Multi-script text versus non-text classification of regions in scene images.
Bowornrat SrimanLambert SchomakerPublished in: J. Vis. Commun. Image Represent. (2019)
Keyphrases
- scene images
- text classification
- text regions
- text extraction
- text documents
- complex background
- text mining
- text segmentation
- text information
- document images
- video frames
- input image
- natural scenes
- connected components
- document classification
- text recognition
- optical character recognition
- text categorization
- outdoor scenes
- bag of words
- infrared images
- video images
- feature selection
- knn
- text processing
- bounding box
- multi label
- n gram
- term frequency
- edge map
- semantic features
- information retrieval
- face detection
- text lines
- natural language processing
- machine learning
- image structure
- lighting conditions