JPH0592767U - Approaching vehicle recognition device - Google Patents

Abstract

(57) [Summary] [Purpose] To accurately detect the direction of a vehicle approaching the vehicle. [Structure] Acoustic signals output from a plurality of microphones 1 and 2 arranged at a predetermined interval are passed through a band pass filter 33,
By 34, it is converted into a corrected acoustic signal from which frequency components in the low frequency band and the high frequency band are removed. The corrected acoustic signal is limited by the band limiter 41 to a predetermined frequency band in which the characteristics of the running sound of the vehicle are well expressed, the power is calculated by the power calculator 42, and the approach determiner 43 approaches when the power is larger than the predetermined value. It is determined that there is a car. The corrected acoustic signal is converted by the noise suppressor 51 into a noise suppression signal from which unnecessary noise components have been removed. The correlation calculator 52 calculates the cross-correlation of a plurality of noise suppression signals, and the time difference calculator 53
The arrival time difference that maximizes the correlation is calculated by. The direction calculator 54 calculates the approach direction of the approaching vehicle from the arrival time difference.

Description

[Detailed description of the device]

[0001]

[Industrial applications]

 The present invention relates to an approaching vehicle recognition device that recognizes a vehicle approaching the vehicle without being affected by the sound generated by the vehicle from the ambient sounds and identifies the arrival direction of the vehicle. In particular, it is installed in automobiles, etc., and is used as a device to support safe driving by recognizing approaching vehicles from the ambient sound at intersections, identifying the direction of arrival of those vehicles, and notifying the driver etc. To be done.

[0002]

[Prior Art]

 Conventionally, a device that is installed in an automobile or the like and recognizes the type and direction of arrival of various alarm sounds such as railroad crossings and emergency automobiles from outside sounds, intermittent rotation (scanning) of directional microphones, and multiple directional microphones A device for detecting the approach and the approach direction of a moving object such as a vehicle is known. There is also a device that recognizes the moving direction of a moving sound source by calculating the cross-correlation value of the acoustic signal waveforms from the acoustic signals as a time function output by the two microphones, and obtaining the difference in the arrival time of the waveforms between the microphones. Are known. Further, a noise suppression method (spectral subtraction method, etc.) for recognizing a voice on which noise is superimposed under noise is known.

[0003]

[Problems to be solved by the device]

 However, in these conventional devices, when detecting the approach and arrival direction of a vehicle or the like, what can be recognized is limited to an alarm sound, and when installed in a vehicle or the like, in order to eliminate the influence of the sound generated by the own vehicle. There was a problem that it was necessary to use a directional microphone.

The present invention has been made to solve the above problems, and an object of the present invention is to detect an approaching vehicle such as an emergency vehicle when detecting the approaching and arrival direction of the vehicle. The approaching vehicle is recognized without being affected by the sound generated by the vehicle, not limited to the vehicle that emits the alarm sound, and the omnidirectional microphone is not limited to the microphone as the directional microphone. It is to be able to identify the direction of arrival and inform the driver etc.

The present inventor has conducted an analysis by a simulation and an in-vehicle experiment in order to develop an approaching vehicle recognition device that solves the above-mentioned problems of the prior art, and as a result, when recognizing a voice on which noise is superimposed under noise. By using the spectral subtraction method, which is a noise suppression method often used for example, it is possible to eliminate the influence of the sound generated by the own vehicle and accurately identify the approaching direction of the approaching vehicle.

[0006]

[Means for Solving the Problems]

 As shown in FIG. 1, the approaching vehicle recognition device of the present invention detects an ambient sound, outputs an acoustic signal corresponding to the ambient sound, and outputs a plurality of acoustic signals arranged at predetermined intervals. The converter X1 and the bandpass filter X2 that removes the frequency components of the low-frequency band and the high-frequency band from the acoustic signals output by the plurality of acoustoelectric converters, respectively, and the vehicle running from the output signals from the bandpass filter A power calculating means X3 for calculating the power of a predetermined frequency band in which the characteristics of the sound are often expressed, a vehicle approach determining means X4 for determining the approach of the vehicle from the power calculated by the power calculating means, and a vehicle approach determining means. When it is determined that a vehicle is approaching, unnecessary noise components are removed from each output signal from the band pass filter, and a noise suppressing unit X5 that outputs each noise suppressing signal and a noise suppressing unit are provided. Of the approaching vehicle from the time difference calculated by the time difference calculation means X6 for calculating the time difference of the vehicle traveling sound reaching the plurality of acoustoelectric converters from the cross-correlation of the plurality of noise suppression signals output by A direction identifying means X7 for identifying the arrival direction of the traveling sound is provided.

