CVFX Lecture 10: Feature descriptors

ECSE-6969 Computer Vision for Visual Effects
Rich Radke, Rensselaer Polytechnic Institute
Lecture 10: Feature descriptors (2/24/14)
0:00:01 Feature descriptors
0:03:37 Why not use blocks around the detected feature?
0:07:26 Dominant gradient orientation
0:17:17 Normalizing block size and intensity
0:20:02 Comparing descriptors
0:23:52 Nearest neighbor distance ratio
0:27:07 The SIFT descriptor
0:39:25 Other descriptors
0:42:40 SURF features
0:45:37 Rotation and affine invariance
0:56:03 FAST corners
0:58:45 MSERs (Maximally stable extremal regions)
Follows Section 4.2 of the textbook. cvfxbook.com
Key references:
D. G. Lowe. Distinctive image features from scale-invariant keypoints. International Journal of Computer Vision, 60(2):91--110, Nov. 2004.
dx.doi.org/10.1...
K. Mikolajczyk and C. Schmid. Scale and affine invariant interest point detectors. International Journal of Computer Vision, 60(1):63--86, Oct. 2004.
dx.doi.org/10.1...
H. Bay, T. Tuytelaars, and L. Van Gool. SURF: Speeded up robust features. In European Conference on Computer Vision (ECCV), 2006.
dx.doi.org/10.1...
E. Rosten and T. Drummond. Machine learning for high-speed corner detection. In European Conference on Computer Vision (ECCV), 2006.
dx.doi.org/10.1...
J. Matas, O. Chum, M. Urban, and T. Pajdla. Robust wide-baseline stereo from maximally stable extremal regions. Image and Vision Computing, 22(10):761--7, 2004.
dx.doi.org/10.1...