Skip to content

Latest commit

 

History

History
15 lines (8 loc) · 570 Bytes

README.md

File metadata and controls

15 lines (8 loc) · 570 Bytes

Scene-Classification

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