JP2012155603A - Intrusion detection device, intrusion detection method and intrusion detection program - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide an intrusion detection device, an intrusion detection method and an intrusion detection program which can detect a human action with high accuracy.SOLUTION: An intrusion detection device 101 includes a space feature amount calculation unit 1 for calculating a space feature amount showing the state of a prescribed area, a change rate calculation unit 2 for calculating a temporal change of the space feature amount calculated by the space feature amount calculation unit 1, and a detection unit 11 for detecting a human action in the prescribed area based on the temporal change calculated by the change rate calculation unit 2.

Description

  The present invention relates to an intrusion detection device, an intrusion detection method, and an intrusion detection program, and more particularly, to an intrusion detection device, an intrusion detection method, and an intrusion detection program that detect human actions using spatial feature amounts.

  Intrusion detection devices that detect human movements in a predetermined area such as a room have been developed. As an example of the intrusion detection method, for example, “Study on indoor intruder detection by UWB-IR” Teijihashi Teijih et al., IEICE Transactions B, Vol. J90-B, No. 1, pp.97-100, 2007 On January 1, 2000 (Non-Patent Document 1), a method using a propagation delay profile, that is, a power delay profile by UWB-IR (Ultra Wide Band-Impulse Radio) is disclosed.

  However, in the method described in Non-Patent Document 1, interference with other wireless services becomes a problem because a wideband signal is used, and because the received signal power is used, it is affected by indoor multipath fading, The detection accuracy may deteriorate.

  As a technique for solving such a problem, for example, Japanese Patent Laid-Open No. 2008-216152 (Patent Document 1) discloses the following configuration. That is, the event detection device calculates an eigenvector, that is, an arrival angle distribution based on the received signal of each array antenna, and calculates an inner product value of the eigenvector and a normal eigenvector as a comparison reference. Then, the event detection device detects an event, that is, an intruder, based on a comparison result between the inner product value and a predetermined threshold value.

"Study on indoor intruder detection by UWB-IR" Keiji Terasaka et al., IEICE Transactions B, Vol. J90-B, No. 1, pp.97-100, January 1, 2007

JP 2008-216152 A

  Here, it can be easily estimated that the inner product of the eigenvectors calculated in the configuration described in Patent Document 1 changes depending on factors other than the target intruder.

  For example, it can be easily assumed that operations of small animals such as mice and cats, and various devices such as electric fans and air conditioners change the radio wave characteristics of the space and affect the arrival angle distribution.

  In Patent Document 1, various event detection is performed. In particular, in an application for detecting a human motion, a variation other than the configuration described in Patent Document 1 is excluded, and variations other than the target intruder are excluded. A technique for suppressing detection is desired.

  This invention was made in order to solve the above-mentioned subject, The objective is to provide the intrusion detection apparatus, the intrusion detection method, and the intrusion detection program which can detect a human operation | movement with high precision. is there.

  (1) In order to solve the above-described problem, an intrusion detection apparatus according to an aspect of the present invention includes a spatial feature amount calculation unit for calculating a spatial feature amount indicating a state of a predetermined area, and the spatial feature amount calculation unit. Based on the temporal change calculated by the change rate calculating unit and the temporal change calculated by the change rate calculating unit, human motion in the predetermined area is detected. And a detecting unit for

  In this way, by detecting the human motion using the temporal change of the spatial feature value instead of the magnitude of the variation amount of the spatial feature value, fluctuation due to factors other than human motion is eliminated, and the detection accuracy Can be improved.

  (2) Preferably, the detection unit determines that a human motion has occurred in the predetermined area when the temporal change is within a predetermined range corresponding to the human motion.

  With such a configuration, it is possible to perform a detection operation focusing on the fact that the speed of the change in the state of the space caused by the human movement is within a predetermined range, so that the detection accuracy can be improved.

  (3) More preferably, when the state where the temporal change is within the predetermined range continues for a predetermined time or more, the detection unit determines that a human motion has occurred in the predetermined area.

  With such a configuration, it is possible to perform a detection operation that focuses on the fact that the change in the state of the space due to human movement continues to some extent, so that the detection accuracy can be improved.

  (4) Preferably, the spatial feature value calculation unit outputs a signal indicating the calculated spatial feature value, and the change rate calculation unit receives the spatial feature value calculation unit received from the spatial feature value calculation unit as the temporal change. The frequency component of the signal is calculated, and the detection unit detects a human motion in the predetermined area based on the frequency component calculated by the change rate calculation unit.

