Text string extraction from periodic overlaying text and background images using a Morphological Approach

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.