eVision Visual Search Technology
News Technology & Products Business Solutions Developers About eVision

Also See:

Online documentation

eVe SDK signup

eVe Technology Overview

eVision Newsgroup

Discussion Boards

 home > developers > sdk > version 2.1 new features
eVe - visual search toolkit
SDK Beta Version 2.1

eVision is excited to announce the release of the version 2.1 public beta of the eVe visual search SDK. This version of eVe adds even more versatility and functionality to the visual search API.

What's new in version 2.1?

eVe SDK Beta 2.1 adds many features that broaden its functionality and strengthen its presence in the content-based image retrieval marketplace. Major enhancements in version 2.1 include:

1.
100% Java implementation.
Java logoVersion 2.1 removes any dependencies on C++ code and is 100% Java-based. This means it can be ported to any Java-supporting platform including Win32, Linux, Solaris, Mac OS X and embedded systems.

View eVe on Mac OS X
2. A new high-level API, making coding easier and more intuitive.
Version 2.1 introduces a set of new methods to make the high-level API more intuitive. These methods provide easier media object management and search. Methods that are cumbersome to use have been removed.
3. Variable region segmentation in analysis.
You can control the number of object regions that are automatically defined during the image analysis (2-4 is usually the best for photographic images). Regions generally correspond to objects or parts of objects and are represented by the Object Map. This gives you the flexibility to select the optimal number of analyzed regions depending on the image types in your data set.
Object Map
4. Single and Multiple region selection in search (called partial image search).
A user can view the Object Map and select one or more of the regions of the image (say, of the sun setting over the water) and find images with a similar color as the sun, or a similar texture as the water. This lets the user specify which aspects of the image are important to the query. Also the user can view the Object Map of each retrieved image; this information assists greatly in refining the query.
5. Faster image analysis and indexing.
Image analysis and indexing performance improvement can be up to 50% depending on image complexity.
6.

Bug fixes and code optimizations.

To be notified when the next beta version of the eVe SDK is available, simply enter your email address below.
email address
(Note: your email address will only be used to notify you of upcoming Beta SDK releases. It will not be used for any other purpose)