Using c++ OpenCV

You are required to create different version of same image (scale, illumination, rotation).

Key stages for SIFT are the following:

scale-space extrema detection,

keypoint localization,

Orientation assignment, and

keypoint descriptor.

In this question, you are asked to implement SIFT and show the results for given images. Please show output of every step separately, make necessary comments in the source code.