Saturday, December 27, 2008
In the first paper we present a new low level compact composite descriptor for Content Based Medical Image Retrieval.
Abstract: The rapid advances made in the field of radiology, the increased frequency in which oncological diseases appear, as well as the demand for prevailing medical checks, led to the creation of a large database of radiology images in every hospital or medical center. There is now an imperative need to create an effective method for the indexing and retrieval of these images. This paper proposes a new method for content based medical image retrieval. The description of images relies on a new Composite Descriptor (CD) which includes global image features, capturing both brightness and texture characteristics at the same time. Image information is extracted using a set of fuzzy approaches. To be applicable in the design of large medical image databases, the proposed descriptor is compact, requiring only 48 bytes per image. Experiments demonstrate the effectiveness of the proposed technique. Authors: Savvas A. Chatzichristofis and Yiannis Boutalis.
The second paper is presenting a method for auto selection the proper compact composite descriptor in order to retrieve natural color images.
Abstract: Compact Composite Descriptors (CCD) are global image features capturing both, color and texture characteristics, at the same time in a very compact representation. In this paper we propose a combination of two recently introduced CCDs (CEDD and FCTH) into a Joint Composite Descriptor (JCD). We further present a method for descriptor selection to approach the best ANMRR that would result from CEDD and FCTH. With our approach the most appropriate descriptor in terms of maximization of information content can be found on a per image basis without knowledge of the data set as a whole. Experiments conducted on three known benchmarking image databases demonstrate the effectiveness of the proposed technique. Authors: Savvas A. Chatzichristofis, Mathias Lux and Yiannis Boutalis.
The descriptors will be added soon in the CCD section.
Monday, December 22, 2008
Mathias Lux is working on a summary tool, which extracts still images from a video. The goal of the tool is to find frames, which describe the image in an optimal way. Now it’s in a (rather) stable state and ready to release. For Windows users it should be a single click to start the thing, for Linux you need to install ffmpeg. Note that the tool is open source & the code is GPL-ed
Saturday, December 20, 2008
All of these options can be selected from the "Any content" drop down in the blue title bar on any search results page, or by selecting one of the "Content types" on the Advanced Image Search page. The good news: no extra typing! In all these examples our query remained exactly the same, we just restricted our results to different visual styles. So whether you're interested holiday wreaths, Celtic patterns, or office clip art, it just became a lot easier to find the images you're looking for.
Thursday, December 18, 2008
Original painting time 2hrs 30mins.
So what are the things to look for if you want to buy digital camera? To be able to answer these, there are 2 sets of information you have to know before you can decide. The first type of information is defining what YOU need and want in a digital camera. To do this, you can ask yourself the following questions:
What do you want to take with your digital camera? Before you buy digital camera, it is important to determine what kind of pictures you want to take with it. If you are a digital photography enthusiast, any digital camera will not just do. You have to look for features that can support the zooming you need, the resolution, etc.
How much is your budget? This is a very important question any person who intends to buy digital camera should ask. Because no matter what your needs and wants are for the device, your financial resource will play a huge part in dictating the type of digital camera you will buy.
What are you resources? When you buy digital camera, sometimes the spending does not end there. You also have to consider the capacity and the power of the computer and the printer you will be hooking your camera with for your editing and printing needs. Editing software are already included when you buy digital camera but other devices aren’t. Aside from a printer, ink and paper for printing, you might also need additional memory cards for your camera and a more powerful computer to support image editing and image storage and retrieval.
Wednesday, December 17, 2008
Recently, standard benchmark databases and evaluation campaigns have been created allowing a quantitative comparison of CBIR systems. These benchmarks allow the comparison of image retrieval systems under different aspects: usability and user interfaces, combination with text retrieval, or overall performance of a system.
1. WANG database
The WANG database is a subset of 1,000 images of the Corel stock photo database which have been manually selected and which form 10 classes of 100 images each. The WANG database can be considered similar to common stock photo retrieval tasks with several images from each category and a potential user having an image from a particular category and looking for similar images which have e.g. cheaper royalties or which have not been used by other media. The 10 classes are used for relevance estimation: given a query image, it is assumed that the user is searching for images from the same class, and therefore the remaining 99 images from the same class are considered relevant and the images from all other classes are considered irrelevant
2. The MIRFLICKR-25000 Image Collection
Access to the collection is simple and reliable, with image copyright clearly established. This is realized by selecting only images offered under the Creative Commons license. See the copyright section below.
Images are also selected based on their high interestingness rating. As a result the image collection is representative for the domain of original and high-quality photography.
In particular for the research community dedicated to improving image retrieval. We have collected the user-supplied image Flickr tags as well as the EXIF metadata and make it available in easy-to-access text files. Additionally we provide manual image annotations on the entire collection suitable for a variety of benchmarks.
MIRFLICKR-25000 is an evolving effort with many ideas for extension. So far the image collection, metadata and annotations can be downloaded below. If you enter your email address before downloading, we will keep you posted of the latest updates.
