Exploiting local dependencies with spatial-scale space (S-Cube) for near-duplicate retrieval.
Xiangang ChengYiqun HuLiang-Tien ChiaPublished in: Comput. Vis. Image Underst. (2011)
Keyphrases
- scale space
- multiscale
- rotationally invariant
- scale spaces
- feature detection
- gaussian scale space
- image structure
- scale selection
- gaussian kernel
- spatial information
- image features
- methods in computer vision
- image analysis
- multiple scales
- scale invariant
- linear scale space
- gaussian derivatives
- scale space analysis
- critical points
- keypoints
- deep structure
- vector valued
- saddle points
- multiscale representation
- singular points
- diffusion equation
- scale parameter
- planar curves
- primal sketch
- neural network
- shape retrieval
- zero crossing
- multiscale segmentation
- nonlinear diffusion
- spatial structure
- image retrieval