JPH09145820A - Signal detecting device - Google Patents

Abstract

PROBLEM TO BE SOLVED: To reduce the lowering of detecting capacity caused by the lowering of output generated in association with Doppler effect generated by the movement of sound source. SOLUTION: One input signal SA (t) between input signals in two places is inputted to the specified number (N) of Doppler correcting parts 11i (i=1-N) to obtain output SDi (t) time-compressed or drawn at N-kind of ratios. The cross-correlation of the output SDi (t) and the other signal SB (t) is obtained by a cross-correlation part 12i. N-number of output Ri (τ) are compared by a selecting part 13i every delay time τ, and the maximum value is selected to obtain RM (τ). With either the time-compressed or drawn ratio approximately coinciding with the ratio of the time-compressed or drawn ratios of sound source signals included in the input signals SA (t), SB (t) in two places, the lowering of correlative output by Doppler effect is reduced so as to obtain output equal to the case of no movement of sound source.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a signal detection apparatus, and more particularly to a method for processing a signal received from a broadband noise signal generated by a sound source moving at two different positions together with ambient noise to detect the presence or absence of a sound source. The present invention relates to a signal detection device.

[0002]

2. Description of the Related Art FIG. 2 shows a conventional signal detector of this kind. In FIG. 2, the cross correlation unit 2 has two signals SA
(T) and SB (t) are input, where the cross-correlation is calculated. As shown in FIG. 3, the signals SA (t) and SB (t) are signals received by the receiver A32 and the receiver B33, which are located at different locations, together with the signal generated by the sound source 31 and the ambient noise.

[0003] SA (t) and SB (t) are represented by the following equations.

SA (t) = S (t) + NA (t) (1) SB (t) = S (t−τ ^AB ) + NB (t) (2) where S (t) is generated by the sound source 31 Τ ^AB is a broadband noise signal, and τ ^AB is a distance difference (LB−L) between the distance LA between the sound source 31 and the receiver A32 and the distance LB between the sound source 31 and the receiver B33.
A) is the time required for the sound wave to propagate. Also, N
A (t) and NB (t) are ambient noises and are unrelated.

[0005] A cross-correlation unit 2 in FIG. 2 receives SA (t) and SB (t) as inputs and obtains a cross-correlation function R (τ). R (τ) is obtained by the following equation.

[0006]

When the integration time T is sufficiently long, R in the equation (3)
(Τ) has a maximum value at τ = τ ^AB because there is no correlation between noises NA (t) and NB (t),

[0008]

[0009] When τ is longer than this, it becomes approximately zero. A conventional signal detection device observes R (τ) based on this property, and detects the presence or absence of a signal by using the fact that the value increases at a specific τ.

[0010]

The sound source 31 shown in FIG.
When is moved, the Doppler effect corresponding to the moving speed of the sound source 31 is generated in the received signals at the receiver A 32 and the receiver B 33. This effect is manifested as a temporal compression or expansion of the received signal. As a result, equations (1) and (2) are represented as equations (4) and (5).

SA (t) = S (α × t) + NA (t) (4) SB (t) = S (β × t−τ ^AB ) + NB (t) (5) where α and β are temporal. The rate of compression or stretching, which differs except in certain cases. The value R (τ ^AB ) of the output R (τ) of the cross-correlation unit 2 in FIG. 2 at τ = τ ^AB sharply decreases as the difference between α and β increases. Therefore, in such a case, even if R (τ) is observed, the value does not increase at a specific τ, and there is a problem that signal detection becomes difficult.

[0012]

SUMMARY OF THE INVENTION A signal detection apparatus according to the present invention detects a presence or absence of a sound source by processing a broadband noise signal generated by a sound source at two different positions together with a signal received together with ambient noise. A plurality of Doppler correction units for time-compressing or expanding one input signal at a plurality of predetermined ratios, respectively, and a plurality of cross-correlations between outputs of the plurality of Doppler correction units and the other signal. And a maximum value selector for selecting the maximum value from the output of each cross-correlator.

