Detection of Text Lines of Handwritten Arabic Manuscripts using Markov Decision Processes.
Youssef BoulidAbdelghani SouharMohamed El Youssfi El KettaniPublished in: Int. J. Interact. Multim. Artif. Intell. (2016)
Keyphrases
- markov decision processes
- handwritten documents
- text lines
- text line segmentation
- document images
- handwriting recognition
- word spotting
- finite state
- optical character recognition
- state space
- connected components
- historical documents
- reinforcement learning
- decision theoretic planning
- policy iteration
- dynamic programming
- optimal policy
- printed documents
- character recognition
- complex background
- planning under uncertainty
- text retrieval
- transition matrices
- hough transform
- reachability analysis
- document image analysis
- infinite horizon
- document analysis
- model based reinforcement learning
- structural features
- markov decision process
- partially observable
- action sets
- discounted reward
- learning algorithm
- writer identification
- action space
- temporal difference