Olex-Genetic Algorithm Based Information Retrieval Model from Historical Document Images

Abstract

Presently, the process of retrieving the historical documents is treated as an important challenge due to the fact that the document possessesindividual structure and level of deprivation. The textual characters present in the printouts come together with the typographical objects. The retrieval and perusal of the visual typographical objects indicates that the content of heritage documents helps to effectively interpret the documents. The extraction of the visual objects finds useful in the interpretation and conveyance of more details regarding diverse practices of demonstration in past documents and the impact in the present status of publication. A pair of essential typographical objects linked to the history of knowledge and information is footnotes and tables. In this paper, the main intention is the detection of the existence of the visual elements from historical printable documents. A new Olex-GA based footnote recognition model (OFR) is developed. The footnote detection model make use of a collection of layout-based models for the extraction of few features concerning to the font and appearance. In addition, Olex-GA algorithm is for the classification purposes. This model is validated using a massive set of 18thcentury printed documents with higher than 32 million images, and the outcome showed their effective outcome of the presented model.