Robust and efficient recognition of low-quality images by cascaded recognizers with massive local features.
Koichi KiseKazuto NoguchiMasakazu IwamuraPublished in: ICCV Workshops (2009)
Keyphrases
- robust recognition
- feature extraction
- low quality images
- object class recognition
- character recognition
- feature vectors
- feature set
- computationally efficient
- low level
- co occurrence
- recognition rate
- highly discriminative
- object recognition
- video sequences
- feature space
- image features
- visual object recognition
- feature extractor
- discrimination power
- noisy environments
- handwritten digits
- word recognition
- handwritten characters
- human activities
- feature analysis
- low quality
- feature descriptors
- data sets
- image matching
- classification accuracy
- data analysis
- decision trees