[0013]

DESCRIPTION OF THE PREFERRED EMBODIMENTS Next, the present invention will be described with reference to the drawings. FIG. 1 is a block diagram of one embodiment of the present invention.
One of the equations (4) and (5), here, SA (t) in the equation (4), is input to a predetermined number N of Doppler correction units 11i (hereinafter, i = 1 to N). You. The Doppler correction unit 11i includes a writing unit 11i1 and a storage unit 11i.
2 and a reading unit 11i3. The writing unit 11i1 writes an input signal waveform to the storage unit 11i2 at a predetermined speed, and the reading unit 11i3 sets the writing unit 11i1.
The signal waveform is read from the storage unit 11i2 at a predetermined speed equal to or different from. If the writing speed and the reading speed are different, the signal is compressed or stretched in time. Let this ratio be ki. The N Doppler correction units 11i include:
All have different ratios ki of temporal compression or stretching.

Output SDi (t) of Doppler correction unit 11i
Is represented by equation (6).

SDi (t) = SA (ki × t) = S (ki × α × t) + NA (ki × t) (6) The output SD1 (t) of the Doppler correction unit 11i is expressed as SB ( Together with t), it is input to a predetermined number N of cross-correlation units 12i. The cross-correlation unit 12i calculates SDi according to equation (7).
A cross-correlation function Ri (τ) between (t) and SB (t) is obtained.

[0016]

When the integration time T is sufficiently long, R in the equation (7)
Since i (τ) has no correlation with the noises NA (ki × t) and NB (t), the maximum value is given at ki × α ≒ β and τ = τ ^AB ,

[0018]

It takes about 0 and becomes approximately 0 in other cases.

The output Ri (τ) of the cross-correlation unit 12 i is input to the selection unit 13. In the selection unit 13, according to equation (8),
The maximum value of Ri (τ) is obtained for each τ, and the output RM (τ) is output.

RM (τ) = max {Ri (τ)} (8) As a result, even when the sound source 31 moves, the correlation output RM is equivalent to the case where the sound source 31 does not move in the conventional method.
(Τ) is obtained.

[0022]

As described above, the signal detection apparatus according to the present invention comprises a predetermined number of Doppler correctors for time-compressing or expanding one input signal at a predetermined ratio, an output of each Doppler corrector and the other. Has a predetermined number of cross-correlation units for calculating the cross-correlation of the signals and a maximum value selection unit for selecting the maximum value from the output of each cross-correlation unit. In this case, there is an effect that the decrease in the correlation output is small and the detection of the presence or absence of a signal is facilitated.

[Brief description of the drawings]

FIG. 1 is a block diagram of one embodiment of the present invention.

FIG. 2 is a block diagram of a conventional signal detection device.

FIG. 3 is a diagram showing positions of a sound source and a receiver.

[Description of Signs] 11i Doppler correction unit 11i1 writing unit 11i2 storage unit 11i3 reading unit 12i cross-correlation unit 13 selection unit 2 cross-correlation unit 31 sound source 32 receiver A 33 receiver B SA (t) reception signal SB of receiver A (T) Received signal of receiver B SDi (t) Output of i-th Doppler correction section Ri (τ) Output of i-th cross-correlation section RM (τ) Output of selection section R (τ) Cross-correlation section Output of LA Distance between sound source and receiver A LB Distance between sound source and receiver B

 ─────────────────────────────────────────────────── ─── Continuation of the front page (72) Inventor Kazunori Ebana 678 Futamata, Ichikawa City, Chiba Prefecture Defense Agency Dormitory 14-504 (72) Inventor Yasushi Sasaki 2-4-16-203 Oyamaguchi, Shirai Town, Inba District, Chiba Prefecture

Claims (4)

[Claims] 1. A receiver that is arranged at two different places and converts sound waves into an electric signal, a plurality of Doppler correction units that compress or expand the time axis of the electric signal from one of the receivers at different ratios, A signal detection device comprising: a plurality of cross-correlation units for obtaining a cross-correlation between an output signal of each of the plurality of Doppler correction units and an electric signal from the other receiver, and detecting a signal based on outputs from the plurality of cross-correlation units apparatus. 2. The signal detection device according to claim 1, further comprising a maximum value selection unit that receives outputs of the plurality of cross-correlation units and selects a maximum value. 3. The signal detection apparatus according to claim 1, wherein said plurality of Doppler correction units include both a unit that compresses a time axis and a unit that extends a time axis. 4. The signal detecting apparatus according to claim 1, wherein said sound wave has a distribution of broadband noise.