DP-PMK: an improved pyramid matching kernel for approximating correspondences in high dimensions.
Jun ZhangGuangzhou ZhaoHong GuPublished in: Neural Comput. Appl. (2012)
Keyphrases
- high dimensions
- feature matching
- high dimensional data
- correspondence matching
- high dimensional
- dynamic programming
- spatial pyramid matching
- matching algorithm
- partial matching
- multiresolution
- image matching
- high dimensional spaces
- transformation parameters
- feature space
- matching cost
- data points
- matching score
- feature points
- training set
- multiscale
- keypoints
- poor quality
- image representation
- visual words
- similarity search
- database
- scale space
- medical images
- image registration
- image features
- feature extraction
- image sequences
- learning algorithm
- data sets