Skeleton Matching based approach for Text Localization in Scene Images

TL;DR

This paper introduces a skeleton matching approach for scene text localization, utilizing morphological skeletonization, template comparison, and geometrical rules to accurately detect text in diverse scene images.

Contribution

It presents a novel skeleton-based method combining morphological techniques and template matching for effective scene text localization.

Findings

Effective detection of texts of various sizes, fonts, and colors

High accuracy demonstrated on standard datasets

Robustness to different scene complexities

Abstract

In this paper, we propose a skeleton matching based approach which aids in text localization in scene images. The input image is preprocessed and segmented into blocks using connected component analysis. We obtain the skeleton of the segmented block using morphology based approach. The skeletonized images are compared with the trained templates in the database to categorize into text and non-text blocks. Further, the newly designed geometrical rules and morphological operations are employed on the detected text blocks for scene text localization. The experimental results obtained on publicly available standard datasets illustrate that the proposed method can detect and localize the texts of various sizes, fonts and colors.