High-quality region-based foreground segmentation using a spatial grid of SVM classifiers.
Xiaohan ZhangCarlos R. del-BlancoCarlos CuevasFernando JaureguizarNarciso GarcíaPublished in: ICCE (2014)
Keyphrases
- svm classifier
- foreground segmentation
- high quality
- support vector machine
- support vector
- support vector machine svm
- image classification
- foreground objects
- background subtraction
- kernel function
- edge information
- feature vectors
- dynamic scenes
- spatio temporal
- image segmentation
- background modeling
- behavior analysis
- spatial information
- space time
- background model
- subspace learning
- temporal coherence
- shadow detection
- pairwise
- training set
- training data
- pixel wise
- semi supervised
- pattern recognition
- learning algorithm
- image quality