International, National and Local Languages OCR Segmentation of Running Hand Scripts
B.Hari Kumar1, P.Chitra2
1B.Hari Kumar*, professor (ECE) Wellfare Institute of science Technology & Management Pinagadi Village, Near Pendurthi, Andhra Pradesh
2P.Chitra, Professor Department of ECE, Sathyabama Institute of Science and Technology
Manuscript received on February 01, 2020. | Revised Manuscript received on February 05, 2020. | Manuscript published on February 30, 2020. | PP: 169-173 | Volume-9 Issue-3, February, 2020. | Retrieval Number: C5028029320/2020©BEIESP | DOI: 10.35940/ijeat.C5028.029320
Open Access | Ethics and Policies | Cite | Mendeley
© The Authors. Blue Eyes Intelligence Engineering and Sciences Publication (BEIESP). This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/)
Abstract: Script Segmentation of running hand scripts is a complicated job because of different hand writing styles and complex formational quality. Segmentation of moving hand script in Indian language is a difficult assignment .The incident of a title, crossway nature in the mid section & half nature makes the segmentation process is more difficult. Sometimes, the import space and noises make line shatter a hard task. Without disconnecting the nudge characters, it will be complicated to recognize the character; hence shatter is needed for the moving texts in a word. So, the technique, according to that first step will be pre-processing of a term, then we can identify the joint points, form the bounding boxes around all perpendicular & parallel lines of the script , finally splint the nudge nature based on their height and width. For non-touching printed scripts and running hand scripts, it gives us 95% and 90%result respectively. This technique fails for some hand written characters because of obligation that means cause will not match always to a lot of variation in the different handwriting scripts i.e., character like occurs problem as space in the middle of parts in scripts that can be resolve by identification system because of this system considers initial part as non-bar character and cut after that, as the future exertion is a concern, in this go these different characters as input to the character identification system. In this method have resolute control of a word which will once more helpful for the identification system to perceive bottom modifiers and higher modifiers. This running hand scripts technique can be applicable to other Indic languages hand written scripts.
Keywords: OCR, Preprocessing, Segmentation, Mat lab code, Image processing.