Given an image the program determines where the image is taken
Created a trained dictionary of visual words from around 1600 images
The training images are not included
Modify evaluateRecognitionSystem.m to test an input image of your choice - previously used to evaluate and compute the efficiency in recognizing 160 images
Scenes trained include Garden,Tennis Court, Computer Room, Ocean, Library, Art Room
Uses Bag-of-words and Spatial-Pyramid-Matching approches for feature extraction
View Writeup for all details of the implementation