Scale-space approach for the comparison of HK and SC curvature descriptions as applied to object recognition.
Erdem AkagündüzOmer EskizaraIlkay UlusoyPublished in: ICIP (2009)
Keyphrases
- scale space
- multiscale
- feature detection
- critical points
- object recognition
- image features
- scale spaces
- scale invariant
- planar curves
- gaussian scale space
- methods in computer vision
- image structure
- multiple scales
- scale selection
- scale space analysis
- multiscale representation
- zero crossing
- scale parameter
- keypoints
- curvature flow
- saddle points
- gaussian filter
- computer vision
- deep structure
- linear scale space
- image segmentation
- singular points
- gaussian kernel
- vector valued
- diffusion equation
- natural images
- object detection
- image analysis
- gaussian derivatives
- curvature scale space
- rotationally invariant
- d objects
- high level
- automatic scale selection