KR20150114373A

KR20150114373A - Appartus and method for monitoring security using variation of correlation coefficient pattern in sound field spectra

Info

Publication number: KR20150114373A
Application number: KR1020140146266A
Authority: KR
Inventors: 박강호; 이성규
Original assignee: 한국전자통신연구원
Priority date: 2014-03-31
Filing date: 2014-10-27
Publication date: 2015-10-12
Also published as: KR102034559B1

Abstract

A security monitoring method according to the present invention includes the following steps: outputting a multi tone sound wave constituted with a linear sum of a sine wave having a plurality of frequency components into a security monitoring space; calculating a sound field from the multi tone sound wave received in the security monitoring space; storing reference sound field information per frequency through the sound field; calculating current sound field information per frequency, and then judging occurrence of sound field variation by comparing the reference sound field information per frequency with the calculated sound field information; and discriminating at least two situations among invasion, movement and temperature variation situation on the basis of correlation between a reference sound field spectrum and a continuous sound field spectrum, by analyzing the sound field variation occurrence collected during a fixed period.

Description

TECHNICAL FIELD [0001] The present invention relates to a security monitoring apparatus using a variation pattern of a correlation coefficient of a sound field spectrum and a security monitoring method thereof. [0002]

The present invention relates to a method and apparatus for estimating a soundness of a sound field based on a change pattern of a correlation coefficient between sound field spectra based on a multi-tone frequency based on a sound field change over time The present invention relates to a security monitoring apparatus that detects occurrence, intrusion, motion, fire, and daily temperature change situations, and a security monitoring method thereof.

Security sensors used to independently detect intrusion and fire situations have long been studied and used. Sensors for detecting intrusions can use one of passive infrared detection (PIR), IR, ultrasonic, sound, vibration, and microwave sensing. Sensors for detecting fire include temperature sensing, A smoke detection method, a gas detection method, and a flame detection method.

However, as described above, these sensors are used as sensors for detecting one of intrusion and fire situations. As a single sensor, the security sensor that senses the intrusion and the fire separately detects the sound field change pattern measured by the sound source having multi-tone frequency and extracts the feature according to the time and frequency change and uses the principle of analyzing it .

As a background art, there is a sound field security pattern technique for calculating an average and a deviation of a sound field and detecting a dangerous situation of an intrusion based on a change value (SNR) of an average value of sound field versus an initial deviation of a reference sound field and informing an alarm -0142499, Security System and Method Through Analysis of Sound Field Change Patterns). However, this patented invention has a disadvantage in that it takes a long time to determine the reference value since it is necessary to perform at least two measurements at first in order to obtain the deviation of the reference sound field, and there is a method of detecting the randomness of the reference deviation obtained by a limited number of measurements and the sound field change There is a weak point that the reliability of the intrusion detection becomes weak due to the inaccuracy. In addition, there was a weak point in this way that it was difficult to distinguish between intrusion, fire, and movement.

As another conventional art, a change in temperature such as a fire is characterized in that the shape of a sound field pattern does not change but moves in a higher frequency direction. In the case of an intrusion situation, There is a technique for distinguishing an invasion situation (Korean Application No. 2013-0122862, security surveillance system and security surveillance method). However, this patented invention also shows a method of quantifying the degree of variation of the pattern shape and a method of deriving the similarity index through the difference index because of the inaccuracy of the method of detecting the sound field change by the change value of the sound field average value relative to the reference sound field deviation It is difficult to precisely quantify the frequency shift index because the randomness is high and the result of obtaining the frequency shift index according to the situation is variable. In addition, there was a weak point that it was difficult to detect motion separately.

It is an object of the present invention to provide a method and apparatus for detecting the occurrence of a dangerous situation causing a sound field change based on a mutual correlation coefficient pattern of a sound field spectrum according to a frequency of a multitone sound source based on time- The present invention provides a highly reliable security monitoring apparatus and method that separately detect daily temperature change situations such as fire, day-to-day, and air-conditioning.

It is an object of the present invention to provide a security monitoring apparatus and method based on pattern variation detection of a correlation coefficient of a sound field spectrum that provides comprehensive security monitoring to detect and distinguish intrusion, have.

A security monitoring method according to an embodiment of the present invention includes: outputting a multi-tone sound wave composed of a linear sum of sine waves having a plurality of frequency components into a security monitoring space; Receiving and calculating a sound field from the multi-tone sound wave; Calculating and storing frequency-specific sound field information through the sound field; Comparing the frequency-dependent reference sound field information with the calculated sound field information to determine whether a sound field change has occurred; Analyzing the occurrence of the sound field change collected for a certain period of time, and classifying at least two situations among the intrusion, motion, and temperature change situations based on the correlation between the reference sound field spectrum and the continuous sound field spectrum.

In an embodiment, the correlation is obtained by calculating a correlation coefficient value between the reference sound field spectrum and the continuous sound field spectrum.

In an embodiment, the correlation is obtained using a correlation coefficient calculated by dividing the covariance value of the reference sound field spectrum and the continuous sound field spectrum by the product of the standard deviation of each sound field spectrum.

In an embodiment,

Comparing the reference sound field spectrum with a current sound field spectrum to determine whether a dangerous situation causing the sound field change occurs; Analyzing the sound field spectrum collected for a predetermined period before the dangerous situation to distinguish the intrusion, the temperature change, or the motion situation; And a step of classifying intrusion, movement, fire, and daily temperature change conditions of the daytime and nighttime on the basis of the correlation coefficient between the reference sound field spectrum and the continuous sound field spectrum.

In an embodiment, a correlation between the reference sound field spectrum and the current sound field spectrum is compared with a set reference value to determine whether a dangerous situation has occurred.

In an embodiment, the correlation coefficient between the reference sound field spectrum and the current sound field spectrum is used, and an initial correlation coefficient at the reference sound field measurement as an index indicating a different degree of mutual sound field spectrums is set from 1 to a limit value, The correlation coefficient with the sound field is compared with the limit value from 1 to determine whether or not a dangerous situation occurs.

In an embodiment, a change pattern of a correlation coefficient between the reference sound field spectrum and the continuous sound field spectrum is used, and a sudden decrease immediately before the occurrence of a dangerous situation is determined as an intrusion, The change situation is judged.

In an embodiment, the average value of the correlation coefficients between the reference sound field spectrum and the continuous sound field spectrum before the occurrence of the dangerous situation is set to a limit value from 1, a sound field spectrum at the time when the dangerous situation occurs, The intrusion and the temperature change situation are distinguished by comparing the ratio of the limit value to the correlation coefficient value between the spectra.

In an embodiment, a temporal variation pattern of the correlation coefficient between the reference sound field spectrum and the continuous sound field spectrum is analyzed, and the variation pattern that is irregularly increasing or decreasing is determined as a motion, The situation is determined.

Comparing the reference sound field spectrum with the current sound field spectrum to determine whether a sound field change has occurred; And analyzing the sound field spectrum of each frequency collected for a certain period of time before the sound field change situation occurs to distinguish the movement of the person / animal, the fire, and the daily temperature change condition of the day and night.

The method may further include comparing a correlation coefficient between the reference sound field spectrum and the current sound field spectrum with a set reference value to determine whether a sound field change situation has occurred.

In an embodiment, a temporal variation pattern of the correlation coefficient between the reference sound field spectrum and the continuous sound field spectrum is analyzed and it is determined that the variation is a motion when the variation pattern is irregular, and the variation pattern gradually decreases gradually It is determined as a temperature change situation.

In the embodiment, the absolute value of the difference between the front and rear correlation coefficients obtained by continuous measurement using the correlation coefficient between the reference sound field spectrum and the continuous sound field spectrum for a predetermined period before the sound field change situation is summed Value and a value obtained by subtracting the correlation coefficient value at a point of time when the sound field change is detected from 1 is set as a motion index, and a motion and a temperature change are distinguished by using the motion index.

In the embodiment, an index indicating the degree of frequency shift of the sound field spectrum is derived based on a correlation coefficient obtained by using multitone frequency between the reference sound field spectrum and the continuous sound field spectrum as a variable, And the temperature change condition of the fire and the daily routine and the heating /

Detecting movement information of a person and an animal inside the security space using sound field information and storing the movement information; And transmitting the detection information to the smartphone and the smart device of the guardian.

In an embodiment, the method further comprises transmitting alarms of an accident occurrence of falling, stalling and non-moving when the movement of people and animals inside the security space is not detected for a predetermined period of time and transmitting the security information.

In an embodiment, the method further includes selectively detecting only a fire situation in the presence of movement of a person or an animal in the security space, transmitting a fire alarm, and transmitting security information.

