MULTIMEDIA SEARCH: FIRST IMAGE INDEXING AND SEARCH SYSTEM MAKES READY

Start-up Recogmission LLC announced an upcoming release of Picollator®, the multimedia search and indexing system, which understands image content and searches through it. Every second photo in the image collections worldwide contains a human face. People manage faces perception in their brains easily, but it is not the same for software, except the innovative solution by Recogmission LLC.

Start-up Recogmission LLC announced an upcoming release of Picollator®, the multimedia search and indexing system, which understands image content and searches through it.

The problem of performing search inside image content is one of the most challenging during last years due to the unprecedented growth of digital multimedia assets in the World. Everybody knows about the difficulty of searching for any particular image in a photo collection or Internet.

Most image library management systems require users to add ‘tags’ to each file, and then use these tags to organize the content. Searches are then made on the basis of the keywords found in those tags. However, tags or text-based search cannot offer the natural way of the information retrieval due to the complexity of the typing proper text and assigning keywords.

Unlike existing solutions, the system developed by Recogmission enables to understand image content. The system works in much the same way as a human brain. The essence of solution is the analogy between image analysis and complex data dynamic performance in the human brain. Picollator® automatically performs objects detection and recognition in images and compose the index to allow users finding the corresponding imaging source as fast as text-based search systems do.

This innovative software is based on original pattern recognition and object detection technologies, developed and patented by Recogmission.

A human face is a key object for Picollator®, because it is the most popular content for digital imaging worldwide. At the same time, faces are probably the most attractive and difficult objects to recognize. Picollator® automatically recognizes all the faces in any set of photos and then links all those faces to events and source files when and where they appear with each other. The result is the index representing the complete set of links between faces and files, created and managed with the minimal user input.

The classification is made on the basis of the content inherent in each image file, so the software minimizes the requirement for human pre-processing. Users get to perform fewer manual operations with image catalogues and they enjoy natural way of search without necessity to keep dates, events and persons in mind.

The most powerful solution provided by Recogmission is the corporate edition which enables local and network file collections processing and fast search by image samples in the distributed environment according to the GRID concept. The home user obtains friendly service for managing photo archives locally.

First release of the software will be demonstrated at the Digital Hollywood Fall in Los Angeles, CA, USA in October 29th - November 1st.

About Company

Recogmission LLC. is focused on the development of the image-based indexing and search system for personal and enterprise multimedia collections. Recogmission has been established in 2006 with the development center located in Samara, Russia. For more information, visit http://www.recogmission.com.