#About

PRec is a panorama recognizer that I wrote while working at Intel Research Lab at the faculty of Computational Mathematics and Cybernetics of the Moscow State University.

PRec automatically recognizes panoramas in a set of input images and stitches them. The process used is described in [1]. PRec is written in C++ and uses Intel Integrated Performance Primitives library for low-level image operations.

The overview of the process is as follows:

• SIFT features are extracted from input images as described in [2].
• SIFT features are matched using fast approximate nearest neighbor search as described in [3].
• For each pair of input images, a RANSAC-class algorithm is used to filter out outlying feature matches.
• Image matches are filtered and a set of panoramas is constructed.
• For each panorama, bundle adjustment is used to jointly optimize camera parameters of input images.

[1] M. Brown and D. G. Lowe. Recognising Panoramas. International Conference on Computer Vision, 2003. [PDF].
[2] D. G. Lowe. Distinctive Image Features from Scale-Invariant Keypoints. International Journal of Computer Vision, 2004. [PDF].
[3] J. S. Beis and D. G. Lowe. Shape indexing using approximate nearest-neighbour search in high-dimensional spaces. Conference on Computer Vision and Pattern Recognition, 1997. [PDF].