Do you have any other good choices? If you do have, please comment. Thanks.

The LIRE (Lucene Image REtrieval) library (Java based)

It is a CBIR system based on Lucene (Java-based) and Caliph and Emir. Caliph-emir is a Java & MPEG-7 based tools for annotation and retrieval of digital photos and images, supporting semantic annotation and content based, metadata based and semantic image retrieval.  Lucene is a full text retrieval package. It is easy to imagine that what jobs LIRE  can do according to its cores.

Also, it’s easy to incorporate context-based image retrieval into LIRE as Lucene  does the job naturally. But Lucene, I think,  is not suitable for academic research.  However,  at lease you can use Caliph-emir to extract the low level features of images, which could also be used in other IR systems.

LIRE can used out-of-the-box.

The GNU Image-Finding Tool(GIFT, C based)

The GIFT (the GNU Image-Finding Tool) is a Content Based Image Retrieval System (CBIRS: http://en.wikipedia.org/wiki/CBIR). It enables you to do Query By Example (QBE: http://en.wikipedia.org/wiki/QBE) on images, giving you the opportunity to improve query results by relevance feedback. For processing your queries the program relies entirely on the content of the images, freeing you from the need to annotate all images before querying the collection.

The GIFT comes with a tool which lets you index whole directory trees containing images in one go. You then can use the GIFT server and its clients, to browse your own image collections

FIRE (CPP)

FIRE is an image retrieval system developed as part of the diploma thesis of Thomas Deselaers. Later, large parts were rewritten to make it more easily maintainable.