Article Elements Recognition of Old Sinhala Newspapers Using Vision Heuristic Algorithm

This paper presents an approach to segmenting degraded and low-quality old Sinhala newspaper articles into separate elements such as Title, Subtitle, Columns, Highlight phrases, special paragraphs, author name or reference phrase, and general content. Human vision inculcated machine vision algorithms were introduced to elicit above categorical content with groundbreaking accuracy. This paper highlights the importance of column-wise analysis of newspaper articles to segment elements. The results of this research are intriguingly comparable to other existing literature. The outcome application of this study is applicable in old Sinhala newspaper digital archival to be enriched with search and retrieval.