A Novel Algorithm for Moving Objects Recognition Based on Sparse Bayesian Classification

This paper deals with the problem of moving objects recognition using Bayesian method. A novel algorithm based on sparse Bayesian classification for the recognition of moving objects is proposed. This approach takes full advantage of sparse Bayesian in solving classification problems. It uses fewer kernel functions in order to reduce the complexity of the computation and resolves the over-fitting problem in recognition system. The experimental results show that this approach distinctly outperforms other classification approaches on this issue. The veracity and velocity are also satisfactory.