虚拟现实的声场合成信息采集建模仿真分析

Virtual Reality Sound Field Synthesis Information Acquisition Modeling Simulation Analysis

目前声场合成信息采集方法存在采集结果与实际结果拟合程度低、采集耗时长的问题,提出基于MEMS和HRIR的虚拟现实声场合成信息采集建模方法。通过麦克风阵列体系得到声场原始信号,并利用欧拉角多数据融合获取人头部姿态信息。根据所得信息计算HRTF函数值,对函数值进行傅里叶变换,得到新生成的声音信号数据,对新生成的信号进行傅里叶逆变换,获得虚拟现实声场合成信息采集结果。利用最小方差模态滤波器能够高效消除期望模态之外模态的能力,结合声场各阶简正波加权系数,获取声场合成信息采集结果滤波后向量值。仿真结果表明,基于MEMS和HRIR的虚拟现实声场合成信息采集建模方法所得结果与实际结果拟合程度高,且信息采集耗时短。上述方法整体运行效率高,声场合成信息采集性能高。

This paper put forward a method for modeling and collecting virtual reality sound field synthesis information based on MEMS and HRIR.First of all,we obtained the original signal of sound field through microphone array system,and then used the multiple data fusion of Eulerian angle to obtain information of human head posture.Secondly,we calculated HRTF function value based on the obtained information and performed Fourier transform on the function value to obtain the new sound signal data.Moreover,we performed the inverse Fourier transform on the new signal to obtain the collection result of virtual reality sound field synthesis information.Through the minimum variance modal filter,we could efficiently eliminate the ability of modality except the excepted modality.Combined with the normal mode weighting coefficient of each order of sound field,we got the vector value after filtering the collection result of sound field synthesis information.Simulation results show that the result obtained by a method for modeling and collecting virtual reality sound field synthesis information based on MEMS and HRIR has high fitting degree with the actual result.The information acquisition usually takes a short time.The overall running efficiency of proposed method is high.Meanwhile,the performance of sound field synthesis information collection is high.