Abstrato

EVENT DETECTION IN CRICKET VIDEO BASED ON VISUAL AND ACOUSTIC FEATURES

Vijayakumar.V , Nedunchezhian

As digital video becomes more pervasive, efficient way of extracting the information from the video becomes an increasingly important. Video itself contains huge amount of data and complexity that makes the analysis very difficult. In this paper, we presented a video and audio features based event detection approach shown to be effective when applied to the cricket sports video. The advantage of this approach is the ability to recognize events that indicate high level of audio response and players crowd which can be correlated to key events. The proposed event detection has two steps. First, audio and visual features are extracted. Next, by defining a set of heuristic rules, the important semantic events such as wicket fall, score events are detected. The proposed framework has been tested with cricket video downloaded from the internet broadcast videos.