机器人听觉定位跟踪声源的研究与进展  

Research and advance for robot audition to localize and track sound sources

作　　者：李从清[1,2] 孙立新[1] 戴士杰[1,3] 李洙梁[1] 

机构地区：[1]河北工业大学机械工程学院,天津300130 [2]天津城市建设学院能源与机械工程系,天津300384 [3]哈尔滨工业大学机器人技术与系统国家重点实验室,黑龙江哈尔滨150080

出　　处：《燕山大学学报》2009年第3期199-205,共7页Journal of Yanshan University

基　　金：河北省自然科学基金资助项目(F2007000118);机器人技术与系统国家重点实验室开放基金研究资助项目(SKLRS200716);河北省教育厅基金资助项目(2007315)

摘　　要：从基于麦克风阵列和基于人耳听觉机理两个方面综述了当前机器人听觉定位跟踪声源目标的研究动态和发展方向。首先回顾了机器人听觉定位声源的研究历史;其次讨论了两种系统的优缺点;最后指出了未来机器人听觉定位跟踪声源的发展趋势。基于麦克风阵列的声源定位系统,盲波束形成技术是未来进一步研究内容。而基于人耳听觉机理的声源定位系统,运用计算听觉场景分析建立声源定位模型将是未来热点研究内容。The trends of research and the direction of development for robot audition to localize and track sound sources are sum- marized from two aspects based on the microphone arrays and the human auditory mechanism. Firstly, the history for robot audition to localize and track sound sources is reviewed. Secondly, the advantages and disadavantages of these systems are discussed. Finally, the development for robot audition to localize and track sound sources is pointed out. For sound localization system based on the microphone arrays, blind beamforming technique will be studied further in the future and for sound localization system based on the human auditory mechanism, it will be hot issue in the future to establish the modle of sound localization by applying computa- tional auditory scene analysis.

关 键 词：机器人听觉 声源定位 麦克风阵列 计算听觉场景分析 

分 类 号：TP242.6[自动化与计算机技术—检测技术与自动化装置] TN912.3[自动化与计算机技术—控制科学与工程]