JPH0370793B2 - - Google Patents

Description

[Detailed Description of the Invention] Industrial Application Field The present invention relates to a sound direction detection method for detecting the direction of arrival of a sound source, and is useful for collecting and photographing a moving object accompanied by sound, such as an automatic tracking camera, an automatic tracking telephoto microphone, etc. It can be used as a service robot or a measuring device to measure the direction of arrival of indirect sound in a room.

Conventional Example Configuration and Its Problems A typical conventional means for detecting the direction of sound is a sound collection method using two microphones. This method uses the time difference between the sounds arriving at the two microphones as information on the direction of arrival, and generally detects the left and right directions, but cannot distinguish between the front and back directions. Furthermore, this method is a sound collection method for so-called stereo reproduction, and no method has been found that has been used as an automatic control means for equipment as a sound direction detection function. Furthermore, so-called telephoto microphones that have high sensitivity in specific directions do not have a sound direction detection function by themselves.

Purpose of the Invention The present invention aims to provide a sound direction detection method that uses a combination of a plurality of microphones and a sound signal processing circuit to obtain an electrical signal output as information on the direction of sound coming from all around. By driving the rotation control system of the device with the obtained electrical signal output, for example, a camera or other photographing device that automatically tracks the subject of sound, a service robot that works by following the person who gives the command, or an automatic tracking telephoto microphone. Moreover, it can be applied to detecting and measuring the direction of incidence of indirect sound in a room.

Structure of the Invention In order to achieve the above object, the present invention includes a sound collection section having a configuration of three or more microphones, and an electric signal processing circuit for converting the electric signal from each microphone into a sound direction detection signal.

DESCRIPTION OF THE EMBODIMENTS The present invention will be described in detail below with reference to the illustrated embodiments. FIG. 1 shows an example of a sound collection section composed of three microphones arranged in the same plane. Three microphones A, B, and C are arranged on the orthogonal coordinate axes to form a triangle as shown in the figure, and the distance between microphones A and B is the sound velocity time difference T _AB , and the time difference between microphones A and C is T _A − Let it be C. Furthermore, 2π centered on microphone A on the coordinate axis plane
Let's take the plane of rotation and assume that a sound is coming from the θ ₁ direction.

On the other hand, Figure 2 shows these microphones A, B, and C.
This is an example of a circuit that processes an electrical signal from a sound direction and obtains a sound direction detection electrical signal. In Figure 2, 1, 2, and 3 are level amplifiers, and microphones A,
This amplifies the weak sound signals of B and C to an appropriate level that is easy to process. 4, 5, and 6 are band-pass filters that extract only the characteristic components of the sound spectrum of the sound source, increasing the SN, detection sensitivity, and accuracy, and filtering out unnecessary sound signals for detection such as background noise. It is to be excluded. 7,
Reference numerals 8 and 9 indicate level comparators, and when a signal level higher than the reference voltage V _TH is applied to the + terminal 10, the output terminals thereof become H (high level). The operation of this level comparator is shown in FIGS. 3a and 3b, where a is the output waveform of the filter 4 and b is the output waveform of the filter 5. Microphone A in Figures 1 and 2,
If B and C, amplifiers, and filters with the same characteristics are used, the level comparators 7, 8, and 9 are applied with sound signals having the same waveform but differing only in time difference. If the reference voltage V _TH is set to a level slightly higher than the level of unnecessary sound signals such as background noise, each level comparator 7, 8, 9
, the moment the reference voltage V _TH is exceeded, the output terminal goes high (high level), and microphones A and
The time of the sound at the moment when it reaches B and C will be detected. Furthermore, the outputs of these level comparators 7, 8, and 9 are connected to the set terminals of flip-flops 11, 12, and 13, and the reset terminal 14 is set to L (low).
state, the set data is held during this time, and this output data is sent to the flip-flops 15 and 1 of the next stage.
When connected to the set and reset terminals 6 and 17 as shown in Fig. 2, the output of the flip-flop 15 has a pulse width equal to the difference in timing between set and reset, that is, the difference in arrival time of sound to microphones A and B. can get. At this time, no output from flip-flop 16 occurs. The reason for this is that the timing of data input to the reset terminal is earlier than the timing of data input to the set terminal. Conversely, when an output is generated at flip-flop 16, no output is generated at flip-flop 15. This inverted _output operation coincides with the _direction of arrival of _the sound in FIG. is output, and is used to determine the front and rear directions of sound.

