Content-Based Image Retrieval Demonstration Software
Fraunhofer Institute of Optronics, System Technologies and Image Exploitation
Content-based Image Retrieval (CBIR) Demonstration Software for searching similar images in databases
Within the EU research project “FAST and efficient international disaster victim Identification” (FASTID) the Fraunhofer-Institute IOSB developed a software module for content based image retrieval. Given an input image, the aim of the software is to find those images within an image database which show the same object or scene as the input image.
The background of the software is to assist the identification of disaster victims in terms of effectiveness and speed. The part of the IOSB within the research project was to support the identification workflow by integrating automated image retrieval methods which operate on images of body modifications such as tattoos . This replaces today's manual classification of tattoos which is always subjective, time consuming and error-prone.
The basic algorithms of the system are however not restricted to images of tattoos but can be utilized in many other domains.
Parts of the system developed in FASTID have therefore been extracted leading to this CBIR software. This enables users to get an idea of the functionality and performance of the algorithms beyond the targets of research projects.
Both the generation of the image database and querying the database can be performed with any image data. Thus, the technology can be easily explored with respect to its applicability in various domains.
Download the demo software now:
- Build your own CBIR image database
- Query your database for similar images
- No internet access needed, your images stay on your computer
- For image databases of up to 50,000 images (ask for a 64bit version for larger databases)
The example image shows how the software is used to search a set of holiday images with a query image from the internet in order to find the holiday images of a distinct object (Golden Gate Bridge).
Make sure you have installed the Microsoft Visual C++ 2008 Redistributable Package (x86).
The research leading to these results has received funding from the European Union Seventh Framework Programme (FP7/2007-2013) under grant agreement no 242339.
 D. Manger, Large-Scale Tattoo Image Retrieval. In 2012 Canadian Conference on Computer and Robot Vision (CRV2012).
 D. Manger, Tattoo Image Retrieval for Forensics. In 6th European Academy of Forensic Science Conference (EAFS2012).
 D. Manger, Content-based Tattoo Image Retrieval. In 21st International Symposium on the Forensic Sciences of the Australian and New Zealand Forensic Science Society (ANZFSS2012).
 D.Manger, Demo: A Tattoo Image Retrieval System, 2012 IEEE International Workshop on Information Forensics and Security (WIFS'12).