In an exemplary embodiment, the method further includes performing image capturing to store image information and verify the situation when a security situation occurs.

In an embodiment, the security monitoring method is interlocked with a smart home appliance including a security camera having a network function, an Internet phone, a smart TV, and an interphone.

In the embodiment, it is implemented in software without adding hardware of the interlocking device in the interlocking.

In an embodiment, when a program in the form of an app related to a user's smartphone or smart device is executed, the security monitoring method transmits security information executed or obtained by remote control.

The security monitoring apparatus according to an embodiment of the present invention includes: a sound generating device that outputs a sound wave according to an input voltage in a security monitoring space; A sound wave receiving device for receiving the sound wave and calculating a sound field using the sound wave; Calculating a cross-correlation coefficient between the continuous sound field spectral information and the reference sound field spectrum information through successive measurements, and calculating a cross-correlation coefficient between the continuous sound field spectral information and the reference sound field spectrum information by using the cross correlation coefficient, A sound field signal processing device for distinguishing at least two situations among the sound field signal processing devices.

In an embodiment, the sound wave is a multi-tone sound wave consisting of a linear sum of sine waves having a plurality of frequency components.

In an embodiment, the apparatus further comprises a memory for storing the reference sound field spectrum information.

In an embodiment, the sound field signal processing apparatus calculates a sound pressure or phase of the sound field using an acoustic transfer function.

FIG. 1 is a block diagram of a security monitoring apparatus that detects and detects daily temperature changes such as intrusion, fire, and cooling / heating and daytime based on a correlation coefficient of a sound field spectrum.
FIG. 2 is a diagram showing a change in a sound field spectrum when a sudden intrusion state occurs in a reference sound field and continuous sound field measurement using a multi-tone sound source having a center frequency of 4 kHz and a frequency interval of 4 Hz.
FIG. 3 is a graph showing changes in the cross-correlation coefficient obtained by the frequency shift between the reference sound field and the continuous number of times of measurement in the sound field spectrum obtained in FIG. 2 for each number of times of 2, 6, and 10 FIG.
FIG. 4 is a diagram showing correlation coefficients (m = 0) without frequency shift between the reference sound field and consecutive times measurement values in the sound field spectrum obtained in FIG.
FIG. 5 is a graph showing the maximum value of the cross correlation coefficient obtained by considering the frequency shift between the reference sound field and the continuous number of times of measurement in the sound field spectrum obtained in FIG. 2 for each number of times.
FIG. 6 is a graph showing a frequency shift index corresponding to a maximum value of a correlation coefficient for each measurement number in order to indicate how much the frequency of the consecutive times of frequency shifts compared with the reference sound field in the sound field spectrum of FIG.
FIG. 7 is a graph showing a gradual change in the sound field spectrum when a temperature change due to a fire situation occurs from the beginning in continuous sound field measurement using a multi-tone sound source with a center frequency of 4 kHz and a frequency interval of 4 Hz.
FIG. 8 is a diagram showing the changes of the cross correlation coefficient obtained by using the frequency shift between the reference sound field and the continuous number of times of measurement in the sound field spectrum obtained in FIG. 7 by the number of times of 2, 6, and 10 times.
FIG. 9 is a graph showing correlation coefficients (m = 0) between the reference sound field and the consecutive times of measurement without frequency shift in the sound field spectrum obtained in FIG.
10 is a graph showing the maximum value of the cross correlation coefficient obtained by considering the frequency shift between the reference sound field and the continuous number of times of measurement in each sound field spectrum obtained in FIG.
FIG. 11 is a graph showing the frequency shift indexes corresponding to the maximum value of the correlation coefficients for each measurement number in order to indicate how frequency the consecutive times of the frequency shifts compared with the reference sound field in the sound field spectrum of FIG.
FIG. 12 is a diagram showing a change in the sound field spectrum when the intruding situation occurs suddenly (15 times) in a continuous sound field measurement with a reference sound field using a multitone sound source with a center frequency of 6 kHz and a frequency interval of 4 Hz.
FIG. 13 is a diagram showing changes in the cross correlation coefficient obtained by using the frequency shift between the reference sound field and the continuous number of times of measurement as variables as a function of the sound field spectrum obtained in FIG. 12 for each of 3 times, 9 times, and 15 times.
FIG. 14 is a diagram showing a change in correlation coefficient (m = 0) without frequency shift between the reference sound field and consecutive times measurement values in the sound field spectrum obtained in FIG.
FIG. 15 is a graph showing the maximum value of the cross correlation coefficient considering the frequency shift between the reference sound field and the continuous number of times of measurement in the sound field spectrum obtained in FIG. 12 for each number of times.
FIG. 16 is a graph showing a frequency shift index corresponding to the maximum value of the correlation coefficient for each measurement number of the monitoring mode in order to show how frequency the consecutive times of frequency shifts compared with the reference sound field in the sound field spectrum of FIG. to be.
FIG. 17 is a graph showing a gradual change in the sound field spectrum when a fire situation occurs from the beginning in continuous monitoring mode sound field measurement using a multi-tone sound source with a center frequency of 6 kHz and a frequency interval of 4 Hz.
FIG. 18 is a diagram showing changes in the cross correlation coefficient obtained by using the frequency shift between the reference sound field and the consecutive number of times of measurement as variables as a function of the sound field spectrum obtained in FIG. 17, by three times, nine times, and fifteen times.
FIG. 19 is a diagram showing a change in correlation coefficient (m = 0) between the reference sound field and the continuous count measurement values in the sound field spectrum obtained in FIG.
FIG. 20 is a graph showing the maximum value of the cross correlation coefficient considering the frequency shift obtained between the reference sound field and the continuous number of times of measurement in each sound field spectrum obtained in FIG.
FIG. 21 is a graph showing a frequency shift index corresponding to the maximum value of the cross correlation coefficient for each measurement number of the monitoring mode in order to indicate how much frequency of the consecutive times of the frequency shifts in comparison with the reference sound field in the sound field spectrum of FIG. FIG.
FIG. 22 is a security monitoring flowchart for detecting intrusion, fire, and daily temperature changes based on the correlation coefficient of the sound field spectrum.
FIG. 23 is a block diagram of a device for monitoring a dangerous situation in which an accident of an elderly person or a pet living alone is distinguished from a fire and a usual temperature change based on a correlation coefficient of a sound field spectrum.
FIG. 24 is a flowchart of a dangerous situation monitoring process in which an accident of an elderly person or a pet living alone is distinguished from a fire and a routine temperature change based on a correlation coefficient of a sound field spectrum.
FIG. 25 is a diagram showing a change in the sound field spectrum in a situation where a person keeps moving in a reference sound field and continuous sound field measurement using a multi-tone sound source having a center frequency of 4 kHz and a frequency interval of 4 Hz.
FIG. 26 is a diagram showing correlation coefficients (m = 0) without frequency shift between the reference sound field and consecutive times measurement values in the sound field spectrum obtained in FIG.
FIG. 27 is a diagram showing characteristics of aspects in which a correlation coefficient (m = 0) in which there is no frequency shift between a reference sound field and a continuous count measurement value due to temperature change, intrusion, and movement changes.
FIG. 28 is a security surveillance flow chart for detecting infiltration, movement, fire, and daily temperature change based on the correlation coefficient of the sound field spectrum.

Hereinafter, embodiments of the present invention will be described with reference to the accompanying drawings so that those skilled in the art can easily carry out the technical idea of the present invention.

FIG. 1 is a block diagram of a security monitoring apparatus 100 that detects and detects daily temperature changes such as an intrusion, a fire, a cooling and heating, and a daytime temperature based on a correlation coefficient of a sound field spectrum. Referring to FIG. 1, the security monitoring apparatus 100 includes a sound generating device 110, a sound receiving device 120, and a sound field signal processing device 130.

The sound generator 110 may output a sound wave according to the input voltage in the security monitoring space. Here, the sound wave output from the sound generator 110 may be a multi-tone sound file composed of a linear sum of sine waves having a plurality of frequency components in an audible frequency range of 20 to 20 kHz and an ultrasonic range of 20 kHz or more. Here, the multi-tone sound wave may be in the form of a continuous wave or a pulse wave.

The sound pressure of the sound generator 110 may be set to an optimal size to drive the sound power of the apparatus at the rated power and detect a sound field change caused by a security situation.

The sound receiving apparatus 120 may receive the sound waves in the security monitoring space, and may obtain the sound pressure from the received sound waves. Here, the acoustic receiver 120 may include a frequency conversion filter for converting the received sound waves into the frequency domain.

