piXlogic's state-of-the-art visual search engine has the following key features:

• extremely fast extraction of logical visual shape information from images (either stills or video)
   
• ability to "see" not only visual objects but also text (language independent)
  
  
• compact mathematical characterization of the extracted objects stored as metadata



• ability to search this metadata rapidly using any image, or portion of an image as a search query

Key value propositions of the piXlogic software:

The amount of digital imagery that is being created and stored is growing exponentially each year. This is occurring across many different marketplaces--private enterprise, government security agencies, and consumers.

We now face a dichotomy--it has never been easier to create and store images, but it is difficult to effectively find and utilize that content without powerful new tools.

The piXlogic software can:

• automate processes that can only be done manually today:
• finding digital images of interest in an asset management system based solely on content
• adding keyword metadata to images based on visual content
• monitoring video streams for targets of interest
• finding objects of interest within video/film
• finding images using text strings that may appear within the picture itself (a street sign, a restaurant name, a logo on a truck or t-shirt, etc.)
• finding video segments by searching for any text that appears on the screen (broadcast news text, scrolling text, commercials, etc.)
   
• use visual search to augment or complement other methods of search (text, audio)
  
  
• Flexible configurations--The software is very compact and multi-threaded. It supports multiple configurations:
  • single user
  • client/server
  • browser, application server
  • mobile devices (PDA, digital camera, camera phones, etc.)
  • real-time video analysis (based on hardware capabilities)

When analyzing an image, the piXlogic software identifies logical visual "objects" within the picture. It stores detailed information about these objects (for example the shape, the color, the position, the orientation, the size, etc. in the image) and does so in a compact mathematical format.  Objects from different pictures can be compared in order to support a visual search.  This compact metadata is stored in database form, and the size of the database typically represents less than 1% of the original size of the image files. piXlogic's patented algorithms represent a new state of the art for accurately indexing and retrieving the unstructured logical content of images and video automatically.

piXlogic offers two products:

• piXserve - A client/server products rich in functionality and suitable for workgroups and for demanding enterprise applications
   
• piXsearch -  An SDK (software developer kit) which can be used to integrate visual search functionality into existing commercial applications.  The API is well-defined, comprehensive and based on industry standard interfaces.