  As described above, the temporal change of the spatial feature amount can be appropriately evaluated by the configuration for calculating the frequency component of the spatial feature amount as the temporal change of the spatial feature amount.

  (5) More preferably, for the target frequency band corresponding to human motion, the change rate calculation unit calculates a component outside the target frequency band among the frequency components of the signal received from the spatial feature amount calculation unit. The detection unit includes a target filter for attenuation, and detects a human motion in the predetermined area based on the signal that has passed through the target filter.

  As described above, the detection accuracy can be further improved by the configuration in which the range of variation of the spatial feature value to be detected is set to a range unique to the object, specifically, a range corresponding to human intrusion behavior. In addition, a configuration in which a bandpass filter is provided instead of a configuration in which Fourier transform is performed eliminates the need for signal buffering, and can increase detection time.

  (6) More preferably, when the level of the signal that has passed through the target filter is equal to or greater than a predetermined threshold, the detection unit determines that a human motion has occurred in the predetermined area.

  As described above, with the configuration in which the determination using the threshold is performed, it is possible to detect the presence or absence of a human action with a simple process.

  (7) More preferably, the change rate calculation unit further includes the non-target frequency band of the signal received from the spatial feature amount calculation unit for the non-target frequency band that is a band other than the target frequency band. Including one or a plurality of non-target filters for attenuating components outside the target frequency band, wherein the detection unit is based on the signal that has passed through the target filter and the signal that has passed through each of the non-target filters. Detect human movements in a given area.

  In this way, the configuration in which not only the band filter corresponding to the human operation but also the other band filter is provided, the change rate becomes a certain level in other bands in addition to the band corresponding to the human operation. In this case, it is possible to accurately detect the presence / absence of a human motion by comparing the level of the change rate in each band.

  (8) More preferably, the detection unit in the predetermined area based on which signal has the maximum level among the signal that has passed through the target filter and the signal that has passed through each of the non-target filters. Detect human movements.

  As described above, with the configuration for detecting the maximum level, it is possible to detect the presence or absence of a human action with a simple comparison process.

  (9) More preferably, when the signal that has passed through the target filter has a maximum level and the level of the signal that has passed through the target filter is equal to or greater than a predetermined threshold, the detection unit It is determined that a human motion has occurred in the area.

  As described above, the presence / absence of human motion can be detected more accurately by the configuration in which the determination using the threshold value is performed in addition to the comparison processing.

  (10) More preferably, the change rate calculation unit performs a Fourier transform on the signal received from the spatial feature amount calculation unit, and an integration in a frequency band corresponding to a human motion is obtained in the result of the Fourier transform. The value is calculated, and the detection unit detects a human motion in the predetermined area based on the integral value calculated by the change rate calculation unit.

  With such a configuration, it is possible to evaluate the temporal change of the spatial feature amount more accurately than the configuration using the bandpass filter.

  (11) More preferably, the detection unit determines that a human motion has occurred in the predetermined area when the integral value is equal to or greater than a predetermined threshold.

  As described above, with the configuration in which the determination using the threshold is performed, it is possible to detect the presence or absence of a human action with a simple process.

  (12) More preferably, the predetermined range is a range of 1 Hz or less.

  As described above, the configuration in which the range of variation of the spatial feature to be detected is set to a range corresponding to a typical human intrusion behavior can improve detection accuracy without fine adjustment. Become.

  (13) More preferably, the predetermined range is a range not less than 1 Hz and not less than 1 Hz and not more than 1 Hz.

  As described above, the detection accuracy can be further improved by removing the low-frequency fluctuation component that is often caused by another environmental factor.

  (14) More preferably, the predetermined range can be set and changed.

  With such a configuration, it is possible to set the range of fluctuation of the spatial feature value to be detected according to the radio wave propagation environment of the target space, and it is possible to perform detection operations specialized for various environments.

  (15) Preferably, the intrusion detection device further includes a regular fluctuation component removing unit for removing a regular fluctuation component from the spatial feature quantity calculated by the spatial feature quantity calculation unit, The change rate calculation unit calculates a temporal change in the spatial feature amount after being removed by the regular fluctuation component removal unit.