3. UW database
The database created at the University of Washington consists of a roughly categorized collection of 1,109 images.These images are partly annotated using keywords. The remaining images were annotated by our group to allow the annotation to be used for relevance estimation; our annotations are publicly available10.The images are of various sizes and mainly include vacation pictures from various locations. There are 18 categories,for example “spring ﬂowers”, “Barcelona”, and “Iran”. Some example images with annotations are shown in Figure 2. The complete annotation consists of 6,383 words with a vocabulary of 352 unique words. On the average, each image has about 6 words of annotation. The maximum number of key-words per image is 22 and the minimum is 1. The database is freely available11. The relevance assessment for the experiments with this database were performed using the annotation: an image is considered to be relevant w.r.t. a given query image if the two images have a common keyword in the annotation. On the average, 59.3 relevant images correspond to each image. The keywords are rather general; thus for example images showing sky are relevant w.r.t. each other,which makes it quite easy to ﬁnd relevant images (high precision is likely easy) but it can be extremely diﬃcult to obtain a high recall since some images showing sky might have hardly any visual similarity with a given query.This task can be considered a personal photo retrieval task,e.g. a user with a collection of personal vacation pictures is looking for images from the same vacation, or showing the same type of building.
4. IRMA-10000 database
5. ZuBuD database
The “Zurich Buildings Database for Image Based Recognition”(ZuBuD) is a database which has been created by the Swiss Federal Institute of Technology in Zurich. The database consists of two parts, a training part of 1,005images of 201 buildings, 5 of each building and a query part of 115 images. Each of the query images contains one of the buildings from the main part of the database. The pictures of each building are taken from diﬀerent viewpoints and some of them are also taken under diﬀerent weather conditions and with two diﬀerent cameras. Given a query image, only images showing exactly the same building are considered relevant.
6. UCID database (Suggested)
The UCID database13 was created as a benchmark database for CBIR and image compression applications. This database is similar to the UW database as it consists of vacation images and thus poses a similar task.For 264 images, manual relevance assessments among all database images were created, allowing for performance evaluation. The images that are judged to be relevant are images which are very clearly relevant, e.g. for an image showing a particular person, images showing the same person are searched and for an image showing a football game, images showing football games are considered to be relevant. The used relevance assumption makes the task easy on one hand,because relevant images are very likely quite similar, but on the other hand, it makes the task diﬃcult, because there are likely images in the database which have a high visual similarity but which are not considered relevant. Thus, it can be diﬃcult to have high precision results using the given rel-evance assessment, but since only few images are considered relevant, high recall values might be rather easy to obtain.
<Yaroslav Bulatov> I've collected this dataset for a project that involves automatically reading bibs in pictures of marathons and other races. This dataset is larger than robust-reading dataset of ICDAR 2003 competition with about 20k digits and more uniform because it's digits-only. I believe it is more challenging than the MNIST digit recognition dataset.
I'm now making it publicly available in hopes of stimulating progress on the task of robust OCR. Use it freely, with only requirement that if you are able to exceed 80% accuracy, you have to let me know ;)
The dataset file contains raw data (images), as well as Weka-format ARFF file for simple set of features.
For completeness I include matlab script used to for initial pre-processing and feature extraction, Python script to convert space-separated output into ARFF format. Check "readme.txt" for more details.
- Database of thousands of weakly labelled, high-res images. Please, click here to download the database.
- Pixel-wise labelled image database v1 (240 images, 9 object classes). Please, click here to download the database. This database was used in paper 1 below and in the above demo video.
- Pixel-wise labelled image database v2(591 images, 23 object classes). Please, click here to download the database.
- Pixel-wise labelled image database of textile materials. Please, click here to download the database.
1. Deselaers, T., Keysers, D., and Ney, H. 2008. Features for image retrieval: an experimental comparison. Inf. Retr. 11, 2 (Apr. 2008), 77-107. DOI=http://dx.doi.org/10.1007/s10791-007-9039-3
2. S. A. Chatzichristofis, K Zagoris, Y. S. Boutalis and Nikolas Papamarkos, “ACCURATE IMAGE RETRIEVAL BASED ON COMPACT COMPOSITE DESCRIPTORS AND RELEVANCE FEEDBACK INFORMATION”, «International Journal of Pattern Recognition and Artificial Intelligence (IJPRAI) », to Appear, 2009
Monday, December 15, 2008
1. Spatial Fuzzy Brightness and Texture Directionality Descriptor (New Descriptor). This descriptor is suittable for Content Based Medical Image Retrieval.
2. New Quantization Tables for Brightness and Texture Directionality Descriptor
3. Gustafson Kessel Classifier for Color Reduction
1. Spatial Fuzzy Brightness and Texture Directionality Descriptor (New Descriptor). This descriptor is suittable for Content Based Medical Image Retrieval.
