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{205415,
        author = {Mohammed Fahaduddin and Ishrat Tamreen and Sobiya Sultana and Sabhah Fatima and Abdur Rahman Fasi},
        title = {DeepVisualPhish: A Hybrid Visual Similarity and Screenshot Intelligence Framework for Detecting Phishing Websites},
        journal = {International Journal of Innovative Research in Technology},
        year = {2026},
        volume = {13},
        number = {1},
        pages = {6706-6711},
        issn = {2349-6002},
        url = {https://ijirt.org/article?manuscript=205415},
        abstract = {Phishing attacks have become increasingly advanced due to the ability of attackers to create webpages that closely resemble trusted websites. Conventional phishing detection mechanisms mainly depend on URL inspection, blacklist databases, and webpage source-code analysis. While these approaches are useful for identifying previously reported threats, they often struggle to detect newly generated phishing websites that imitate the visual appearance of authentic platforms. Cybercriminals now focus heavily on duplicating webpage layouts, logos, login portals, and branding elements to manipulate users into disclosing confidential information. This research introduces DeepVisualPhish, a hybrid phishing detection framework that emphasizes webpage screenshot intelligence and visual similarity evaluation. The proposed system examines screenshots of webpages and combines multiple analysis techniques including Convolutional Neural Networks (CNN), Optical Character Recognition (OCR), logo matching, structural similarity analysis, and layout inspection. Instead of relying entirely on textual or domain-based indicators, the framework evaluates the overall visual identity of a webpage. The model compares suspicious webpages against a repository of legitimate website templates and analyzes similarities in branding, login forms, navigation structure, and graphical content. Additional behavioral indicators such as deceptive overlays and replicated authentication interfaces are also incorporated into the detection process. Experimental evaluation indicates that the proposed framework performs effectively against visually deceptive and zero-day phishing attacks that commonly bypass traditional security filters. The study demonstrates that integrating visual intelligence with machine learning significantly improves phishing detection accuracy and reduces dependency on blacklist-based approaches. The framework can support future browser security systems, automated cyber defense tools, and intelligent phishing prevention platforms.},
        keywords = {Phishing Website Detection, Screenshot Analysis, Visual Intelligence, Deep Learning, OCR, CNN, Structural Similarity, Cyber Security.},
        month = {June},
        }