The sound field signal processing device 130 is an apparatus for discriminating an intrusion or a fire situation using a sound field change in a security monitoring space, and can be implemented through a processor such as a smart device and a digital signal processor (DSP). Sound field values can be represented by sound pressure and phase, and sound pressure and phase can be used individually or in combination. However, in this embodiment, the sound pressure is shown as an example, and the sound pressure level which is the magnitude of the sound pressure is used as the object of signal processing. Here, the sound pressure level can be generally expressed by a logarithmic function, and the value obtained by measuring the sound pressure within the security monitoring space by the sound receiving apparatus 120 can be a sound pressure level. Here, the sound pressure within the security monitoring space is the sound pressure generated by the sound pressure output from the sound generating device 110 spreading into the security monitoring space.

Accordingly, the sound field signal processing apparatus 130 uses the sound pressure P of sound in the preparation mode to calculate the reference sound pressure information (the magnitude of the reference sound pressure Amp = 20 logP) or the phase of the reference sound pressure Ph = ang (P) ) Can be calculated. In this case, the sound field signal processing device 130 can measure the change pattern of the sound field spectrum over time in order to solve the problem of malfunction due to the change in the sound pressure P due to the environmental change such as a gradual change in the temperature and humidity of the air . The sound field signal processing device 130 can analyze the measured sound field spectrum change patterns by time and set the initialization time period of the reference sound field and the security situation determination reference value.

The sound field signal processing apparatus 130 uses the sound transfer function P 'in the monitoring mode to calculate the current sound pressure information (the current sound pressure level Amp = 20 log (P')) or the current sound pressure phase Ph = ang ( P ')), and then compare the reference sound pressure information with the current sound pressure information to determine whether a fire or an intrusion occurs in a security situation.

More specifically, the sound field signal processing device 130 can determine that a dangerous situation such as an intrusion or a fire has occurred through several methods. The conventional method compares the reference deviation (Noise) with the signal value (hereinafter, referred to as 'reference deviation to sound pressure change rate: SNR') and judges that a security situation has occurred when the comparison result is not less than the discrimination reference value. Here, the reference deviation may be a maximum value of the deviation of the frequency-dependent reference sound pressure information, and the signal value may be the absolute value (20log (P ')) between the average of the reference sound pressure information by frequency and the average of the current sound pressure information by frequency, ) -20 log (P)).

In this case, the sound field signal processing device 130 can reset the reference sound field value according to the initialization time period in order to prevent the sound pressure P from changing due to a gradual change in the temperature and humidity of the atmosphere or the like. Such a reset can be performed by calculating the mean and variance of the frequency-specific sound pressure information at the initialization time period intervals in the monitoring mode. The measured sound field itself may be set as the reference sound field if it is not a dangerous situation.

Likewise, as the speed of the sound wave changes due to the temperature change of the air in the fire monitoring room, a change of the sound field occurs, and the sound receiving device installed inside the security monitoring space changes the sound field of the sound wave according to the temperature distribution state You can detect it differently.

1, when the intrusion situation occurs in the security monitoring space, the boundary condition is changed, so that the acoustic transfer function changes and the sound field changes accordingly. Such a sound field change phenomenon occurs in an acoustic space Can be better generated within the body. By detecting the sound field change in this manner, it is possible to detect a fire in a blind spot where an intruder in a blind spot or a fire or smoke is not observed. On the other hand, there may be a dangerous situation mistake due to the sound field change caused by the heating / cooling or the daytime difference. Procedures and procedures are needed to distinguish these situations.

The existing security monitoring method quantifies the degree of change based on the signal to noise ratio (SNR) of the reference sound field, which is obtained by performing a plurality of measurements at the time of measuring the sound field change. This method takes time to measure the initial sound field noise, and also leads to inaccuracies due to a limited number of reference deviations. In addition, when the deviation of the initial sound field is 0, the problem of calculation error must be considered. Furthermore, in the case of a method of obtaining a sound field change at a change value of the average value relative to the reference deviation at each frequency, when the sound pressure at the frequency of the destructive interference is very low, there is a limit that the reference deviation is relatively large and the error becomes large.

Meanwhile, the security monitoring method according to the embodiment of the present invention calculates a correlation coefficient for accurately deriving the similarity between the reference sound field spectrum and the changed current sound field spectrum, and quantifies the degree of the sound field change based on the correlation coefficient Therefore, it is possible to improve the reliability in detecting and distinguishing a dangerous situation.

FIG. 2 is a diagram showing a change in a sound field spectrum when a sudden intrusion state occurs in a continuous sound field measurement with a reference sound field using a multitone sound source having a center frequency of 4 kHz and a frequency interval of 4 Hz. Referring to FIG. 2, there is shown an experimental result obtained by measuring the reference sound field spectrum in the security monitoring space and measuring the sound field spectrum continuously 10 times at a time interval of 8 seconds. 1 to 9 are the sound field spectra before the intrusion, and 10 are the sound field spectra obtained after the intrusion situation occurs.

The multitone sound source used for the sound field measurement has a center frequency of 4 kHz and a frequency interval of 4 Hz, all of which have a frequency of 17 channels. The sound generator 110 generates a sound source for 0.5 seconds. The acoustic receiver 120 receives the generated acoustic signal. The sound field signal processing device 130 frequency-filters the sound signal to obtain a sound field spectrum. Referring back to FIG. 2, the sound field spectrum before intrusion is almost unchanged. However, after the intrusion, the condition is changed by the intruder, and the corresponding sound field spectrum is greatly changed.

FIG. 3 is a diagram showing a cross-correlation coefficient obtained by using a frequency shift between a reference sound field spectrum and each sound field spectrum measured continuously, six times, and ten times as a variable. The cross correlation coefficient can be expressed by the following equation.

, m < 0

In Equation (1), R _i _{, j} is a cross correlation coefficient between the i-th measured sound field S _i and the j-th measured sound field S _j , N is the number of channels of the multitone sound source, Which is the neighboring frequency gap of the multitone sound source.

Specifically, in the case of m = 0, the cross correlation coefficient is obtained by dividing the covariance value of the two sound field spectra without frequency shifting by the product of the standard deviation values of the respective sound field spectra measured at the i-th and j- to be. Also, when m is not 0, the correlation coefficient between the i-th sound field spectrum and the j-th sound field spectrum shifted by m is calculated.

Referring to FIG. 3, the cross correlation coefficients R ₀ _{, 2} (m) and R ₀ _, ₆ (m) between the reference sound field spectrum and the sound field spectrum before the intrusion (2 times, 6 times) are almost similar. However, the cross correlation coefficient R ₀ _, ₁₀ (m) between the reference sound field spectrum and the sound field spectrum after penetration (10 times) is greatly changed.

FIG. 4 is a diagram showing correlation coefficients (m = 0) without frequency shift between the reference sound field and consecutive times measurement values in the sound field spectrum obtained in FIG. Referring to FIG. 4, the correlation coefficient in the case of m = 0 in Equation (1) is close to one before the intrusion. On the other hand, after the invasion, the correlation coefficient decreases rapidly to about 0.91.

FIG. 5 is a graph showing the maximum value of the cross correlation coefficient obtained by considering the frequency shift between the reference sound field and the continuous number of times of measurement in the sound field spectrum obtained in FIG. 2 for each number of times. Referring to FIG. 5, when m is not 0 in Equation (1), the maximum value of the cross correlation coefficient with the reference sound field spectrum is shown for each measurement number in consideration of all cases of frequency shift. For example, all the values are the maximum values only when m = 0, which is the same result as in Fig.

FIG. 6 is a graph showing a frequency shift index corresponding to a maximum value of a correlation coefficient for each measurement number in order to indicate how much the frequency of the consecutive times of frequency shifts compared with the reference sound field in the sound field spectrum of FIG. Referring to FIG. 6, as shown in FIG. 5, when m is 0, all of the frequency shifts are zero because the correlation coefficient is the maximum value.

FIG. 7 is a graph showing a gradual change in the sound field spectrum when a temperature change due to a fire situation occurs from the beginning in continuous sound field measurement using a multi-tone sound source with a center frequency of 4 kHz and a frequency interval of 4 Hz. Referring to FIG. 7, the result obtained by measuring the reference sound field spectrum in the security monitoring space, then constructing an artificial fire situation using an electric heater, and measuring the sound field spectrum continuously 10 times with a time interval of 8 seconds. The sound field spectrum is gradually shifted in the high frequency direction as shown in Fig.

In general, the speed (v) of a sound wave can be expressed as shown in Equation (2) as follows and is proportional to the temperature (T) of the air.

