Font Size: a A A

Study On Acoustic Source Localization And Tracking Based On Microphone Array

Posted on:2015-01-30Degree:MasterType:Thesis
Country:ChinaCandidate:S LiuFull Text:PDF
GTID:2298330467974616Subject:Signal and Information Processing
Abstract/Summary:PDF Full Text Request
Acoustic source localization and tracking is an important subject in human-computerinteraction. It is widely used in digital hearing aids, multimedia system, video conference systemand intelligent robot system. For example, in video conference system, it can get real-timedirectional information to control the camera moving with the speaker. It can automatically detectand point to the speaker in digital hearing aids, so that the voice of speaker can be directionalamplification and shielding the independent noise. The robot navigation also based on the locationinformation of acoustic source. The method of acoustic source localization and tracking based onmicrophone arrays has the ability of all-direction locating and tracking, and has much lowercomputational complexity. So, it has become the most important solution to the acoustic sourcelocalization and tracking problem.In real room environment, there are always exist noise and reverberation which can influencethe ability of the acoustic source localization and tracking system. First, the theory of sourcelocalization based on microphone is studied in this paper. The background knowledge of arraysignal processing and microphone array audio signal processing model is also described. Thentraditional localization methods are discussed and an improved acoustic source localization methodbased on SRP-PHAT is proposed. Simulation results show that the improved method can achievealmost same localization performance whereas the computation load is much lower compared withtraditional SRP-PHAT method. The basic theory of particle filter is described and the standardparticle filter is studied at the same time. Then genetic particle filter is described and a generalframework for tracking acoustic source using particle filters is formulated. An improved methodbased on genetic particle filter is proposed under this framework. Experiment results clearly showthat the improved method can effectively improve the tracking ability of the acoustic sourcetracking system.
Keywords/Search Tags:acoustic source localization and tracking, microphone array, genetic particle filter
PDF Full Text Request
Related items