  This makes it possible to remove fluctuation components that are much more dominant than fluctuations due to intruders, such as regular fluctuations in spatial feature amounts due to some device, and can stably obtain high detection accuracy.

  (16) In order to solve the above problem, an intrusion detection method according to an aspect of the present invention includes a step of calculating a spatial feature amount indicating a state of a predetermined area, and calculating a temporal change of the calculated spatial feature amount. And detecting a human motion in the predetermined area based on the calculated temporal change.

  In this way, by detecting the human motion using the temporal change of the spatial feature value instead of the magnitude of the variation amount of the spatial feature value, fluctuation due to factors other than human motion is eliminated, and the detection accuracy Can be improved.

  (17) In order to solve the above-described problem, an intrusion detection program according to an aspect of the present invention includes a step of calculating a spatial feature amount indicating a state of a predetermined area in a computer, and a time of the calculated spatial feature amount. A program for executing a step of calculating a change and a step of detecting a human motion in the predetermined area based on the calculated temporal change.

  In this way, by detecting the human motion using the temporal change of the spatial feature value instead of the magnitude of the variation amount of the spatial feature value, fluctuation due to factors other than human motion is eliminated, and the detection accuracy Can be improved.

  According to the present invention, human motion can be detected with high accuracy.

It is a figure which shows the structure of the intrusion detection apparatus which concerns on the 1st Embodiment of this invention. It is the flowchart which defined the operation | movement procedure at the time of the intrusion detection apparatus which concerns on the 1st Embodiment of this invention performing intrusion detection operation | movement. It is a figure which shows the experiment environment and experiment content using the intrusion detection apparatus which concerns on the 1st Embodiment of this invention. It is a figure which shows the inner product value obtained by the experiment using the intrusion detection apparatus which concerns on the 1st Embodiment of this invention. It is a figure which shows the inner product value obtained by the experiment using the intrusion detection apparatus which concerns on the 1st Embodiment of this invention. FIG. 6 is a diagram illustrating a calculation result of a change rate in case 1. FIG. 10 is a diagram illustrating a calculation result of a change rate in case 2. It is a figure which shows the structure of the change rate calculation part in the intrusion detection apparatus which concerns on the 1st Embodiment of this invention. It is a figure which shows the structure of the modification of the change rate calculation part in the intrusion detection apparatus which concerns on the 1st Embodiment of this invention. It is a figure which shows the structure of the change rate calculation part in the intrusion detection apparatus which concerns on the 2nd Embodiment of this invention. It is a figure which shows the structure of the comb filter in the intrusion detection apparatus which concerns on the 2nd Embodiment of this invention. It is a figure which shows the structure of the intrusion detection apparatus which concerns on the 3rd Embodiment of this invention. It is the flowchart which defined the operation | movement procedure at the time of the intrusion detection apparatus which concerns on the 3rd Embodiment of this invention performing intrusion detection operation | movement. It is a figure which shows the structure of the intrusion detection apparatus which concerns on the 4th Embodiment of this invention.

  Hereinafter, a first embodiment of the present invention will be described with reference to the drawings. In the drawings, the same or corresponding parts are denoted by the same reference numerals and description thereof will not be repeated.

<First Embodiment>
[Configuration and basic operation]
FIG. 1 is a diagram showing a configuration of an intrusion detection apparatus according to the first embodiment of the present invention.

  The intrusion detection apparatus 101 detects a human motion in the area by monitoring a change in a spatial feature amount indicating a state of a predetermined area such as a room. For example, the intrusion detection device 101 monitors the state in a predetermined area based on the nature of wave propagation such as reflection and diffraction. Specifically, the intrusion detection apparatus 101 detects human movements by monitoring arrival angle distributions calculated based on reception signals of a plurality of antennas provided in a predetermined area.

  With reference to FIG. 1, the intrusion detection apparatus 101 includes a spatial feature amount calculation unit 1, a change rate calculation unit 2, and a detection unit 11. The detection unit 11 includes a change rate determination unit 3 and a change duration calculation unit 4.

  The spatial feature amount calculation unit 1 calculates a spatial feature amount indicating a state of the predetermined area for a predetermined area where a human motion is to be detected.

  The change rate calculation unit 2 calculates a temporal change in the spatial feature amount calculated by the spatial feature amount calculation unit 1.

  The detection unit 11 detects a human motion in a predetermined area based on the temporal change calculated by the change rate calculation unit 2.

