Text Line Segmentation of Handwritten Historical MODI Documents

Tremendous growth in digital technology introduced new magnitudes to the digital world also raised various challenges and opened new area for researchers. Text line segmentation of historical handwritten documents became mystery for many researchers. Degradation, presence of noise, fluctuation in illumination, multiple skews, overlapping lines, touching lines, curved lines, slanting lines, uneven alignments, additional modifiers, embellishments and presence of decorators motivated researchers to develop suitable segmentation algorithm for historical documents. MODI was an ancient script of India and vanished from usual transactions, available only in old historical documents with cursive and complex type of nature. Principal interest of this work was to address various hurdles occurred during text line segmentation and propose robust method for text line segmentation from historical handwritten MODI documents. According to best of our knowledge this was the first attempt taken towards segmentation and document analysis of historical handwritten MODI documents.