Therefore, even if the frequency f of the sound wave is the same, the wavelength l increases in proportion to the temperature T of the air by the following equations (3) and (4).

As the temperature of the air inside the security space rises, the speed of the sound waves increases. As a result, the wavelength increases proportionally at the same frequency. Since the size of the inside of the security space is fixed, when the temperature is increased, the wavelength of the sound wave must be constant so that the acoustic receiver at the same position has the same sound pressure. As a result, the sound pressure level pattern moves in the high frequency direction without changing its shape. At this time, the change value? F of the moving frequency can be simply expressed as shown in Equation (5).

Since the speed change? V of the sound wave is proportional to the temperature change? T in Equation (2), the frequency change? F is proportional to the frequency of the sound wave and proportional to the temperature change as shown in Equation (6).

It is difficult to simplify the temperature change of the air due to the actual fire due to the increase of the total temperature. The local temperature change around the fire and the overall temperature change are complicated. However, the degree to which the sound pressure level pattern shifts to high frequency due to the rise in temperature can generally monitor changes in the internal air temperature. For example, when the center frequency of the multitone frequency is 4 kHz and the frequency gap is 4 Hz, the temperature change amount? T is expressed by Equation (7) below, which corresponds to 0.57 占 폚 at room temperature (T = 18 占 폚).

FIG. 8 is a diagram showing cross-correlation coefficients obtained by using the reference sound field spectrum and frequency shifts between the respective sound field spectra measured two times, six times, and ten times continuously measured. Referring to FIG. 8, the cross-correlation coefficient gradually changes in comparison with the case of the intrusion shown in FIG. (c), the maximum value of the cross correlation coefficient between the reference sound field spectrum and 10 measurements is shown at m = 1. This is consistent with the results of the sound field spectrum shown in FIG. 7 shifted by about 4 Hz with high frequency.

Fig. 9 is a diagram showing the change in the correlation coefficient when there is no frequency shift of m = 0 as shown in Fig. 4 according to the number of measurements. Referring to FIG. 9, the correlation coefficient is gradually decreased near 1 in comparison with the situation of the intrusion shown in FIG. This means that the sound field spectrum has progressively shifted to high frequency as the temperature rises.

10 is a graph showing the maximum value of the cross correlation coefficient obtained by considering the frequency shift between the reference sound field and the continuous number of times of measurement in each sound field spectrum obtained in FIG. All cases where m = 0, i.e., frequency shifting, as shown in FIG. 5, are considered. The cross correlation coefficient is the maximum value when the high frequency is shifted by 4 Hz in the 10 times measurement shown in FIG. Therefore, although the results of Figs. 9 and 10 are different for the correlation coefficient at 10 times, the difference is not large.

FIG. 11 is a graph showing the frequency shift indexes corresponding to the maximum value of the correlation coefficients for each measurement number in order to indicate how frequency the consecutive times of the frequency shifts compared with the reference sound field in the sound field spectrum of FIG. Referring to FIG. 11, the frequency shift value at which the cross correlation coefficient reaches a maximum value is shifted to 4 Hz high frequency only at 10 times of measurement.

12 to 21, the reference sound field spectrum is measured using a 6 kHz center frequency, which is higher than the center frequency in the security monitoring space, and the 15-time sound field spectrum is continuously measured with a time interval of 8 seconds These are the drawings on the results. In this case, the sound field signal processor 130 generates a sound signal having a frequency interval of 4 Hz and having a frequency of 17 channels, generates a sound source in the sound generator 110 for 0.5 seconds and receives the sound signal received by the sound receiver 120, The sound field spectra obtained by filtering were analyzed.

FIG. 12 is a diagram showing a change in the sound field spectrum when the intruding situation occurs suddenly (15 times) in a continuous sound field measurement with a reference sound field using a multitone sound source with a center frequency of 6 kHz and a frequency interval of 4 Hz. Referring to FIG. 12, it can be seen that the sound field spectrum changes abruptly at the time of 15 times of the occurrence of an intrusion situation.

FIG. 13 is a diagram showing changes in the cross correlation coefficient obtained by using the frequency shift between the reference sound field and the continuous number of times of measurement as variables as a function of the sound field spectrum obtained in FIG. 12 for each of 3 times, 9 times, and 15 times. Referring to FIG. 13, a cross correlation coefficient between the reference sound field spectrum and 3 times, 9 times, and 15 times is shown, and a sudden change occurs at 15 times when an intrusion situation occurs.

FIG. 14 is a diagram showing a change in correlation coefficient (m = 0) without frequency shift between the reference sound field and consecutive times measurement values in the sound field spectrum obtained in FIG. Referring to FIG. 14, the correlation coefficient when m = 0 without a frequency shift is close to 1 before invasion but does not change but decreases sharply at 15 times when an invasion situation occurs and is smaller than 0.1. This indicates that the sound field spectrum in the case of an intrusion situation is completely different from the reference sound field spectrum.

FIG. 15 is a graph showing the maximum value of the cross correlation coefficient considering the frequency shift between the reference sound field and the continuous number of times of measurement in the sound field spectrum obtained in FIG. 12 for each number of times. Referring to FIG. 15, the maximum value of the cross-correlation coefficient obtained by considering all cases where m is not 0 in consideration of frequency shift is a value between 0.5 and 0.6.

FIG. 16 is a graph showing a frequency shift index corresponding to the maximum value of the correlation coefficient for each measurement number of the monitoring mode in order to show how frequency the consecutive times of frequency shifts compared with the reference sound field in the sound field spectrum of FIG. to be. Referring to FIG. 16, the corresponding frequency movement index (m) is 6, and since the cross correlation coefficient is not large even considering the frequency shift, the frequency shift index does not indicate that the actual sound field spectrum shifts in frequency. This is because the frequency movement index does not have a significant meaning unlike the case where the sound field spectrum moves to high frequency.

FIG. 17 is a graph showing a gradual change in the sound field spectrum when a fire situation occurs from the beginning in continuous monitoring mode sound field measurement using a multi-tone sound source with a center frequency of 6 kHz and a frequency interval of 4 Hz. Referring to FIG. 17, it can be seen that the sound field spectrum is gradually shifted to a high frequency from 1 to 15 times when a fire situation occurs from one time.

FIG. 18 is a diagram showing changes in the cross correlation coefficient obtained by using the frequency shift between the reference sound field and the consecutive number of times of measurement as variables as a function of the sound field spectrum obtained in FIG. 17, by three times, nine times, and fifteen times. Referring to FIG. 18, it can be seen that the frequency shift index corresponding to the maximum value gradually increases when the cross-correlation coefficient between the reference sound field spectrum and 3 times, 9 times, and 15 times is seen.

FIG. 19 is a diagram showing a change in correlation coefficient (m = 0) between the reference sound field and the continuous count measurement values in the sound field spectrum obtained in FIG. Referring to FIG. 19, it can be seen that the correlation coefficient when m = 0 without frequency shift is close to 1 at the beginning of the fire situation but gradually decreases with time and decreases to 0.1 to 0.2 when it reaches 15 times.

FIG. 20 is a graph showing the maximum value of the cross correlation coefficient considering the frequency shift obtained between the reference sound field and the continuous number of times of measurement in each sound field spectrum obtained in FIG. Referring to FIG. 20, considering the frequency shift, the maximum value obtained by considering all cases where m is not 0 changes periodically between 0.8 and 1.

FIG. 21 is a graph showing a frequency shift index corresponding to the maximum value of the cross correlation coefficient for each measurement number of the monitoring mode in order to indicate how much frequency of the consecutive times of the frequency shifts in comparison with the reference sound field in the sound field spectrum of FIG. FIG. Referring to FIG. 21, it can be seen that the frequency shift index (m) gradually increases from 0 to 1 to 6, 1 to 7 to 12, and 2 to 13 to 15 times. Since the cross correlation coefficient is close to 1 when the frequency shift is considered, the frequency shift index indicates that the actual sound field spectrum is shifted to the high frequency, so that it can be relied upon that there is a temperature change. The higher the center frequency, the higher the degree of frequency shift. The security monitoring method using the variation pattern of the correlation coefficient between the sound field spectrum implemented through the sound field measurement is as follows.

A security monitoring method according to an embodiment of the present invention calculates a cross correlation coefficient between a reference sound field and a predetermined number of consecutive sound fields in a sound field spectrum obtained through continuous measurement, When the correlation coefficient is monitored and the value becomes smaller than a predetermined reference value smaller than 1, it is determined that a security situation such as an intrusion or a fire has occurred. The correlation coefficient is a criterion that quantifies the correlation that determines how similar the two spectra are. In the case of the discriminant reference value, it may be set to a constant value of 1 or less depending on the environment or the condition. In the embodiment of the present invention, the discrimination reference value may be set to about 0.95, but the present invention is not limited thereto.

