Near-duplicate image detection from million images (using C#)

Hi,

We need to develop near-duplicate image detection from a database of million images

The basic concept is to use SURF (or SIFT) with min-hash and index it using LSH.

* SIFT for image signature
* min-hash for finding similarity
* locality-sensitive-hashing (LSH) to perform indexing

To be developed in Visual Studio as a C# console app.
Sample images and corresponding paper will be shared in message.

Thanks