Simulation Techniques Towards Recognizing Tamil Characters Inscribed in Antiquated Copper Plates Application of Complex Deep Learning Machine on Eco Friendly Cleansed Plates

Abstract

Refurbishing schemes towards restoring and protecting classical metal-plates with inlaid inscriptions and/or carvings reflect artistic impressions and useful information of our history. They enable us to comprehend individual traits, regarding various phases of time possessing different propensities and customs. Every antiquated composition is available in different forms of structures like stones, carvings in metal-plates, inscriptions written in stacks of palm-leaf and original copies of documented papers. As such, hundreds and thousands of authenticated copper-specific ancient landmarks, significant data relating the history and culture of that era. However, these copper-based antiques have been gradually deteriorating and crumbling as a result of ecological contaminations and other organic and related anthropogenic exercises. Therefore, there is an imminent need to preserve these copper landmarks and to re-establish (refurbish) them for our future use and references. Hence, Copper Plate Optical Character Recognition (CPOCR) for the included contents is an open territory for research and testing. Relevantly, this thesis explores the underlying, traditional aspects of data acquisition in copper plates and identifies customary interpretation techniques. With regard to such efforts, an interactive tool is proposed here for epigraphist compatibility towards rescuing valuable inscriptions of the past so as to identify the tarnished characters inscribed in copper-plates. The proposed technique targets retrieving valuable information from copperplates in an efficient and reliable method with minimal processing time. Objectively, the proposed scheme in this thesis would help viably identifying and transcribing the smudged and/or problematic inscriptions/carvings seen in copper- plates of Tamil writings/images; and, relevant restorations can be undertaken thereof in character recognition that forms the fundamental aspect of Tamil text image- processing strategies. Relevant problems in character recognition is generally