In another embodiment, the security monitoring method includes calculating a value obtained by subtracting a correlation coefficient from 1 as an index indicating a degree of difference between mutually different sound field spectra, The occurrence of a dangerous situation can be determined by comparing the two values obtained at the current sound field. A correlation coefficient between the limit value and the real time sound field and the initial sound field can be used as a ratio of the limit value to the critical value by using a correlation coefficient obtained by measuring a certain number of times of the initial number of times of sound field measurement . Similar to the sound field change detection method using the noise field and the signal value, the average and deviation of the index indicating the different degrees of the initial measurement and the index indicating the difference in the real time sound field measurement are compared with each other, A method of discrimination can be applied. Compared to conventional methods, this method can greatly improve its reliability and sensitivity.

The security monitoring method according to another embodiment of the present invention minimizes the problem of erroneous operation of the sound field security that can be caused by the external noise or the electrical noise of the acoustic element and the sound field value can instantaneously have erroneous data, It is possible to add a step of repeating the measurement of the sound field or increasing the size of the sound source to re-measure the sound transfer function to check the change of the sound field due to the occurrence of the dangerous situation. Since the correlation coefficient is based on the relative deviation from the average irrespective of the absolute size of the sound field, the sound source has the same value regardless of the size of the sound source when the sound source is very large. Therefore, even if the sound pressure level of the acoustic receiving element itself is used instead of the acoustic transfer function considering the applied voltage of the acoustic generating element, the same result can be obtained regardless of the size of the sound source.

Since the sound field spectrum can be sensitively changed not only by intrusion, fire, but also by daily temperature changes such as daily drift and heating and cooling, it is difficult to distinguish the security situation by a simple measurement of the correlation coefficient without considering the frequency shift. First, in order to solve this problem, it is necessary to distinguish between a situation in which the sound field changes very suddenly, such as an intrusion, and a situation in which the sound field gradually changes, such as a temperature change such as fire, In the case of intrusion, as shown in FIG. 4, the correlation coefficient is maintained before the intrusion. When the intrusion occurs, the correlation coefficient decreases very rapidly. In the case of fire and ordinary temperature change, Likewise, the correlation coefficient decreases gradually.

The security monitoring method according to the embodiment of the present invention for distinguishing the change pattern of the correlation coefficient according to the time variation is characterized in that the correlation of the predetermined number of times of the interval before the correlation coefficient becomes smaller than the reference value due to intrusion or temperature change The comparison of the mean value (R _a ) of the coefficients and the correlation coefficient (R _b ) at the time of the occurrence of the intrusion or fire hazard situation is used. Equation (8) can be used for a specific and quantitative comparison.

This ratio indicates how many times the difference between the reference sound field and the sound field at the time of occurrence of the security situation occurs, compared to the average difference between the reference sound field and the average sound field before the security situation occurs. The V value in Equation (8) has the meaning of an index indicating how rapidly the sound field changes sharply at the time of detection of the final sound field change.

For example, in the case of intrusion, R _a shown in FIG. 4 is 0.9995 and R _b is 0.9103, so that the index (V) value indicating the degree of abrupt change of the correlation coefficient is 193.37, Since R _a is 0.9774 and R _b is 0.9247, the V value is 3.33, which is very large. Therefore, by setting a constant reference value, the state of the intrusion and the temperature change can be distinguished.

However, in a normal environment, intrusion and temperature change can be mixed, and intrusion, movement, and temperature change of a slight object can not be detected within a predetermined measurement interval. In some cases, these two situations occur sequentially, and sound field changes can be detected. When the sound field change is detected by the movement of a slight object, which is regarded as an intrusion in the state where there is a temperature change, the condition that the change value of the correlation coefficient due to intrusion becomes larger than the specific value of the change value of the correlation coefficient due to fire is approximately 6. Thus, in general, if V is less than 6, it is a fire, and if V is greater than 6, it can be classified as an intrusion. However, the present invention is not limited to this value. In the invasion situation of FIG. 14 and the temperature change situation of FIG. 19, V is 376.61 and 2.63, respectively.

Even if a change in the sound field is detected when a slight change in the sound field is detected due to intrusion or movement of a slight object, since the change in the sound field does not occur above the reference value determined within the security space, V, which is an exponent representing the change of the sound field, is lower than 6. In this case, there is no big problem in the reliability of the security detection system even if the alarm is not triggered because it is classified as a temperature change.

The security monitoring method according to the embodiment of the present invention can prevent the sound field value from being erroneously erroneous due to external noises or electrical noises of the acoustic elements and if the sound field change is detected with a temperature change and a slight intrusion situation In order to minimize the problem of the sound field security malfunction caused by this, and to increase the reliability, a step of re-confirming the sound field change due to the occurrence of a dangerous situation is added by repeating the sound field measurement or increasing the size of the sound source to re- . This improves the accuracy of the intrusion detection and confirms that the situation is safe, and then proceeds to measure the reference sound field again.

In the security monitoring method according to the embodiment of the present invention, when an intruding situation is determined, an alarm indicating that the intruding situation is transmitted, and a risk situation can be coped with by capturing, storing, and transmitting an image of the security space. However, when it is determined that the temperature changes, it is necessary to separate the daily temperature change situations such as fire, daylight, and heating / cooling, so a follow-up process is required. In this case, the temperature change situation can be discriminated by discriminating whether the frequency shift of the sound field spectrum moves to a high frequency or a low frequency and whether a frequency shift is continuously occurring.

And in the case of temperature rise, the frequency movement index moves in the high-frequency direction as shown in Figs. 11 and 21. Fig. In the case of the temperature drop, the frequency shift is shifted in the low frequency direction. In addition, if the temperature rises continuously through repeated measurement of a certain period, it is monitored whether or not the phenomenon continues in the subsequent repetitive measurement, and it is judged whether the temperature increases due to daily or temporary temperature rise due to heating or diurnal change, Can be distinguished.

As described above, the security monitoring method based on the sound field change pattern detection using the correlation coefficient of the sound field spectrum according to the embodiment of the present invention distinguishably detects a security risk situation in an initial state of an intrusion situation or a fire situation, Or alert according to the fire situation. In addition, when the security monitoring system is interlocked with a camera module such as CCTV, the security monitoring system can save shot images related to an intrusion or a fire situation or transmit the captured images to an established destination. Here, the target may be a smart device such as a certain person's automobile wireless remote device, a smart phone and a tablet PC, a guard room server, a security company server, a fire department server, or a police server.

FIG. 22 is a security monitoring flow chart for detecting intrusion, fire, and daily temperature changes based on a correlation coefficient of a sound field spectrum according to an embodiment of the present invention.

FIG. 22 can be implemented by executing the operation of the sound field signal processing device 130 of FIG. 1. Based on the sound field pattern change detection using the correlation coefficient of the sound field spectrum, it is possible to detect an ordinary temperature change such as an intrusion, a fire, Detect the situation separately. Referring to FIG. 22, the security monitoring method according to the embodiment of the present invention is largely divided into a preparation mode and a monitoring mode.

The preparation mode may include an initialization step S2200, a time-domain sound field spectrum measurement step S2210, a time-domain sound field spectrum analysis step S2220, and a security monitoring condition setting step S2230.

The monitoring mode includes a sound field spectrum change measurement step S2300 for measuring a change in the correlation coefficient, a dangerous situation occurrence judgment step S2310, an intrusion / temperature change classification step S2320 through correlation coefficient analysis, S2330), an intrusion image acquisition step S2340, an intrusion alert announcement and information transmission step S2350, a fire / routine temperature change classification step S2360 in which a frequency movement index is analyzed, a fire situation determination step S2370, A step S2380, and a fire alarm announcement and alarm transmission step S2390.

In the initial setting step S2200, the sound generator 110 is operated to output a sound wave into the security monitoring space according to a constant input voltage. In addition, the sound receiving apparatus 120 is operated to receive sound waves in the security monitoring space. The sound field signal processing device 130 measures the sound field spectrum of the reference sound field (sound pressure and phase) information for each frequency provided from the sound receiving device 120. The calculated information is stored in the internal DRAM or flash memory.

In a time-domain sound field spectrum measuring step S2210, the sound field signal processor 130 measures a sound pressure signal according to a time-dependent change in frequency in order to measure a sound field spectrum over time, and outputs the measured result to sound pressure spectrum information Compare.

