A MORPHOLOGICAL APPROACH TO TEXT STRING EXTRACTION FROM REGULAR PERIODIC OVERLAPPING TEXT/BACKGROUND IMAGES

Abstract

The extraction of text strings from images, particularly those with regular periodic
overlapping text/background patterns, presents a significant challenge in the field of
computer vision and document analysis. In this project, we propose a morphological
approach aimed at accurately extracting text strings from such complex images. Our
method leverages morphological operations, including dilation, erosion, and
connected component analysis, to detect and segment text regions from the
background. By exploiting the periodic nature of the text/background patterns, our
approach effectively distinguishes between text and non-text regions, overcoming
issues commonly encountered in conventional text extraction methods. Furthermore,
we employ techniques for handling text overlaps and enhancing text string
connectivity to ensure comprehensive extraction results. Experimental evaluations
conducted on a diverse set of images demonstrate the effectiveness and robustness of
our proposed method in accurately extracting text strings from regular periodic
overlapping text/background images. The outcomes of this project hold promising
implications for various applications in document processing, optical character
recognition (OCR), and information retrieval systems.