| Island security is of great significance in national defense.Island security work is an important guarantee for national territorial sovereignty and the security of people’s property.The most important responsibilities are to intercept intrusions into national territory by illegal persons and illegal targets and to maintain stability in coastal and border areas.With the deepening of our country’s opening to the outside world,the proportion of people entering and leaving the country is rising,and border and sea defense island invasion,escape,smuggling,smuggling,drug trafficking,and other criminal acts are increasingly rampant.At present,the domestic response to such problems is more than the use of border and sea defense island fighters on duty,radar technology,and image video for security monitoring.However,these technologies have high environmental requirements,easy to produce misjudgment,concealment is not strong,and there are blind spots in security monitoring and other problems,can not meet the current requirements of the border and sea defense island security.In view of the shortcomings of the above methods,it is proposed to use acoustic means to make up for them,so as to develop a set of island security systems based on acoustic recognition technology,combined with video and radar technology to work in a coordinated manner,which can further improve the probability of detection of targets on the border and sea defense islands.The harsh and complex environment and climate of the border defense islands,make the target sound masked by noise,in order to be able to quickly and accurately identify the target sound to achieve island security monitoring,this paper proposes an adaptive noise reduction method based on multi-objective genetic algorithm optimization of weak target sound processing to improve the signal-to-noise ratio,adding a new idea to improve the signal-to-noise ratio of weak signals.In order to speed up data processing and further reduce the impact of noise on the signal,this paper adopts a double-threshold endpoint detection algorithm based on short-time energy and short-time over-zero rate,and extracts useful signal segments of the sound signal based on the thresholds set by both of them for target sound signal segmentation processing;for the problem of low recognition rate based on the traditional single Meier frequency cepstrum characteristic parameter(MFCC),this paper combines the MFCC with good stability and In this paper,the MFCC static features with good stability and anti-interference property and the first-order differential MFCC dynamic features are fused as the feature parameters in this paper,which effectively improves the recognition accuracy;the pattern recognition method is studied,and the DTW model is selected for recognition matching according to the characteristic that the feature vector length of the sound emitted by the same sound source at different times is different.In view of the problems of large computational volume and low real-time performance caused by the traditional DTW model for global matching,this paper improves the traditional DTW model through the core ideas of segmental matching and successive screening,and proposes an improved DTW model,which reduces the computational volume and shortens the recognition time,and improves the recognition accuracy of this system.On the basis of the theoretical research,the system hardware and software platform is completed according to the system requirements set in this paper.The MCU performs a series of operations on the collected data,such as noise reduction,multi-parameter extraction and recognition matching,to recognize the target sound signal and send the result to the Window terminal and display the recognition result to the user.The recognition result is sent to Window and displayed to the user.Finally,the whole system is simulated and tested,and the results show that the system can identify the required target sound signals to monitor the security of the island.The system can effectively make up for the shortcomings of radar technology and video images in low visibility for security monitoring and has good market prospects. |
