Segmentation stage of a PHMM-based model for off-line recognition of Arabic handwritten city names

Off-line recognition of handwritten words is a difficult task due to the high variability and uncertainty of human writing. The majority of the recent systems have some constraints such as the limitation of the size of the lexicon to deal with. However, the recognition of city names involves the use of a large vocabulary. Therefore, a segmentation stage is necessary to reduce the complexity of the problem. We present the segmentation stage relative to a Planar HMM-based model (PHMM) for off-line recognition of Arabic cursive Tunisian city names. We discuss the different segmentation steps and the variety of problems encountered when performing them. We especially focus on the median zone whose model depends highly on natural and vertical segmentation results.