More specifically, first, the spatial feature quantity calculation unit 1 extracts a spatial feature quantity using the arrival angle distribution, for example, similarly to the configuration described in Patent Document 1. That is, the spatial feature quantity calculation unit 1 extracts the spatial feature quantity P (t) by calculating the inner product of the eigenvectors. This inner product indicates the amount of change from the initial vector serving as a comparison reference. Assuming that the initial vector, that is, the eigenvector when there is no intruder is vno and the eigenvector at the observation time t is vob (t), the spatial feature P (t) is expressed by the following equation.
P (t) = vno · vob (t)

  Next, the change rate calculation unit 2 calculates the change rate of the spatial feature amount. In other words, the change rate of the spatial feature value is an amount indicating a temporal change of the spatial feature value, for example, a periodic change. The change rate calculation unit 2 calculates and outputs the change rate of the spatial feature amount every 0.1 seconds, for example.

  If the spatial feature value appropriately represents the state quantity of the predetermined space, it is considered that the change rate of the spatial feature value has a characteristic specific to the target. For example, since small animals move faster than humans, the change rate of small animals is greater than the change rate of humans.

  In the intrusion detection apparatus according to the first embodiment of the present invention, a detection operation focusing on the following points (1) and (2) is performed in consideration of human intrusion behavior and space state change. That is, (1) The speed of the change in the state of the space due to human movement falls within a predetermined range. For example, the change in the state of the space by birds is very instantaneous. (2) The change in the state of the space due to human movement continues to some extent. That is, the intruder needs to stay in a predetermined area for a certain time after the intrusion in order to secure the object.

  Specifically, the detection unit 11 determines that a human motion has occurred in a predetermined area when the temporal change in the spatial feature amount falls within a predetermined range corresponding to the human motion. In addition to this condition, the detection unit 11 may determine that a human motion has occurred in a predetermined area when the state in which the temporal change is within a predetermined range continues for a predetermined time or longer.

  FIG. 2 is a flowchart defining an operation procedure when the intrusion detection device according to the first embodiment of the present invention performs an intrusion detection operation. The intrusion detection apparatus 101 reads out and executes each step of the flowchart from a memory (not shown). This program can be installed externally. The installed program is distributed in a state stored in a recording medium, for example.

  Referring to FIG. 2, first, the spatial feature quantity calculation unit 1 calculates a spatial feature quantity indicating the state of a predetermined area (step S1).

  Next, the change rate calculation unit 2 calculates a temporal change in the calculated spatial feature amount (step S2).

  Next, the detection unit 11 detects a human motion in a predetermined area based on the calculated temporal change.

  More specifically, when the change rate calculated by the change rate calculation unit 2 is not within a predetermined detection target range corresponding to a human motion (NO in step S3), the detection unit 11 measures the measured space. It is determined that the variation in the feature amount is not a variation due to human movement, and it is determined that there is no intruder (step S6). This is a determination based on the point (1).

  In addition, when the state in which the change rate calculated by the change rate calculation unit 2 is within the detection target range is not continuous (NO in step S4), the detection unit 11 determines that the variation in the measured spatial feature amount is It is determined that the change is not caused by human movement, and it is determined that there is no intruder (step S6). This is a determination based on the point (2).

  On the other hand, the detection unit 11 is measured when the state in which the change rate calculated by the change rate calculation unit 2 is within the detection target range is continuous (YES in step S3 and YES in step S4). It is determined that the change in the spatial feature amount is a change due to human action, and it is determined that there is an intruder (step S5).

  FIG. 3 is a diagram showing an experimental environment and experimental contents using the intrusion detection device according to the first embodiment of the present invention.

  With reference to FIG. 3, using the intrusion detection apparatus 101, an experiment was performed in a room as shown in FIG. Specifically, a transmitter and a receiver were provided in the room, and the fluctuation of the inner product value of the eigenvector when the subject moved in the room according to the order of arrows 1 to 4 in the figure was measured. Here, in the calculation of the inner product value of the eigenvector, the eigenvector in a state before the subject enters the room is set as the initial vector vno.

  FIG. 4 is a diagram illustrating inner product values obtained by experiments using the intrusion detection device according to the first embodiment of the present invention.

  FIG. 4 shows the variation of the inner product value when the subject who has entered the room moves at a normal walking speed (hereinafter referred to as case 1).

