Automated Text line Segmentation and Table detection for Pre-Printed Document Image Analysis Systems

摘要

Document Image Processing is considered one of the widely used technologies today. It has proved to be extremely useful in digitizing physical documents to preserve them for longer times. The primary challenge in this work is to extract the overlapping lines from handwritten document images and table detection from pre-printed documents. In this research paper, we present a line segmentation and table detection algorithm suitable for pre-printed and handwritten documents. Initially, the input document image is preprocessed and a baseline is created for offline every peak to carry out line segmentation. Further, table detection is accomplished with the use of edge detection techniques and morphological operations.