In the time-domain sound field spectrum analysis step S2220, the sound field signal processing device 130 analyzes the measured sound field spectrum in a time domain, and then stores a correlation coefficient, which is a time domain sound field variation index value.

In the security monitoring condition setting step S2230, the sound field signal processing device 130 sets the security situation occurrence criterion value of the initialization time period and the correlation coefficient by referring to the correlation coefficient value of the stored time.

In the sound field change measurement step (S2300) in the security monitoring mode, the sound field signal processing device 130 measures the current sound pressure spectrum by frequency and calculates a correlation coefficient with the reference sound field spectrum. In this case, the sound field signal processing device 130 can reset the reference sound field spectrum at an initialization time period interval.

As another embodiment, it is also possible to set a measured sound field before a predetermined interval as a reference sound field and move the reference sound field backward one time at a time every time the current real time sound field is measured. The advantage of this approach is that it allows comparison of sound field variations of the same period compared to a pre-determined threshold. In addition, the measured initial sound field is fixed to the reference sound field until the measurement of the sound field of a predetermined number of times is completed, and the method of extending the reference sound field to the virtual preceding sound field is selected within the corresponding interval, If a sound field change occurs before the passage of time, a method of considering it as an unconditional invasion situation may be selected. In general, this interval may be set to an initialization period.

In a dangerous situation occurrence determination step S2310 such as fire or intrusion, the sound field signal processing device 130 compares the current sound field spectrum for each frequency with the reference sound field spectrum for each frequency to determine whether a security situation has occurred. Specifically, when the correlation coefficient between the reference sound field spectrum and the sensed sound field spectrum is smaller than a set reference value, the sound field signal processing device 130 determines that a dangerous situation causing a sound field change has occurred.

In the correlation coefficient analysis step S2320 of the sound field spectrum, when it is determined that a dangerous situation has occurred, the sound field signal processing device 130 analyzes the change of the correlation coefficient with the sound field spectrum of a predetermined section just before the security situation occurs. At this time, the reference sound field spectrum is reset to the initial sound field spectrum before the predetermined interval recovery. Here, it is possible to distinguish whether the change of the correlation coefficient is caused by intrusion or temperature change by classifying whether the change of the correlation coefficient varies rapidly with time or gradually. An exponent value indicating the degree of abrupt change in the correlation coefficient can be used as in the method applied as the embodiment as shown by V in Equation (8).

If it is determined that intrusion is detected in the intrusion state determination step S2330, in order to accurately verify the intrusion image acquisition step S2340, the camera module is operated under the control of the sound field signal processing device 130, Is performed.

In the intrusion alert announcement and information transmission step S2350, the sound field signal processing device 130 may issue an intrusion alert sound or may transmit an intrusion alert to an automobile wireless remote device or the like. In addition, images photographed through the camera module can be transmitted to a server such as a mobile phone, a smart device, a security room, a security company, and a police station through a wired / wireless communication network. In the case of a general automobile without a network function, a wireless remote device such as a remote control may be used to activate or deactivate the alarm function.

In the intrusion discrimination step S2330, it is determined as a temperature change state, and when the frequency shift analysis step S2360 is performed, the frequency shift index is derived by analyzing the change of the cross correlation coefficient according to the frequency shift. If it is in the low frequency direction, the sound field change measurement step S2300 is performed again as a normal state by discriminating the effect of cooling. However, since the fire is suspected in the case of moving from the frequency movement index to the high frequency direction, it is determined whether or not the fire situation occurs in the fire suspicion step (S2370) based on how long the situation lasts for a certain period.

In a specific embodiment, the process of measuring the sound field change (S2300) to the intrusion situation (S2330) and the frequency shift analysis (S2360) to the fire suspicion (S2370) Is detected, the fire situation can be determined.

If it is determined that the fire situation is determined in the fire situation determination step S2370, the camera module is operated under the control of the sound field signal processing device 130 to accurately verify this in the fire image acquisition step S2380, Storage is performed.

In the fire alarm announcement and information transmission step S2390, the sound field signal processing device 130 may issue a fire alarm sound or may transmit a fire alarm to the automobile wireless remote device or the like. In addition, images photographed through the camera module can be transmitted to a server such as a mobile phone, a smart device, a security room, a security company, and a police station through a wired / wireless communication network. In the case of a general automobile without a network function, a wireless remote device such as a remote control may be used to activate or deactivate the alarm function. Each step shown in FIG. 22 may be omitted, if necessary, or added with another process.

In the embodiment, if the analysis interval of the sound field spectrum is set to two or three short term intervals in the correlation coefficient analysis step S2320 of the sound field spectrum, the rapidly changing sound field spectrum can be selectively detected. Most of the intrusions can be detected by comparing the changes of the correlation coefficient with the relatively slow temperature changes without using the rapid change index of Equation 8 defined above, and the intrusion situation is selectively detected even in the most temperature change situations .

In this case, however, detailed changes in the sound field spectrum due to temperature changes can not be detected. Therefore, by performing the calculation process of analyzing and setting the analysis section unit of the sound field spectrum as a comparatively long section, the detection results of the sound field spectrum changes according to the two section conditions are compared with each other, and accordingly, Can be used. In this case, sensitivity of intrusion or fire detection can be adjusted by differentiating the judgment reference value of the change of correlation coefficient in intrusion or fire in short-term section or long-term section. Also, when classified as a temperature change situation, a risk situation can be distinguished by using a process of dividing the temperature rise or fall to analyze the frequency shift index.

The security monitoring apparatus to which such a security monitoring method is applied may be connected to an internet telephone, and may be used as an integrated type or an external type. The above security monitoring methods can be applied to various types of smart devices, for example smart phones, smart TVs, smart cars, smart home appliances including safes or intercoms.

A module with security monitoring function can be installed in a home, office, shop, factory, warehouse or the like which is set as security space, and each module can operate independently or can be connected to each other by wired or wireless connection. It is essential that a pair of the sound generating device and the sound receiving device is configured as a detection module so that the sound field signal processing is integrally formed. However, if it is difficult to securely perform the sound field security monitoring because the security space range is too wide or the structure is complicated A system configuration in which a plurality of sound generating apparatuses and a pair of sound sensing apparatuses are connected by a wired connection or a short-range communication module such as a WiFi is possible with a system operating as a sound field signal processing apparatus as a center.

In a situation in which people are all out of the office and operate in the audible frequency range and are restricted to a designated indoor space or in a sleeping state, by setting a door or a window as a prudent security space, a multitone sound source The noise problem can be solved. In the audible frequency range of 20 ~ 15kHz, there is no blind spot due to structure inside the security space due to the large sound wave wavelength. However, it is possible to perform a wide range of security surveillance. However, in the case of hearing or non- Is possible.

Another security surveillance method according to an embodiment of the present invention detects and stores sound field information in real time in order to monitor the movement of the elderly person living alone or a pet in the house or uses the sound field information in real time to detect falls, A method of transmitting a danger alert of a user to a guardian's smartphone can also be provided. In such a case, a hearing and non-audible frequency region of 15 kHz or more in which humans and animals can not hear well can be used as a sound source. It can be implemented in such a way as to monitor the motion detection of elderly persons or pets alone for a predetermined period of time.

In this case, it is necessary to have a function to distinguish the fire situation by disregarding the intrusion situation or daily temperature change, and to operate a procedure of issuing alarms of falling, stalling, and malfunction incapable when motion detection is not performed for a long time This is possible. 23 and 24 are a conceptual diagram and a flowchart showing such a function.

FIG. 23 is a diagram showing that the sound transfer function changes and the sound field changes according to a change in the boundary conditions due to a situation where the elderly person or the pet moves alone in the security monitoring space. Likewise, changes in the sound field occur as the speed of the sound wave changes due to the occurrence of fire in the security monitoring space, or the temperature change of the air due to the heating / cooling or the daytime difference. It is necessary to detect it and to detect the fire situation because there is a risk of accident, such as falls, stomach, and inability to move when there is no movement for a certain period of time. However, due to the sound field change caused by the heating / Procedures and procedures are also necessary to distinguish these situations, since misidentification can occur.

FIG. 24 is a flowchart of a dangerous situation monitoring process in which an accident of an elderly person or a pet living alone is distinguished from a fire and a routine temperature change based on a correlation coefficient of a sound field spectrum. Referring to FIG. 24, the security monitoring method according to the embodiment of the present invention is divided into a preparation mode and a monitoring mode.

The preparation mode may include an initialization step S2400, a time-domain sound field spectrum measurement step S2410, a time-domain sound field spectrum analysis step S2420, and a security monitoring condition setting step S2430.

