machine learning - Searching an Image Database Using SIFT -


several questions have been asked sift algorithm, seem focussed on simple comparison between 2 images. instead of determining how similar 2 images are, practical use sift find closest matching image out of collection of thousands of images? in other words, sift scalable?

for example, practical use sift generate keypoints batch of images, store keypoints in database, , find ones have shortest euclidean distance keypoints generated "query" image?

when calculating euclidean distance, ignore x, y, scale, , orientation parts of keypoints, , @ descriptor?

there several approaches.

one popular approach called bag of words representation matching based solely upon how many descriptors match, ignoring location part consisting of (x, y, scale, , orientation) , @ descriptor.

efficient querying of large database may use approximate methods locality sensitive hashing

other methods may involve vocabulary trees or other data structures.

for efficient method takes account location information, check out pyramid match kernels


Comments

Popular posts from this blog

c# - How to set Z index when using WPF DrawingContext? -

razor - Is this a bug in WebMatrix PageData? -

visual c++ - Using relative values in array sorting ( asm ) -