Copyright © 2026 Authors retain the copyright of this article. This article is an open access article distributed under the Creative Commons Attribution License which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

@article{184285,
        author = {Arpitha B V and Dr SUNITHA G P and Mr SANTHOSH S G},
        title = {Real-Time Text Extraction from Advertisement Board Images Using Image Processing Approaches},
        journal = {International Journal of Innovative Research in Technology},
        year = {2025},
        volume = {12},
        number = {4},
        pages = {826-834},
        issn = {2349-6002},
        url = {https://ijirt.org/article?manuscript=184285},
        abstract = {This paper presents a real-time OCR system designed to extract English text from advertisement board images containing a variety of fonts, layouts, and backgrounds. The system leverages two open- source OCR engines PaddleOCR and EasyOCR to directly recognize text from full images without requiring separate text detection stages. Implemented as a Streamlit-based web application, the system enables users to upload images, adjust confidence thresholds, and view side-by-side recognition results. Experimental evaluation shows that PaddleOCR performs well on clean, structured text, while EasyOCR offers flexibility in handling moderately stylized and curved fonts. The proposed approach simplifies the pipeline and provides an efficient solution for extracting text from real-world advertising scenarios. The system is particularly designed to handle the challenges posed by stylized fonts, varied text alignments, and visually complex backgrounds found in real-world advertising scenarios. By bypassing traditional multi-stage detection-recognition pipelines and leveraging pre- trained deep learning models, the approach achieves both speed and simplicity. The side-by-side comparison of OCR results also enables researchers and developers to identify model-specific strengths and weaknesses, facilitating better decision-making for deployment in practical environments. Overall, the solution offers a robust framework for rapid and accurate extraction of scene text from diverse advertisement board images without requiring manual region marking or extensive preprocessing.},
        keywords = {Advertisement boards, EasyOCR, PaddleOCR, Scene text recognition, Streamlit OCR app.},
        month = {September},
        }