A novel IPTV framework for automatic TV commercials detection, labeling, recognition and replacement

• Published in: Multimedia tools and applications 2023-03, Vol.82 (6), p.8561-8579
• Main Authors: Arsan, Taner, Bulut, Enes Emre, Eren, Berk, Uzgor, Ahmet, Yolcu, Selcuk
• Format: Article
• Language: English
• Subjects: 1207: Innovations in Multimedia Information Processing & Retrieval; Algorithms; Commercials; Computer Communication Networks; Computer Science; Customers; Data Structures and Information Theory; Multimedia Information Systems; Real time; Recognition; Special Purpose and Application-Based Systems
• ISSN: 1380-7501; 1573-7721
• DOI: 10.1007/s11042-021-11563-y

Advertisements are one of the most important way for companies to access their customers. In this context, televison commercials are gaining significant importance in many sectors daily, and it is crucial for companies to promote their products in the best way. This creates a big rivalry between companies. From this point of view, we have created an IPTV Framework that can automatically detect commercials of rival companies and replace them with desired commercials for companies to help them highlight their products to their customers. We have benefited from monochrome frames to detect the Livestream commercial block and proposed a fingerprint algorithm to create an automatic commercial database. We can easily recognize the commercials, and we can mask the commercials of rival companies with these techniques. We have tested our algorithm in real-time by streaming a recorded broadcast from a server of a specific TV channel. Experimental results show that our algorithm provides high accuracy in real-time commercial recognition.