  FIG. 5 is a diagram showing inner product values obtained by experiments using the intrusion detection device according to the first embodiment of the present invention.

  FIG. 5 shows the variation of the inner product value when the subject entering the room moves at a speed about twice that of Case 1 (hereinafter referred to as Case 2).

  Here, for example, a section with a large fluctuation indicated by peak A in FIG. 4 corresponds to the timing at which the subject passes near the receiver or the transmitter. When the subject passes through such a position, the magnitude of fluctuation, that is, the inner product value, changes greatly. For this reason, it can be said that it is generally difficult to identify an object based on the fluctuation amount itself.

  Therefore, the intrusion detection device 101 does not use the amount of variation of the spatial feature amount as in the technique described in Patent Document 1, but uses the temporal change of the variation amount.

  FIG. 6 is a diagram illustrating a calculation result of the change rate in case 1. FIG. 7 is a diagram illustrating the calculation result of the change rate in Case 2. In FIG.

  FIG. 6 and FIG. 7 show the frequency analysis result of the fluctuation of the inner product value, that is, the result of the discrete Fourier transform performed on the inner product value, and this result shows the change rate.

  Normally, when there is human intrusion behavior, as shown in FIG. 6 and FIG. 7, the fluctuation component of the inner product value is concentrated in the range of 1 Hz or less and has a higher density than the range of 1 Hz or more.

  In addition, since human behavior is generally irregular, the characteristics of the change rate do not have a peak at a specific frequency but are distributed over a certain frequency band.

  Based on the result of this experiment, the intrusion detection apparatus 101 employs a criterion of determining that there is an intruder when the change in the spatial feature amount corresponds to a change rate that may be caused by the intruder.

  Referring to FIG. 1 again, for example, the spatial feature amount calculation unit 1 outputs a signal indicating the calculated spatial feature amount to the change rate calculation unit 2.

  The change rate calculation unit 2 calculates the frequency component of the signal received from the spatial feature amount calculation unit 1 as a temporal change in the spatial feature amount.

  The detection unit 11 detects a human motion in a predetermined area based on the frequency component calculated by the change rate calculation unit 2.

  FIG. 8 is a diagram showing a configuration of a change rate calculation unit in the intrusion detection device according to the first embodiment of the present invention.

  Referring to FIG. 8, change rate calculation unit 2 includes bandpass filters 21 to 24, power calculation units 25 to 28, threshold value determination units 29 to 32, and maximum value determination unit 33.

  The band pass filters 21 to 24 are differential filters that pass, for example, a change rate within a predetermined range. That is, the band pass filter 21 is a target frequency band corresponding to a human action, for example, other than 0.1 to 1.0 Hz, among the frequency components of the signal indicating the feature quantity P (t) received from the spatial feature quantity calculation unit 1. Is attenuated.

  The band pass filters 22 to 24 are non-target frequency bands that are bands other than the target frequency band among the frequency components of the signal indicating the feature quantity P (t) received from the spatial feature quantity calculation unit 1, for example, 1.0. Attenuate components other than ~ 2.0 Hz, 2.0-3.0 Hz, and 3.0-4.0 Hz.

  In addition, the pass band of each band pass filter may not be a fixed width | variety, and may mutually overlap. By overlapping the passbands of the bandpass filters with each other, it is possible to increase the resolution for detecting the variation of the spatial feature amount.

  The detection unit 11 detects a human motion in a predetermined area based on the signal that has passed through the band pass filter 21 and the signal that has passed through the band pass filters 22 to 24. For example, the detection unit 11 performs a human operation in a predetermined area based on which signal has the maximum level among the signal that has passed through the band pass filter 21 and the signal that has passed through each of the band pass filters 22 to 24. Is detected. More specifically, the detection unit 11 has a predetermined area when the signal that has passed through the band-pass filter 21 has the maximum level and the level of the signal that has passed through the band-pass filter 21 is equal to or greater than a predetermined threshold. It is determined that a human motion has occurred.

  More specifically, the power calculation units 25 to 28 calculate the power of the signals that have passed through the band pass filters 22 to 24. Note that the power calculation units 25 to 28 are not limited to power, and may calculate the absolute values of these signals, or may calculate the squares of these signals.

