Dimension Reduction of the SIFT descriptor and its Object Recognition

..................................................................................................................................... 3 Introduction ................................................................................................................................ 4 Stage (I) ....................................................................................................................................... 6 Overview: ................................................................................................................................ 6 (1.1) Brief introduction to OpenCV .......................................................................................... 6 (1.2) A simple SIFT feature matching Experiment using brute force approach .......................... 7 (1.3) Finding “Strong” SIFT Keypoints....................................................................................... 9 (1.4) Compare Average Models ............................................................................................. 12 (1.5) Compute Probability Density Function and Confusion Probabilities ............................... 14 Stage (2) .................................................................................................................................... 17 Overview ........................................................................................................................... 17 (2.1) SIFT keypoints ranking system ....................................................................................... 17 (2.2) Reduce dimensionality of SIFT feature descriptors using t-SNE ...................................... 19 (2.3) A simple object classifier ............................................................................................... 22 Summary and Possible Future Works ........................................................................................ 24 Reference ................................................................................................................................. 24 Appendix ................................................................................................................................... 26 Functional Requirements: ..................................................................................................... 26 Object Identification: ......................................................................................................... 26 Program Input: .................................................................................................................. 26 Program Output: ............................................................................................................... 27 Non-functional Requirements: .............................................................................................. 27 Object recognition ............................................................................................................. 27 Development/Usage platform ........................................................................................... 27