[0007]

[Action]

 Ambient sound of the vehicle is picked up by a plurality of acoustoelectric converters X1, and an acoustic signal corresponding to the ambient sound is output from the acoustoelectric converter X1. The acoustic signal is input to the band pass filter X2, and low frequency components and high frequency components are removed. As a result, noise components that do not primarily contribute to the running noise of the adjacent vehicle are removed. Next, the power calculation means X3 calculates the power of a predetermined frequency band in which the characteristics of the traveling sound of the approaching vehicle are well expressed from each output signal from the band pass filter X2, and the calculated power is used to calculate the vehicle power. The approach determination means X4 determines the approach of the vehicle.

When it is determined that the vehicle is approaching, the noise suppression unit X5 removes unnecessary noise components from the output signals corresponding to the acoustoelectric converters from the bandpass filter X2, and the noise suppression signals are removed. Is output. After that, the time difference calculation means X6 calculates the time difference of the vehicle traveling sound reaching the plurality of acoustoelectric converters from the cross-correlation of the plurality of noise suppression signals, and the direction identification means X7 uses the calculated time difference to detect the arrival of the approaching vehicle. The coming direction is identified.

As described above, in the present invention, when identifying the direction of arrival of an approaching vehicle, the cross-correlation values of the acoustic signals as a plurality of time functions output by the plurality of acoustic-electric converters are not calculated, but The spectrum of the acoustic signal is analyzed and the spectral component of the sound generated by the vehicle is subtracted by the spectral subtraction method.Then, the noise suppression signal as a time function is returned to remove the effect of the sound generated by the vehicle. The cross-correlation value of the noise suppression signal is calculated and the time difference of the sound waves reaching the acoustoelectric converters is calculated to identify the direction of arrival of the approaching vehicle.

[0010]

[Effect of the device]

 Since the approaching vehicle recognition device of the present invention determines the approaching of the vehicle from the power of the signal in the band in which the characteristics of the vehicle running sound often appear, the approaching of the vehicle can be recognized without being limited to the vehicle that particularly emits the warning sound. Further, since the device of the present invention eliminates the influence of the sound generated by the own vehicle by the noise suppression means, it is possible to use the omnidirectional acoustoelectric converter without using the directional acoustoelectric converter. The direction of arrival of an approaching vehicle can be accurately identified without being affected by the sound generated by. In addition, the device of the present invention recognizes the approaching and arrival direction of a vehicle from the sound emitted by the vehicle and notifies the driver, etc. from the sound emitted by the vehicle when the visibility is poor and the approaching vehicle cannot be seen. It is possible to support safer driving.

[0011]

【Example】

 Hereinafter, the present invention will be described based on a specific embodiment. FIG. 2 is a diagram showing the configuration of the apparatus of the embodiment. The approaching vehicle recognition device determines the approach of a vehicle from two microphones 1 and 2 that collect sounds around the vehicle, an acoustic signal input unit 3 that samples an acoustic signal from the microphones, and a sampled acoustic signal. A vehicle approach determination unit 4, a direction identification unit 5 that identifies the arrival direction of the vehicle when the vehicle approach determination unit 4 determines that the vehicle is approaching, and a buzzer sound is output when the approach is determined. An alarm device 61 and a display device 62 for displaying the result of identification of the vehicle arrival direction by the direction identification unit 5 are configured.

FIG. 3 shows a case in which two microphones are installed at the front end of the host vehicle A in the horizontal direction at a distance d and the approach direction of the vehicle B is recognized to identify the direction of arrival. Further, the microphone can be installed on the roof as shown in FIG. 4 or in the left and right door mirrors (or fender mirrors) not shown.

Here, in the acoustic signal input section 3, first, the acoustic signals from the microphones 1 and 2 are input to and amplified by the microphone amplifiers 31 and 32. The acoustic signal is input to the band pass filters 33 and 34 in order to remove the influence of wind in the low frequency range and the noise generated by the vehicle and the influence of noise in the high frequency range, and the low frequency component and the high frequency component are removed. Converted to a corrected sound signal. After that, the corrected acoustic signal is converted from analog data to digital data by the A / D converter 35.

Next, in the vehicle approach determination unit 4, the corrected acoustic signal corresponding to the acoustic signal output from one of the acoustic signal input units 3, for example, the right microphone 1 is output to the approaching vehicle by the band limiter 41. It is limited to the frequency band in which the characteristics of the running sound of the vehicle appear. Then, the power calculator 42 calculates the power of the limited band of the corrected acoustic signal. Next, the approach level determiner 43 compares the power level with a predetermined threshold value, and determines that the vehicle is approaching when the calculated power level of the running sound exceeds a certain value. As another determination method, it is also possible to determine that the vehicle is approaching when the degree of increase in the power of the traveling sound is higher than a certain rate.

When the approach of the vehicle is determined by the vehicle approach determination unit 4, the alarm 61 is activated and a buzzer sound or the like is output to a speaker or the like to notify the driver of the approach of the vehicle. Then, the direction identifying unit 5 is operated to identify the arrival direction of the approaching vehicle. To the direction identifying unit 5, the corrected acoustic signal obtained by correcting the acoustic signals from the two microphones 1 and 2 from the acoustic signal input unit 3 by the band pass filters 33 and 34 is input. When the microphone is installed at the front end of the vehicle, the noise generated by the vehicle is large, so the noise suppressor 51 performs noise suppression processing on the input corrected acoustic signal. As the noise suppression method, the spectral subtraction method, which is often used when performing speech recognition under noise, is used.