Here, the output pulse width of flip-flop 15 is
T ₁ is T ₁ = T (A - B) sin θ ₁ (where 0 < θ ₁ < π). Similarly, the output pulse T ₁ ' of the flip-flop 16 is T ₁ ' = T (A - B) sin θ ₁ (where π <θ ₁ <2π) On the other hand, the flip-flop 17 determines whether the sound direction is left or right, and resets the output of the microphone A with the output of the microphone C. That is, when the sound direction θ ₁ is in the range of π/2<θ ₁ <-π/2, the arrival time of the sound to the microphone A is always earlier than that to the microphone C, so that the data applied to the set terminal of the flip-flop 17 is applied to the reset terminal. Since it is faster than that of , a pulse output appears at the output terminal (out3). In the range of π/2<θ ₁ <3/2π, the reset timing is earlier, so no output is generated at the output terminal (out3). In other words, this operation determines whether the sound direction is left or right.

The output characteristics of the output terminals out1 to out3 of the flip-flops 15, 16, and 17 are shown in FIG. In FIG. 4, the vertical axis shows the width of the output pulse, and the horizontal axis shows the sound direction angle. Depending on the synchronization and presence/absence of output pulses from these output terminals out1 to out3, the sound direction and all directional angles of front, rear, left, and right on the same plane can be detected.

The above detection operations are sequentially performed by a reset signal sent to the reset terminal.

The sound direction detection method according to the present invention detects the upper and lower angles of the sound direction by placing each microphone arrangement plane on the vertical axis, and by adding one more microphone, it is possible to detect both horizontal and vertical sound directions. It has a detection function for.

Effects of the Invention According to the present invention, the sound direction detection output makes it possible to cause a controlled device to work on an object that emits a sound, for example, the following can be done.
For example, if the rotation of the photographing direction rotation mechanism of the camera is controlled by the main detection output, an automatic tracking function can be provided.

When applied to voice recognition type robots, it is also possible to have a commanding person perform a specific task.

If applied to directional control of a telephoto microphone,
Automatic tracking allows you to collect only the necessary sounds with good SN.

If the reset terminal in FIG. 2 is replaced with a measurement timing gate, it is possible to detect the direction of incidence of indirect sound from a room or the like that is successively incident, which can be utilized for sound field analysis and measurement.

The output of the amplifier of each microphone can be used as is for sound recording, and at the same time, by inserting a phase shift circuit between each output, it is possible to provide a telephoto microphone function with arbitrary directional characteristics, for example.

[Brief explanation of drawings]

Fig. 1 is a layout diagram of the microphone sound collection section in the present invention, Fig. 2 is an electrical signal processing circuit diagram for obtaining sound direction detection outputs on the same plane from three microphones, and Fig. 3 is the same as Fig. 2. An explanatory diagram of the operation of the electric signal processing circuit, and FIG. 4 is an output characteristic diagram of the electric signal processing circuit of FIG. 2. A, B, C are microphones, 1, 2, 3 are level amplifiers, 4, 5, 6 are band pass filters,
7, 8, and 9 are level comparators, and 11 to 17 are flip-flops.

Claims (1)

[Scope of Claims] 1. A sound collection section composed of at least three or more microphones arbitrarily arranged to form a triangle in the same plane, and an electric signal from each microphone is transmitted through an SN improvement means. Compare it with the reference voltage, operate the logic circuit with this comparison output,
a first signal processing circuit section that obtains pulse outputs A, B, and C that are proportional to the difference in sound arrival time between the respective microphones; A sound direction detection system comprising: a second signal processing circuit section that obtains left and right discrimination pulses for sound directions based on A and C; 2. The sound direction detection method according to claim 1, further comprising a sound collection section configured with at least four or more microphones arbitrarily arranged to form a triangle in each of a plurality of planes.