The monitoring mode includes a sound field spectrum change measurement step S2500, a sound field change discrimination step S2510, a movement / temperature change discrimination step S2520, a movement state discrimination step S2530, (Step S2540), a fire situation determination step (S2550), a fire image acquisition step (S2560), a fire alarm announcement and alarm transmission step (S2570), and a sound field change discrimination step An accident suspicion step 2580, an accident image acquisition step (S2590), and an accident alert notification and information transfer (S2600).

The steps from the initial setting step (S2400) of the security monitoring preparation mode to the sound field change measuring step (S2500) of the security monitoring mode are the same as in Fig. 22, but the multi- tone sound source used can only use the deaf or the non- It is different. In the case of FIG. 22, sudden infiltration is monitored while there is no movement, but in the case of FIG. 24, the state of continuous motion is monitored.

In the sound field change determination step S2510, the sound field signal processing device 130 compares the current sound field spectrum for each frequency with the reference sound field spectrum for each frequency to determine whether a sound field change has occurred. Specifically, when the correlation coefficient between the reference sound field spectrum and the sensed sound field spectrum is smaller than the set reference value, the sound field signal processing device 130 determines that a sound field change has occurred.

In the correlation coefficient analysis step S2520 of the sound field spectrum, when the sound field change is determined to have occurred, the sound field signal processing device 130 analyzes the change in the correlation coefficient with the sound field spectrum of the predetermined constant interval immediately before the occurrence of the security situation. In this case, however, it is difficult to use an exponential value indicating the degree of abrupt change in the correlation coefficient as shown in Equation (8), since the condition in which human and animal are continuously moving is common. It is possible to distinguish by using the index indicating whether it is continuously decreased or irregularly changing in the time interval. At this time, the reference sound field is reset to the initial sound field of the sound field spectrum before the predetermined interval before the occurrence of the security situation.

As shown in FIG. 22, as another embodiment, it is also possible to set the measured sound field before the predetermined number of times to the reference sound field as a reference sound field, and also to move the reference sound field annually as the current sound field is measured. At this time, a section for analyzing the number of initialization cycles may be set.

FIG. 25 is a diagram showing experimental results obtained by continuously measuring 10 sound field spectra with a time interval of 8 seconds after measuring the reference sound field spectrum in the security monitoring space. FIG. Referring to FIG. 25, the sound field spectrum obtained in a situation where a mannequin assuming one to ten persons moves slowly. The multitone sound source used for the sound field measurement has a center frequency of 4 kHz, a frequency interval of 4 Hz, all of which have a frequency of 17 channels, and a sound source is generated in the sound generator 110 for 0.5 seconds and the sound obtained from the sound receiver 120 The sound field spectrum changes suddenly as a result of the sound field spectrum obtained by performing the frequency filtering of the signal by the sound field signal processing device 130.

26 is a diagram showing a correlation coefficient when m = 0 between the reference sound field spectrum and the sound field spectrum of each measurement number. Referring to FIG. 26, it can be seen that the correlation coefficient rapidly drops below 0 from the initial measurement, and the increase / decrease change irregularly appears depending on the number of measurements. The fact that the correlation coefficient changes around zero means that the sound field spectra are completely different from each other. This may include dynamic characteristics depending on human motion. In the actual security setting situation, the sound field changes from the initial measurement should be detected and used for analysis considering all the sound field spectra stored in the predetermined interval.

FIG. 27 is a diagram showing a representative form of a correlation coefficient (m = 0 without frequency shift) between a reference sound field spectrum and a sound field spectrum of a predetermined section up to a point of occurrence of a sound field change situation. In general, the number of times (M) can be set by the number of times of initialization period of the reference sound field spectrum. A graph indicated by a dotted line indicates a gradual temperature change type, and a graph B indicates a correlation coefficient It shows the form of change. And the C graph is a typical intrusion type, and the D and E graphs are a form of correlation coefficient for a situation where a person is moving continuously.

Summing the absolute value of the difference of correlation coefficient obtained in a continuous measurement from the total period, and the correlation value at the time a field change is detected the value in the 1 (R _b) to restrict the movement index divided by the equation (9) values Respectively. In A, B, and C, which are graphs of rapid temperature change and intrusion, only the decrease of the correlation coefficient occurs continuously. Therefore, this motion index value is close to 1. However, since the correlation coefficient increases and decreases repeatedly in the continuously moving graphs D and E, the motion index value becomes much larger than one.

Therefore, the motion index value MOVE obtained in Equation (9) is calculated, and when the value is equal to or greater than a predetermined value, for example, 2 or more, .

Since invasions are not considered, using this method can distinguish between motion and temperature changes. However, in the present invention, this method is not limited thereto, but various types of methods are applicable.

Actually, the motion index value (MOVE) obtained from the result of FIG. 26 using Equation (9) is 6.8. This is much larger than 2. However, in the case of the temperature change of FIG. 9, since the value of Equation (9) is almost 1, it is possible to distinguish the motion state from the temperature change state by obtaining the motion index value.

When the movement situation is determined in the movement situation determination step S2530, the elderly living alone or the pet animal is still moving in the room, so the process returns to the sound field change measurement step S2500. There may be a situation where the sound field temporarily changes due to external noises or electrical noise of the apparatus. Therefore, as another embodiment of the present invention, a process of confirming a sound field change by measuring a sound field again using the same condition or a larger sound pressure Can be added.

If motion is not detected, it is a sound field change state due to a temperature change. Therefore, in the frequency shift analysis step S2540, a fire and a normal temperature change are distinguished. If the fire is confirmed in the fire suspicion step S2550, Acquires an image at the fire image step S2560, issues a fire alarm, and transmits information (S2570). The image captured through the camera module can be transmitted to a server such as a mobile phone, a smart device, a guard room, or a fire department through a wired / wireless communication network.

If no sound field change is detected in the sound field change determination step S2510, it is determined whether movement has not been detected for the set time interval in the suspicion step S2580 to determine occurrence of an accident of falling, stasis, or inability to move, The sound field signal processing device 130 acquires an accident image, and transmits an alarm announcement and information. The image captured through the camera module can be transmitted to a server such as a smart phone, a smart device, a guard room, or a hospital through a wired / wireless communication network.

In the embodiment of the present invention, in the security monitoring concept and the flow chart of FIG. 1 and FIG. 22, the intruder moves slowly in the security space, or there is a minute movement in the security space 27, when the correlation coefficient of the sound field spectrum is not abruptly changed and the intrusion state can not be detected because it can not be distinguished from the aspect of the temperature change, the motion detection index of Equation (9) The process of distinguishing between change and movement of a person can be added.

For example, by using Equation (8), it is possible to divide the interval between the intrusion and the temperature change, or the interval in which the change in the correlation coefficient is judged into two sections, that is, the short term and the long term, And then the intrusion detection alarm is transmitted in the presence of movement to distinguish the movement from the temperature change using the motion index of Equation (9) in the long-term section, Can be further improved. In particular, in the case of an intruder who intrudes by moving very slowly in anticipation of a loophole in the sound field security method, an intrusion can be detected through such a method.

28 shows a flow chart of security surveillance for detecting intrusion, fire, and daily temperature changes as well as movement in the embodiment of the present invention. 22, a step of analyzing the motion index (S2360) and a step of distinguishing the motion and the temperature change (S2370) are added. If it is determined as a motion, it is regarded as an intrusion, and the process goes to a step of acquiring an intrusion image and issuing an alarm. If it is discriminated as a temperature change, if the fire is discriminated through a step S2390 A step of issuing an alarm is executed.

In addition, it is possible to implement a fire safety monitoring function that selectively ignores the movement of people in a space where people are active and selectively detects the fire situation. The scope of the present invention encompasses all such variations and modifications.

In the embodiment of the present invention, a security monitoring method for classifying the risk situation by analyzing the temporal change pattern of the correlation coefficient of the sound field spectrum is proposed. However, the change of the correlation coefficient by the center frequency before and after the occurrence of the dangerous situation is analyzed It can also be applied to security monitoring methods that distinguish temperature change and intrusion / movement. For example, if the center frequency is set to 1 kHz, 2 kHz, 4 kHz, 6 kHz, and the frequency interval is set to 4 Hz, the sound field spectrum is obtained. The correlation coefficient with the sound field spectrum at the time of occurrence of a situation tends to be proportionally smaller.

However, in the case of intrusion or movement of objects, this proportional relation is not consistently displayed and it shows irregular patterns. Therefore, it is possible to compare the sound field spectrum measured at the current point with the reference sound field spectrum It is possible to distinguish between the intrusion / movement of object and the temperature change by observing only the change of correlation coefficient by frequency. It is also possible to apply the motion index obtained by expressing the motion index of Equation (9) on the frequency axis instead of the measurement time (time) axis. A relatively long time is required to measure a large number of correlation coefficients of the center frequency, and the sound pressure of the partial frequency is low due to a large number of center frequency measurements. However, in the special case, the security monitoring using this method is also selected .