  Here, the graphs shown in FIGS. 6 and 7 correspond to signals output from the power calculators 25 to 28. Also, the graphs shown in FIGS. 4 and 5 correspond to signals output from the spatial feature quantity calculation unit 1.

  The threshold determination units 29 to 32 compare the power calculated by the power calculation units 25 to 28 with a predetermined threshold, and output the comparison result and the power to the maximum value determination unit 33. It is desirable to estimate and set an appropriate value for this threshold in an environment without intruders.

  Here, the detection unit 11 determines that there is no intruder when there is no signal exceeding a predetermined threshold in the threshold determination units 29 to 32 among the output signals of the band pass filters 21 to 24.

  The maximum value determination unit 33 selects a signal having the maximum power value from among signals output from the bandpass filters 21 to 24 that have exceeded a predetermined threshold in the threshold determination units 29 to 32. Then, the maximum value determination unit 33 outputs a signal indicating a change rate corresponding to the selected signal to the maximum value determination unit 33 in the detection unit 11. For example, the maximum value determination unit 33 outputs a signal indicating 0.1 to 1.0 Hz when the pass signal of the band pass filter 21 is selected.

  In the detection unit 11, the change rate determination unit 3 determines whether or not the change rate extracted by the change rate calculation unit 2 is included in the detection target range. The change rate determination unit 3 determines that no intruder exists when the change rate is outside the detection target range.

  This detection target range is desirably 1 Hz or less based on the above experimental results. Furthermore, since a fluctuation component having a frequency that is too low is likely to be caused by another environmental factor, it is desirable to remove a component of 0.1 or less that is close to the direct current component. In addition, it is desirable that the high frequency region is set to about 2 Hz with some margin. In this case, the pass band of the band pass filter 21 is set to 0.1 to 2.0 Hz.

  Here, the change rate also depends on the radio wave propagation environment. In the intrusion detection apparatus 101, the setting of the detection target range can be changed. This makes it possible to adjust the detection target range according to the propagation environment in the target space for intrusion detection.

  Specifically, before actually operating the intrusion detection apparatus 101, a human moves in the room in a simulated manner as shown in the experiment shown in FIG. 3 to determine in advance how much the fluctuation rate is at a normal walking speed. Keep measuring. And a detection object range is determined using this measurement result.

  Based on the determination result of the change rate determination unit 3, the change duration calculation unit 4 monitors whether or not the state in which the change rate is included in the detection target range continues for a predetermined time.

  This monitoring time is preferably set to 2 seconds or more, for example. This is because a series of intrusion actions such as intrusion, coloration of the room, and escape do not occur instantaneously but are performed over a predetermined time. Thereby, it is possible to perform determination with higher accuracy.

  FIG. 9 is a diagram showing a configuration of a modification of the change rate calculation unit in the intrusion detection device according to the first embodiment of the present invention.

  Referring to FIG. 9, this change rate calculation unit 2 includes a DFT (Discrete Fourier Transform) calculation unit 81, integration calculation units 82 to 85, threshold value determination units 29 to 32, and a maximum value determination unit 33. .

  The change rate calculation unit 2 performs a Fourier transform on the signal received from the spatial feature amount calculation unit 1, and calculates an integral value in a frequency band corresponding to human movement in the result of the Fourier transform.

  The detection unit 11 detects a human motion in a predetermined area based on the integral value calculated by the change rate calculation unit 2. For example, when the integration value is equal to or greater than a predetermined threshold, the detection unit 11 determines that a human motion has occurred in a predetermined area.

  More specifically, the DFT calculation unit 81 performs DFT calculation on the signal indicating the spatial feature value P (t) received from the spatial feature value calculation unit 1, and outputs a signal indicating the calculation result.

  The integral calculation units 82 to 85 correspond to the corresponding bands, for example, 0.1 to 1.0 Hz, 1.0 to 2.0 Hz, 2.0 to 3.0 Hz, and 3. An integral value at 0 to 4.0 Hz is calculated, and a signal indicating the calculation result is output.

  The threshold value determination units 29 to 32 compare the integration value calculated by the integration calculation units 82 to 85 with a predetermined threshold value, and output the comparison result and the integration value to the maximum value determination unit 33.

