you can use MATLAB for this there are inbuilt algorithms for SURF and you can export your results to a excel file also which is more convenient for analysis 

http://in.mathworks.com/help/vision/ref/detectsurffeatures.html

http://in.mathworks.com/help/matlab/import_export/exporting-to-excel-spreadsheets.html

I think you have two options here that I can think of. (perhaps there are a lot more efficient)

So first one would be to write a custom function that will take the output image and then query the database.

So, I think here you would need to extract info from output and use it to query the database. Perhaps you would need to devise your own custom search/similarity function. (or one from the literature).

The other option would be to convert database format to algorithm's data format in one-shot and never touch it again.

Then, you would not need to pass each output to an extra function as above, but could directly query the database.

Maybe you can use bag-of-word terminology, which transforms the SUFR features to a histogram of visual vocabulary and after that, you can use similarity method to that histogram.

Some reference:

http://jialu.cs.illinois.edu/technical_notes/CBIR_BoW.pdf

http://www.escience.cn/system/download/65857

You can use Emgu CV; Emgu CV is a cross platform .Net wrapper to the OpenCV image processing library. Allowing OpenCV functions to be called from .NET compatible languages such as C#, VB, VC++, IronPython etc. The wrapper can be compiled by Visual Studio, Xamarin Studio and Unity, it can run on Windows, Linux, Mac OS X, iOS, Android and Windows Phone.

See EmguCV with C# - Tutorial List

http://fewtutorials.bravesites.com/tutorials