In another embodiment of the present invention, if the analysis interval unit of the sound field spectrum is set to two to three short term intervals as described above, the rapidly changing sound field spectrum can be selectively detected. Most of the intrusion situations can be detected through this configuration, simply by comparing the changes in the correlation coefficient without using the rapid change index in Equation (8) defined above and distinguishing it from the relatively slow temperature change. In the analysis of the long term interval in parallel with the above, it is possible to distinguish the temperature change and the motion and to monitor the intrusion or motion detection not occurring for a certain period of time, so that the elderly living alone in the security space can observe fall, A function for detecting a situation can be implemented. In this case, the judgment reference value of the correlation coefficient for determining the intrusion / movement and the temperature change in the short-term and long-term sections can be appropriately selected in accordance with the environment.

The security monitoring apparatus to which the security monitoring method is applied may be connected to an internet telephone, and may be used as an integrated type or an external type. A security monitoring method according to an embodiment of the present invention can be applied to smart appliances including various kinds of smart devices, for example, smart phones, smart TVs, and intercoms.

The security monitoring methods using the variation pattern of the correlation coefficients of the sound field spectrum according to the embodiment of the present invention do not necessarily require hardware changes of existing Internet telephones or smart devices. That is, if only related algorithms are built in the internal processor, it is possible to use interlocking.

The security information detected according to the embodiment of the present invention can be transmitted to various smart devices connected to the network as multimedia information such as characters and images. Furthermore, when a user of a smart phone or a smart device accesses the related security system in the form of an app (App), it is possible to provide various security related services.

While the invention has been shown and described with reference to certain preferred embodiments thereof, it will be understood by those skilled in the art that various changes and modifications may be made therein without departing from the spirit and scope of the invention. Therefore, the scope of the present invention should not be limited to the above-described embodiments, but should be determined by the equivalents of the claims of the present invention as well as the claims of the following.

110: Sound Generating Device
120: Acoustic receiver
130: Sound field signal processing device
100: Security monitoring device

Claims

A security monitoring method for a security monitoring apparatus comprising:
Outputting a multi-tone sound wave composed of a linear sum of sine waves having a plurality of frequency components into a security monitoring space;
Receiving the multi-tone sound wave and calculating a sound field;
Calculating and storing frequency-specific sound field information through the sound field;
Comparing the frequency-dependent reference sound field information with the calculated sound field information to determine whether a sound field change has occurred;
And analyzing whether or not the sound field changes are collected for a certain period of time to classify at least two of the intrusion, movement, and temperature change situations based on the correlation between the reference sound field spectrum and the continuous sound field spectrum. Way.

The method according to claim 1,
Wherein the correlation is obtained by calculating a correlation coefficient value between the reference sound field spectrum and the continuous sound field spectrum.

The method according to claim 1,
Wherein the correlation is obtained using a correlation coefficient calculated by dividing the covariance value of the reference sound field spectrum and the covariance value of the continuous sound field spectrum by a product of standard deviation of each sound field spectrum.

The method according to claim 1,
Comparing the reference sound field spectrum with a current sound field spectrum to determine whether a dangerous situation causing the sound field change occurs;
Analyzing the sound field spectrum collected for a certain period of time before the dangerous situation to distinguish intrusion, temperature change, or motion situation; And
Further comprising the step of classifying a temperature change state of intrusion, movement, fire, and daily diurnal change and cooling / heating based on a correlation coefficient between the reference sound field spectrum and the continuous sound field spectrum.

5. The method of claim 4,
And comparing the correlation coefficient between the reference sound field spectrum and the current sound field spectrum with a set reference value to determine whether a dangerous situation has occurred.

5. The method of claim 4,
The correlation coefficient between the reference sound field spectrum and the current sound field spectrum is used and an initial correlation coefficient at the reference sound field measurement as an index indicating the degree of difference between the sound field spectrums is set to 1 at a limit value and a correlation coefficient between the current sound field and the reference sound field And the limit value is compared with 1 to determine whether or not a dangerous situation occurs.

The method according to claim 1,
Using a variation pattern of the correlation coefficient between the reference sound field spectrum and the continuous sound field spectrum,
A sudden decrease immediately before the occurrence of a dangerous situation is determined as an intrusion,
Wherein the progressive decrease immediately before the occurrence of the risk event is determined as a temperature change situation.

The method according to claim 1,
Wherein a mean value of correlation coefficients between the reference sound field spectrum and the continuous sound field spectrum before the occurrence of a dangerous situation is set to 1 and a correlation value between the sound field spectrum and the reference sound field spectrum at the time of occurrence of the dangerous situation 1, the ratio of the limit value is compared, thereby distinguishing the intrusion from the temperature change situation.

The method according to claim 1,
Analyzing a temporal variation pattern of a correlation coefficient between the reference sound field spectrum and the continuous sound field spectrum,
It is determined that the above-described change pattern of irregular increase / decrease is motion,
Wherein the suddenly decreasing change pattern is identified as an intrusion situation.

The method according to claim 1,
Comparing the reference sound field spectrum with a current sound field spectrum to determine whether a sound field change has occurred; And analyzing the sound field spectrum of each frequency collected for a certain period of time before the sound field change situation occurs to distinguish the movement of the person / animal, the fire, and the daily temperature change of the day and night, Monitoring method.

11. The method of claim 10,
And comparing the correlation coefficient between the reference sound field spectrum and the current sound field spectrum with a set reference value to determine whether a sound field change situation has occurred.

The method according to claim 1,
Analyzing a temporal variation pattern of a correlation coefficient between the reference sound field spectrum and the continuous sound field spectrum,
When the variation pattern is irregular, it is determined to be a motion,
Wherein the temperature change state is determined to be a temperature change state when the change state gradually decreases gradually.

The method according to claim 1,
Using the correlation coefficient between the reference sound field spectrum and the continuous sound field spectrum for a predetermined period before the sound field change situation occurs,
A ratio of a value obtained by summing the absolute values of the differences of the front and rear correlation coefficients obtained by the continuous measurement and a value obtained by subtracting the correlation coefficient value at the point of time when the sound field change is detected from 1 is added to the motion index,
And a motion and a temperature change are distinguished by using the motion index.

The method according to claim 1,
An index indicating a frequency shift of a sound field spectrum is derived based on a correlation coefficient obtained by using a multitone frequency between the reference sound field spectrum and the continuous sound field spectrum as a variable,
And a temperature monitoring unit for monitoring a temperature of the air conditioner,

11. The method of claim 10,
Detecting and storing motion information of a person and an animal inside the security space using sound field information; And
And transmitting the detection information to the smartphone and the smart device of the guardian.

11. The method of claim 10,
Further comprising the step of transmitting alarms of falls, fainting and non-malfunctioning accidents and transmitting security information when the movement of persons and animals inside the security space is not detected for a predetermined period of time.

11. The method of claim 10,
Further comprising selectively sensing only a fire situation, communicating a fire alarm, and transmitting security information in the presence of movement of a person or an animal inside the secure space.

The method according to claim 1,
And performing image capturing to store the image information and verify the situation when a security situation occurs.

The method according to claim 1,
The security surveillance method is a security surveillance method linked to a smart home appliance including a security camera having a network function, an Internet phone, a smart TV, and an interphone

The method according to claim 1,
Security monitoring method implemented by software without adding hardware of interlocking device when interworking.

21. The method of claim 20,
Wherein the security monitoring method is executed by remote control or transmits obtained security information when a program in the form of an app related to a user's smartphone or smart device is executed.

An acoustic generator for outputting a sound wave according to an input voltage within a security monitoring space;
A sound wave receiving device for receiving the sound wave and calculating a sound field using the sound wave; And
Calculating cross-correlation coefficients between the continuous sound field spectral information and the reference sound field spectral information by continuously measuring sound field spectrum information for the sound field, calculating cross-correlation coefficients between the continuous sound field spectrum information and the reference sound field spectral information through the cross- And a sound field signal processing device for distinguishing at least two situations.

23. The method of claim 22,
Wherein the sound wave is a multi-tone sound wave having a linear sum of sine waves having a plurality of frequency components.

23. The method of claim 22,
And a memory for storing the reference sound field spectrum information.

23. The method of claim 22,
Wherein the sound field signal processing apparatus calculates the sound pressure or phase of the sound field using an acoustic transfer function.