A Principal Component Clustering Approach to Object-Oriented Motion Segmentation and Estimation.
Yun-Ting LinYen-Kuang ChenSun-Yuan KungPublished in: J. VLSI Signal Process. (1997)
Keyphrases
- motion segmentation
- principal components
- spectral clustering
- object oriented
- motion estimation
- model selection
- image sequences
- motion model
- principal component analysis
- clustering method
- dynamic scenes
- moving objects
- optical flow
- clustering algorithm
- multibody
- k means
- data clustering
- flow field
- epipolar geometry estimation
- missing data
- multiple moving objects
- dimensionality reduction
- unsupervised learning
- kernel space
- feature tracking
- cluster analysis
- optical flow estimation
- motion field
- principal component regression
- data points
- high dimensional data
- density estimation
- categorical data
- similarity measure
- motion compensation
- training data
- covariance matrix
- pattern recognition
- feature vectors