A framework of human detection and action recognition based on uniform segmentation and combination of Euclidean distance and joint entropy-based features selection.
Muhammad SharifMuhammad Attique KhanTallha AkramMuhammad Younus JavedTanzila SabaAmjad RehmanPublished in: EURASIP J. Image Video Process. (2017)
Keyphrases
- human detection
- action recognition
- euclidean distance
- histograms of oriented gradients
- hog features
- feature vectors
- activity recognition
- histogram of oriented gradients
- similarity measure
- human actions
- background subtraction
- computer vision
- object detection
- surveillance system
- body parts
- distance measure
- data points
- image segmentation
- feature extraction
- pedestrian detection
- human pose
- co occurrence
- face detection
- image features
- feature space
- three dimensional
- video surveillance
- machine learning
- multiscale
- probabilistic model
- dimensionality reduction