contain lire classes implementing the LireFeature
interface, but currently that is not possible because the lire classes are not
serialisable. Could the LireFeature interface be modified to implement
serialisable? Alternatively the individual classes implementing LireFeature
could implement serialisable instead.
Thanks
Mike
Original issue reported on code.google.com by [email protected]
on 16 Feb 2015 at 8:23
What steps will reproduce the problem?
CEDD cedd = new CEDD();
cedd.Compact = true;
cedd.extract( ImageIO.read(new File("c://myfile.jpg")));
System.out.println(cedd.getStringRepresentation());
What is the expected output? What do you see instead?
desired:
cedd 60 5 3 1 3 1 0 0 0 0 0 1 2 0 0 2 0 0 0 0 0 4 6 0 3 4 0 0 0 3 0 1 3 0 1 4 0
0 0 1 0 0 1 0 0 1 0 0 0 0 0 1 2 0 0 1 0 0 0 0 0
instead: (filling with 0s)
cedd 144 5 3 1 3 1 0 0 0 0 0 1 2 0 0 2 0 0 0 0 0 4 6 0 3 4 0 0 0 3 0 1 3 0 1 4
0 0 0 1 0 0 1 0 0 1 0 0 0 0 0 1 2 0 0 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
What version of the product are you using? On what operating system?
0.9.4 (trunk)
Please provide any additional information below.
If thats how it should be the tanimoto similarity is slower than having only to
loop 60 bits
Original issue reported on code.google.com by [email protected]
on 29 May 2013 at 11:38
in
net.semanticmetadata.lire.imageanalysis.sift.Feature.getByteArrayRepresentation
method
descriptor is a double array. so that SerializationUtils.toBytes(descriptor[0])
return 8 bit byte array. but only first 4 byte is using to serilazing and
deserilazing.
i attached patch code.
Original issue reported on code.google.com by [email protected]
on 20 Jan 2015 at 4:36
Attachments:
A declarative, efficient, and flexible JavaScript library for building user interfaces.
🖖 Vue.js is a progressive, incrementally-adoptable JavaScript framework for building UI on the web.
TypeScript is a superset of JavaScript that compiles to clean JavaScript output.
An Open Source Machine Learning Framework for Everyone
The Web framework for perfectionists with deadlines.
A PHP framework for web artisans
Bring data to life with SVG, Canvas and HTML. 📊📈🎉
JavaScript (JS) is a lightweight interpreted programming language with first-class functions.
Some thing interesting about web. New door for the world.
A server is a program made to process requests and deliver data to clients.
Machine learning is a way of modeling and interpreting data that allows a piece of software to respond intelligently.
Some thing interesting about visualization, use data art
Some thing interesting about game, make everyone happy.
We are working to build community through open source technology. NB: members must have two-factor auth.
Open source projects and samples from Microsoft.
Google ❤️ Open Source for everyone.
Alibaba Open Source for everyone
Data-Driven Documents codes.
China tencent open source team.