Video Deblurring via Semantic Segmentation and Pixel-Wise Non-Linear Kernel.
Wenqi RenJin-shan PanXiaochun CaoMing-Hsuan YangPublished in: CoRR (2017)
Keyphrases
- pixel wise
- pixel level
- superpixels
- video sequences
- video data
- spatial information
- foreground objects
- background subtraction
- denoising
- video surveillance
- appearance model
- video frames
- computer vision
- support vector
- conditional random fields
- segmentation algorithm
- natural images
- feature space
- moving camera
- object detection
- information extraction
- feature vectors