Automated Audio Monitoring Approach for Radio Broadcasting Channels in Sri Lanka

Most of the people like to listen music. When music repository is growing, it is very hard to find a music song without knowing the exact Meta data of it like title, of the song singer, musician and so on. As a solution to this problem, many researchers have tried to provide a way to identify music songs by analyzing the content of it. This is known as content based audio identification. In this research we extend this idea to monitor and identify broadcasted radio streams since there are so many applications and business values of it. Before analyzing the content of radio stream, we converted the audio stream into machine friendly format. Then we tried to remove irrelevant audio objects like commercials and talks during the preprocessing stage. After that using a powerful content base audio identification algorithm, we recognized songs objects and identified them. We used a very large audio repository and store songs as fingerprints. At the end of this process we provide a detailed report of song broadcasting history of a given radio channel. Even though there were many useful applications of this research we focus only on the protecting the copy right ownership of artists.