FIG. 5 shows an algorithm of the spectral subtraction method. Each corrected acoustic signal output from the bandpass filter X2 is Fourier transformed by the FFT processing 100 to obtain a frequency function. Next, the amplitude spectrum is calculated by the square process 101 of the absolute value of the frequency function. Next, the amplitude spectrum of the own vehicle noise is subtracted from the amplitude spectrum of the detected sound 102, and the amplitude spectrum of the approaching vehicle traveling sound is estimated. Next, the absolute value of the amplitude of the frequency function corrected by the square root process 103 of the corrected amplitude spectrum is calculated. From the absolute value of the amplitude and the phase information, the inverse FFT processing 104 generates a time function for the corrected frequency function, that is, a noise suppression signal. As a result, a signal in which the vehicle noise component is suppressed can be obtained.

The cross-correlation of the noise suppression signals as two time functions corresponding to the two acoustic signals output from the two microphones 1 and 2 is calculated by the correlation calculator 52. Then, as shown in FIG. 6, the time difference calculator 53 calculates the time difference when the value of the cross-correlation becomes maximum. This time difference is the signal arrival time difference τ when the sound waves emitted by the approaching vehicle reach the two microphones 1 and 2. From the arrival time difference τ, the direction calculator 54 calculates the arrival direction θ of the approaching vehicle. When the arrival direction is calculated, the result is notified to the driving vehicle or the like by the display device 62 such as a display and a lamp.

The relationship between the arrival time difference τ and the arrival direction θ of the approaching vehicle has the following relationship as shown in FIG.

## EQU00001 ## cos .theta. =. Tau.v / d where v is the speed of sound and d is the distance between the two microphones 1 and 2.

As described above, in the device of the present embodiment, the approach of the vehicle is determined from the power of the signal in the band in which the characteristic of the vehicle running sound often appears. Therefore, the approach of the vehicle is not limited to the vehicle that particularly emits the alarm sound. Can be recognized. Further, since the apparatus of the embodiment eliminates the influence of the sound generated by the own vehicle by performing the noise suppression, it is not limited to the directional microphone and is not affected by the sound generated by the own vehicle by the omnidirectional microphone. The direction of arrival of an approaching vehicle can be accurately identified. In addition, the device of the embodiment recognizes the approaching and arrival direction of a vehicle from the sound emitted by the vehicle when the visibility is poor and the approaching vehicle cannot be seen at an intersection or the like, and notifies the driver of the safety of driving. Can help.

[Brief description of drawings]

FIG. 1 is a block diagram showing the concept of the present invention.

FIG. 2 is a configuration diagram of an approaching vehicle recognition device according to a specific embodiment of the present invention.

FIG. 3 is an explanatory view showing a specific installation place of a microphone.

FIG. 4 is an explanatory view showing a specific installation place of a microphone.

FIG. 5 is a block diagram showing a processing procedure of a spectral subtraction method for obtaining a noise suppression signal.

FIG. 6 is an explanatory diagram showing a method of calculating the direction of an approaching vehicle.

[Explanation of symbols]

 1, 2 ... Microphone (acoustoelectric converter) 3 ... Acoustic signal input unit 4 ... Vehicle approach determination unit 5 ... Direction identification unit 33, 34 ... Band pass filter 41 ... Band limiter (power calculation means) 42 ... Power calculator (Power calculation means) 43 ... Approach determination device (vehicle approach determination means) 51 ... Noise suppressor (noise suppression means) 52 ... Correlation calculator (time difference calculation means) 53 ... Time difference calculator (time difference calculation means) 54 ... Direction calculation Device (direction identifying means) 62 ... Display device

Claims (1)

[Scope of utility model registration request] 1. An ambient sound is detected, an acoustic signal corresponding to the ambient sound is output, and a plurality of acoustoelectric converters arranged at predetermined intervals, and the plurality of acoustoelectric converters output the acoustic signals. In each of the acoustic signals, a bandpass filter that removes frequency components in a low frequency band and a high frequency band, and a power of a predetermined frequency band in which the characteristic of the running sound of the vehicle is often expressed is calculated from each output signal from the bandpass filter. And a vehicle approach determination unit that determines the approach of the vehicle from the power calculated by the power calculation unit, and when the approach of the vehicle is determined by the vehicle approach determination unit, Noise suppression means for removing unnecessary noise components from each output signal and outputting each noise suppression signal, and the plurality of noise suppression signals output by the noise suppression means From the cross-correlation between the acoustoelectric converters, a time difference calculating means for calculating a time difference of arrival of the vehicle running sound, and a direction for identifying the arrival direction of the running sound of the approaching vehicle from the time difference calculated by the time difference calculating means. An approaching vehicle recognition device, comprising: identification means.