2. Get the Precision->Recall Graph after the retrieval procedure.
Download the latest version
Friday, December 12, 2008
And what's more, Multiple Image Resizer .NET is FREE for personal and educational use! Commercial users of Multiple Image Resizer .NET should buy a commercial use license from us.
Multiple Image Resizer .NET also has a completely customisable user interface that you can arrange to suit yourself.
Have a look at the features page for more information about what Multiple Image Resizer .NET is capable of.
Latest News, 3rd December 2008:
The software has been updated to support the Greek language.
Multiple Image Resizer .NET's user interface now supports 14 different languages. If you would like to see the software in your own language then why not provide a translation - see the translate page for more information.
Multiple Image Resizer .NET Version 220.127.116.11 is now available for download.
Images can be scenes; landmarks; objects; graphics; people or even personalities. Irrespective of the size of the image collection Imprezzeo Image Search helps users find the right image - fast. For a demonstration, visit www.imprezzeo.com.
The potential for this technology is huge. Stock photo libraries and news agencies can provide more relevant search results to image buyers (and sell content that might not have been found using traditional text-based search); search engines can provide users with a far more sophisticated image search experience than is currently available; photo sharing sites can offer search by example image, rather than search that solely relies on user’s tagging; consumers can organise their personal photo collections by content and person rather than by date; even retailers can recommend similar products for purchase (for example, if a consumer is searching for a ‘red handbag’, Imprezzeo could be used to find all similar products).
The proprietary search software is more sophisticated than any other image-based search technology on the market, combining content-based image retrieval (CBIR) and facial recognition technologies. Imprezzeo’s sophisticated image analytics ensures the Imprezzeo platform can deliver great results when applied to a whole range of different image content and when used in a whole range of applications.
The new technology generates image search results that closely match a sample image either chosen by the user from an initial set of search results that can then be refined, or from an image uploaded by the user.
Dermot Corrigan, CEO of Imprezzeo, says: “This will fundamentally change the way users and consumers expect to search for images, whether that’s in a photo stock library, the desktop or the web. We know that currently many image searches are abandoned at the first set of results, because the returned results are not what the user is looking for. This technology changes that.”
Thursday, December 11, 2008
Sunday, December 7, 2008
1. Tamura Texture Directionality Histogram (Bug Fixed)
2. Auto Correlograms using several methods and max distances
3. Color Histogram Crisp Linking
4. Brightness and Texture Directionality Descriptor (New Descriptor)
5. Scalable Fuzzy Brightness and Texture Directionality Descriptor (New Descriptor)
6. Spatial Color Layout (Beta Version) (New Descriptor)
7. Color Reduction Using Gustafson Kessel
8. Joint Composite Descriptor (Final Version) (New Descriptor)
9. Auto Descriptor Selector (Final Version)
10. Color Histograms (RGB)
11. Auto Correlogram
12. Tamura Texture
13. Evaluate retrieval results using ANMRR and/or Mean Average Precision
14. Faster creation of index files
15. .Net Framework 3.5 Support
16. Retrieve images form sketches using the beta version of "Spatial Color Layout" (New Descriptor)
17. Retrieve images from Sketches using "Color Layout Descriptor"
18. Retrieve images form sketches using the beta version of "Spatial Color Layout" (New Descriptor)
Thursday, December 4, 2008
The idea of ImageSorter is to find images of which you remember how they look but you forgot in which folder they were. If one or several folders are selected, all images from these folders will be visually arranged such that similar images are close to each other. In this sorted display it will be much easier to find a particular image. Selected images can be copied, moved or deleted (right mouse click).ImageSorter does cache thumbnails and sortings, therefore after images have been loaded once, everything will be much faster.
The current version of ImageSorter is 3.0 BETA 3 for Windows and V2.0.2 for Mac OS X. ImageSorter 3 introduced an Internet image search (Yahoo! and Flickr) and the possibility to search for similar images on the local disk or the Internet. The software profits from a greatly improved stability and run-time performance. Furthermore quite a few suggestions made in the forum have been included. See the change log for a detailed list of changes
Download the ImageSorter software
Tuesday, December 2, 2008
Monday, December 1, 2008
The proposed scope of CAIP09 includes, but not limited to, the following areas:
* 3D Vision
* 3D TV
* Color and texture
* Document analysis
* Graph-based Methods
* Image and video indexing and database retrieval
* Image and video processing
* Image-based modeling
* Kernel methods
* Medical imaging
* Mobile multimedia
* Model-based vision approaches
* Motion Analysis
* Non-photorealistic animation and rendering
* Object recognition
* Performance evaluation
* Segmentation and grouping
* Shape representation and analysis
* Structural pattern recognition
Invited speakers (incomplete yet)
David G. Stork (Ricoh Innovations and Stanford University, USA) Aljoscha Smolic (Fraunhofer Institute for Telecommunications, Germany)