Project 3 SIFT Matching by Binary SIFT
Experiment Setting Image database – UKBench dataset: images from 2550 categories Feature extraction – SIFT feature The source code will be provided –
Tasks Binary SIFT Generation – Transform each SIFT to a 256-bit binary signature Feature matching based on q Binary SIFT – Two features from two images are considered as a match if the Hamming distance between their binary signatures is less than t – Test the impact of t on feature matching performance – Select 5 relevant image pairs to conduct feature matching – Select 5 irrelevant image pairs to conduct feature matching
Implementation Program with C++ or Matlab – You may need Open CV when programing with C++ OpenCV 210 library files are provided – You do not need to install the OpenCV source file Refer to OpenCV China for instructions to set programming environment – Refer to OpenCV China for instructions to process images : –