Abstract

Sports events videos-based semantic content analysis and hidden markov model application

Author(s): Yongbing Chen, Geng Du

With sports video relative data rapidly development, howto find out required parts of information is focus of current research, the paper proposed sports video semantic analysis method after analyzing hidden Markov model, and proposes Out _ of _ play and In _ play contents analysis method flow, only selects partial segments to test and it gets that themodel accuracy has arrived at around 90%, besides it also explains the model can be convenient to lots of information statistics and train as well as other advantages, so the model has universality