https://github.com/alexkch/scalable-imagevocab-tree
Image matching on DVDs using scalable vocabulary trees
https://github.com/alexkch/scalable-imagevocab-tree
image-processing matlab sift vocabulary-tree
Last synced: 8 months ago
JSON representation
Image matching on DVDs using scalable vocabulary trees
- Host: GitHub
- URL: https://github.com/alexkch/scalable-imagevocab-tree
- Owner: alexkch
- Created: 2017-12-25T16:26:22.000Z (over 8 years ago)
- Default Branch: master
- Last Pushed: 2018-04-07T23:02:21.000Z (about 8 years ago)
- Last Synced: 2025-05-16T23:12:25.261Z (about 1 year ago)
- Topics: image-processing, matlab, sift, vocabulary-tree
- Language: Matlab
- Homepage:
- Size: 35.3 MB
- Stars: 1
- Watchers: 1
- Forks: 0
- Open Issues: 0
-
Metadata Files:
- Readme: README.md
Awesome Lists containing this project
README
# Scalable-ImageVocab-Tree
by: Alex Chang
Scalable Image Vocabulary Tree
Implementing [1] using Stanford Mobile Visual Search imageset [2] as test and training dataset
[1] implementation details:
(a) Keypoint detection
(b) Keypoint description with rotation and scale in-variance
(c) Building and using the vocabulary tree (i.e. quantization)
(d) Hierarchical scoring
(e) Retrieval
STEP 1: For a test image, retrieve top 10 matches (DVD cover images from the database) returned
via the implemented approach.
Then... using RANSAC:
STEP 2: Compute a homography with your implementation
from above for each retrieved DVD cover image and the test image.
STEP 3: Find the DVD cover image from (4) with the highest number of inliers. Plot the test
image with the localized DVD cover as well as the best retrieved DVD cover
Reference:
[1] Nister, Stewenius, Scalable Recognition with a Vocabulary Tree, CVPR 2006,
http://www-inst.eecs.berkeley.edu/~cs294-6/fa06/papers/nister_stewenius_cvpr2006.pdf
[2] https://sites.google.com/site/chenmodavid/datasets