CN205139359U - Based on indoor sound localization system of FPGA microphone array - Google Patents

Abstract

The utility model relates to a based on indoor sound localization system of FPGA microphone array is including microphone array structure module, speech signal preliminary treatment, the algorithm realization of delay location, FPGA control system, five parts of camera, six microphones are fixed and constitute toper microphone array in rectangular coordinate are, the voice signal is gathered to the microphone, through AD sampling and band -pass filter, generate six tunnel pending input speech signal, each coordinate epaxial a pair of microphone collected voice signal's time delay estimation is accomplished with broad sense cross -correlation algorithm, obtain 3 pairs of microphones the time delay the how much location algorithms by hexa -atomic microphone array, can confirm the azimuth and elevation tracking unit angle of pitch of sound source, FPGA is as the rotation of controller control camera, make its direction towards the spokesman.

Description

A kind of based on the indoor sonic location system of FPGA microphone array

Technical field

The utility model relates to a kind of sonic location system, is a kind of based on the indoor sonic location system of FPGA microphone array specifically.

Background technology

The aerial propagation of sound completes the transmission of information, achieves interpersonal basic communication.Microphone is one of the most basic voice signal receiving equipment, when carrying out video conference, after microphone obtains the voice signal of speaker, need to determine fast the position of spokesman and camera is turned to spokesman, traditional single-sensor affects by sound source position, if sound source position changes, need again to put microphone, the voice signal of better quality could be picked up.And in indoor environment, sound source may be subject to the pollution of coloured noise, correlated noise, reflection and reverberation, greatly reduces the quality of speech signal of single microphone, thus affects the effect of speech processes.

Therefore, design a kind of microphone array utilizing multi-microphone to form to carry out speech processes and be necessary.Microphone array shows very strong spatial selectivity, can realize the source of location and tracking sound automatically, therefore can be applied to speech enhan-cement, auditory localization and echo cancellor etc., have higher researching value.In the last few years, along with developing rapidly of microphone array technology, people propose higher expectation to auditory localization, and wish that its application is wider, have more hommization and easier, therefore actual demand facilitates the applied research of auditory localization greatly.

Utility model content

For solving the problem, the purpose of this utility model is to provide a kind of hexa-atomic microphone array system to realize indoor auditory localization.

For achieving the above object, the technical scheme that the utility model is taked is: a kind of based on the indoor sonic location system of FPGA microphone array, comprise microphone array construction module, speech signal pre-processing module, FPGA control system, camera four part, speech signal pre-processing module comprises A/D modular converter and sampling processing module, microphone array is connected with A/D modular converter, and FPGA control system is connected with camera.

Native system adopts top-down design, makes full use of the base resource that development environment provides, and plays the advantage of FPGA to greatest extent, shortens the design cycle.

Native system FPGA is as controller, and the characteristic utilizing point sound source to propagate with spherical wave form, gathers voice signal at diverse location, and by specific mathematical model, the also position of acoustic source, reliability is high, practical.

Described based in the indoor sonic location system of FPGA microphone array, the hexa-atomic plain microphone array of institute's taper that uses, does not limit by sound source position, has full spatial domain auditory localization ability, improve positioning precision simultaneously.

Accompanying drawing explanation

Fig. 1 is described in the utility model based on the indoor sonic location system structured flowchart of FPGA microphone array.

Fig. 2 is the hexa-atomic microphone array scheme of installation of system taper described in the utility model.

Embodiment

Refer to shown in Figure of description 1, the utility model is a kind of based on the indoor sonic location system of FPGA microphone array.It comprises microphone sensor array module (1), speech signal pre-processing module (2), FPGA control module (3) and camera (5) composition.

Wherein said microphone sensor array module (1) forms a space six array element microphone array by microphone (6) (7) (8) (9) (10) (11).Speech signal pre-processing module (2) comprises (12) (13) (14), carries out pre-service respectively to three pairs of microphone signals.

Accompanying drawing 2 is the hexa-atomic microphone array scheme of installation of taper, and six microphones are fixed in three-dimensional cartesian coordinate system.Sound source coordinate is T (x, y, z), and it is r to the distance of coordinate origin, and position angle is the angle of pitch is θ.Sound source is respectively r to the distance of each array element ₁, r ₂..., r ₆.The time interval that sound wave arrives array element 1 and 3 is t ₁₃, the time interval arriving 2 and 4 is t ₂₄, the time interval arriving 5 and 6 is t ₅₆.Can in the hope of position angle according to geometric approximation and pitching angle theta. $\mathrm{\θ}=arctan\sqrt{t_{13}^2+t_{24}^2}/t_{56}.$

Voice signal is band-limited signal, energy mainly concentrates between 300-3400Hz, the signal that microphone collects comprises interchannel noise and ground unrest, after A/D modular converter and sampling processing module, need to fall with the noise filtering of bandpass filter by low frequency and high band, desirable upper cut off frequency f _h=4000Hz, lower limiting frequency f _l=200Hz, sampling rate is f _s=20kHz.Obtain pending input speech signal x1 (n) and x2 (n).

Microphone sensor array module (1) is connected with pretreatment module (2), pretreatment module (2) is connected with FPGA control module (3), FPGA control module (3) is connected with camera (5), broad sense cross-correlation module (15), time delay estimation module (4) and Mike's subarray geometry computing module (16) are connected with FPGA control module (3), and time delay estimation module (4) can obtain the time delay between 3 pairs of microphones respectively.

Obtain between 3 pairs of microphones time delay, the position angle of sound source can be tried to achieve by formulae discovery and pitching angle theta, $\mathrm{\θ}=arctan\sqrt{t_{13}^2+t_{24}^2}/t_{56}.$

After the position angle obtaining sound source and the angle of pitch, FPGA controls the rotation of camera as controller, makes it towards the direction of spokesman.

Claims (4)

1. one kind based on the indoor sonic location system of FPGA microphone array, it is characterized in that: comprise microphone array construction module, speech signal pre-processing module, FPGA control system, camera four part, speech signal pre-processing module comprises A/D modular converter and sampling processing module, microphone array is connected with A/D modular converter, and FPGA control system is connected with camera.

2. one according to claim 1 is based on the indoor sonic location system of FPGA microphone array, it is characterized in that: A/D modular converter and sampling processing module form speech signal pre-processing module.

3. one according to claim 1 is based on the indoor sonic location system of FPGA microphone array, it is characterized in that: be fixed in three-dimensional cartesian coordinate system with six microphones and form taper microphone array, just can be obtained position angle and the angle of pitch of sound source position by time delay value.

4. one according to claim 1 is based on the indoor sonic location system of FPGA microphone array, it is characterized in that: FPGA, as controller, controls the direction that camera turns to spokesman.