A Fast Vehicle Horn Sound Location Method with Improved SRP-PHAT

Traffic noises, especially the vehicle horn sounds, affect the normal life of people and researchers have done many researches on the recognition and location of the traffic noises. This paper proposes a vehicle horn sound location method with an improved SRP-PHAT. Through analyzing the sounds, the frequency domain features for recognizing the vehicle horn sound are selected and the corresponding sound recognition method based on time-frequency transformation (VHSR-TFT) is presented. Complying with the large time cost of SRP-PHAT method for sounds location, the stochastic region contraction method is utilized to improve the real-time performance. The results of simulations and experiments demonstrate that the proposed method has better properties in accuracy and efficiency than the competing methods.