Research On 3D Space Fast Localization Of Moving Sound Source And Sound Source Intensity Monitoring

Acoustic beamforming technology is based on the inverse reconstruction process of acoustic radiation.It reconstructs the distribution of the sound field at a fixed depth from the array by delaying and summing the signals obtained by the microphone array.The location of the maximum response in the sound field is the sound source.The beamforming technology can reconstruct the transient and steady state sound field,and has a good ability to identify the middle and far field sound sources.The traditional beamforming reconstruction of the planar sound field requires scanning and reconstruction of a plane at a certain depth from the microphone array.If the threedimensional sound field in space is needed to be reconstructed,it also needs to be scanned in depth,which has a large amount of computation.This also leads to the inability to quickly and accurately obtain the depth of the moving sound source,and thus it is impossible to use the sound intensity scaling method to accurately calculate and dynamically monitor the sound power of the sound source.Binocular stereo vision and image processing technology is used to remove the invalid sound field reconstruction area in beamforming in this study,which greatly improves the reconstruction speed.In the beamforming and binocular stereo vision technology fusion method,the target object containing the sound source is captured by the binocular camera to obtain the depth image of the camera’s field of view the beamforming and binocular stereo vision technology fusion method.The threedimensional surface of the target object is obtained through image processing and used as the beamforming three-dimensional sound field reconstruction area.Finally,locating the 3D coordinates of the sound source through the largest response position in the 3D sound field.And combining the beamforming-based sound intensity scaling method to achieve sound power acquisition.The method of obtaining the three-dimensional surface of the object to be recognized from the camera depth image is studied in detail based on the above theory,and the spatial angle and corresponding relationship between the microphone array and the stereo camera are analyzed.The method of transforming the two spatial coordinate systems of acoustic array and stereo vision is solved,and the spatial transformation between the three-dimensional surface acquired by the camera and the reconstructed surface of the beam forming sound field is completed.The 3D sound field reconstruction method that combines acoustic beamforming and stereo vision has the ability to quickly reconstruct and dynamically obtain the 3D localization and sound power level of the sound source.The number of reconstructed focal points on the 3D surface of the measured target is less than the number of focal points in the traditional regular reconstruction surface.The principle of numerical analysis is used to compare the computational complexity of the fusion method with the conventional beamforming.A verification experiment was carried out in a hemianechoic room.The results show that the algorithm used in this paper can effectively reduce the amount of calculation,and can complete the 3D positioning and sound power monitoring of the moving sound source in the operating mechanism.At last,the LabVIEW development platform is used to design,build and test the3 D sound field reconstruction and sound source state monitoring system.The threedimensional sound field on the surface of the target object can be quickly reconstructed and visualized through the system.Then,the 3D positioning of the moving sound source and the sound power monitoring of the sound source are also achieved.