A parameterless scale-space approach to find meaningful modes in histograms - Application to image and spectrum segmentation.
Jérôme GillesKathryn HealPublished in: CoRR (2014)
Keyphrases
- scale space
- multiscale
- gaussian scale space
- multiple scales
- scale spaces
- image structure
- image analysis
- scale selection
- image features
- multiscale segmentation
- gaussian derivatives
- image segmentation
- multiscale representation
- gaussian kernel
- vector valued
- edge detection
- linear scale space
- gaussian smoothing
- scale space analysis
- segmentation method
- zero crossing
- singular points
- automatic scale selection
- feature detection
- keypoints
- segmentation algorithm
- critical points
- grey level
- gaussian filter
- scale parameter
- image processing
- primal sketch
- input image
- morphological scale space
- planar curves
- energy functional
- edge detector
- pattern recognition
- gray value
- pixel wise
- level set
- similarity measure
- computer vision
- diffusion equation
- scale invariant
- active contour model
- feature vectors