Outdoor Impulse Sound Event Detection Method Based On Microphone Array

Some dangerous events happened as the impulsive sound events occur.The effective detection of impulsive sound events can help to find and analyze the dangerous situation in time and thus plays an important role in preventing and combating violent terrorist crimes.However,the time-frequency characteristics of the signal are relatively complex.It is an urgent problem to be solved.On the other hand,public places such as squares and green spaces contain a large amount of environmental noise,which often directly submerges the impulsive sound at low signal-to-interference-plus-noise ratios.It is necessary to further research the interference suppression method.Based on the cooperation project with enterprises,this paper focuses on the recognition and detection of impulsive sound in typical outdoor areas like squares,green spaces and other public places.In view of the above situation,focused on the muzzle blast,and taken other meaningful impulsive sounds like firecrackers and glass breaking into account.Rely on the idea of detection-by-recognition method,a multi-scale subband energy set is proposed.The feature set is used as the feature vector for support vector machine and the experimental results with high signal-to-interference-plus-noise ratios recordings of gunshot demonstrate that both the precision and recall ratios are larger than 93%,and F-score is better than 0.95.Due to the outdoor environment noise is inevitable,it further analyzes the influence of environmental noise.The impulsive sound is disturbed by background environments according to different signal-to-interference-plus-noise ratios to simulate the public environment.It is found that the detection rate gets lower when the signal-to-noise ratio decreases.When the signal-to-interference-plus-noise ratio is less than 10 dB,the missed detection rate is up to 70%.To solve the problem,this paper further used microphone array to effectively enhance the outdoor impulsive sound based on adaptive source direction estimation and generalized sidelobe canceller.The experiment shows that the adaptive interference suppression method can effectively suppress the background noise.Finally,the detection-by-recognition technology is used to detect the outdoor impulsive sound from the continuous monitoring recordings after noise suppression.The experimental results show that the impulsive sound signal can be effectively detected by the proposed method.Although there are still a few false alarms,the detection rate is much higher than the result before using adaptive interference suppression,and missing alarms situation no longer exists,which provides a good foundation for the acoustic monitoring system in the field of outdoor public safety.