Answers are grounded in this paper and cite their sources.
In Odia script, certain matras that cover the alphabet from the top pose a hindrance in character segmentation. This paper focuses on separating those matras that cover the Odia alphabet from the top. For this, an approach has been followed that investigates different partitions in an Odia alphabet. Two types of partitions are found in an Odia alphabet, one is open and another is closed. The approach explained in this paper searches for open partitions from the top side of the alphabet. Based on the presence of open partitions, the system decides whether a matra is covered with a matra at the top or not. The system has been tested with the data set which has been created using the Odia typing software 'Akruti'. Different fonts with different font sizes have been used to create the data set. The data set consists of several alphabets with those matras that cover the alphabet. The correctness accuracy is calculated by dividing the total number of cases that correctly segment the matras from the alphabet by the total number of inputs provided to the system multiplied by hundred. The system has achieved 99.05% accuracy.
No takes yet. Share an insight, caveat, or question.
Aradhana Kar (2024) studied this question.
Synapse has enriched 8 closely related papers on similar clinical questions. Consider them for comparative context: