Efficient keypoint detection and description using filter kernel decomposition in scale space.
Ryo OkutaniKenjiro SugimotoSei-ichiro KamataPublished in: ICIP (2016)
Keyphrases
- scale space
- gaussian kernel
- multiscale
- keypoint detection
- keypoints
- scale spaces
- scale parameter
- multiple scales
- gaussian filter
- image analysis
- scale invariant
- vector valued
- gaussian scale space
- zero crossing
- image structure
- feature detection
- scale space analysis
- image features
- critical points
- multiscale representation
- deep structure
- edge detector
- scale selection
- linear scale space
- planar curves
- high level
- diffusion equation
- gaussian derivatives
- scale space theory
- reproducing kernel hilbert space
- methods in computer vision
- automatic scale selection
- feature extraction
- median filter
- loss function
- kernel methods
- primal sketch
- riemannian manifolds
- saddle points
- curvature flow