  The maximum value determination unit 33 selects a band in which the integrated value is maximum from the bands in the threshold determination units 29 to 32 that exceed a predetermined threshold among the bands in the calculation result by the DFT calculation unit 81. Then, the maximum value determination unit 33 outputs a signal indicating the selected band. For example, the maximum value determination unit 33 outputs a signal indicating 0.1 to 1.0 Hz when a band corresponding to the integration calculation unit 82 is selected.

  By the way, in the method described in Non-Patent Document 1, interference with other wireless services becomes a problem because a wideband signal is used, and it is affected by multipath fading indoors because the power of the received signal is used. As a result, the detection accuracy may deteriorate. In addition, the inner product of the eigenvectors calculated in the configuration described in Patent Document 1 varies depending on factors other than the target intruder, and thus may cause false detection.

  On the other hand, in the intrusion detection device according to the first embodiment of the present invention, the spatial feature amount calculation unit 1 calculates a spatial feature amount indicating the state of a predetermined area. The change rate calculation unit 2 calculates a temporal change in the spatial feature amount calculated by the spatial feature amount calculation unit 1. Then, the detection unit 11 detects a human motion in a predetermined area based on the temporal change calculated by the change rate calculation unit 2.

  In this way, by detecting the human motion using the temporal change of the spatial feature value instead of the magnitude of the variation amount of the spatial feature value, fluctuation due to factors other than human motion is eliminated, and the detection accuracy Can be improved.

  Further, in the intrusion detection device according to the first embodiment of the present invention, the detection unit 11 is configured such that the temporal change calculated by the change rate calculation unit 2 falls within a predetermined range corresponding to human movement. Then, it is determined that a human motion has occurred in a predetermined area.

  With such a configuration, it is possible to perform a detection operation focusing on the fact that the speed of the change in the state of the space caused by the human movement is within a predetermined range, so that the detection accuracy can be improved.

  In the intrusion detection device according to the first embodiment of the present invention, the detection unit 11 continues the state in which the temporal change calculated by the change rate calculation unit 2 is within a predetermined range for a predetermined time or more. In this case, it is determined that a human motion has occurred in a predetermined area.

  With such a configuration, it is possible to perform a detection operation that focuses on the fact that the change in the state of the space due to human movement continues to some extent, so that the detection accuracy can be improved.

  Moreover, in the intrusion detection apparatus according to the first embodiment of the present invention, the spatial feature amount calculation unit 1 outputs a signal indicating the calculated spatial feature amount. Then, the change rate calculation unit 2 calculates the frequency component of the signal received from the spatial feature amount calculation unit 1 as a temporal change. Then, the detection unit 11 detects a human motion in a predetermined area based on the frequency component calculated by the change rate calculation unit 2.

  As described above, the temporal change of the spatial feature amount can be appropriately evaluated by the configuration for calculating the frequency component of the spatial feature amount as the temporal change of the spatial feature amount.

  Further, in the intrusion detection device according to the first exemplary embodiment of the present invention, the band-pass filter 21 includes the frequency component of the signal received from the spatial feature amount calculation unit 1 for the target frequency band corresponding to the human motion. The component outside the target frequency band is attenuated. And the detection part 11 detects the human motion in a predetermined area based on the signal which passed the band pass filter 21. FIG.

  As described above, the detection accuracy can be further improved by the configuration in which the range of variation of the spatial feature value to be detected is set to a range unique to the object, specifically, a range corresponding to human intrusion behavior. In addition, a configuration in which a band-pass filter is provided instead of the configuration in which the DFT operation unit 81 as shown in FIG. 9 is provided eliminates the need for signal buffering, and can increase the detection time.

  Further, in the intrusion detection device according to the first embodiment of the present invention, the detection unit 11 performs human action in a predetermined area when the level of the signal that has passed through the band-pass filter 21 is equal to or higher than a predetermined threshold. Is determined to have occurred.

  As described above, with the configuration in which the determination using the threshold is performed, it is possible to detect the presence or absence of a human action with a simple process.

  Here, when an animal or an object passes in the vicinity of the transmitter or the receiver, the level of the change rate becomes high not only in a band corresponding to human movement but also in other bands. For this reason, in a configuration using only a bandpass filter corresponding to human movement, it may be difficult to distinguish between a case where an animal or an object has passed near the transmitter or receiver and a case where there is an intruder. .

  On the other hand, in the intrusion detection device according to the first embodiment of the present invention, the band pass filters 22 to 24 perform the non-target frequency band that is a band other than the target frequency band from the spatial feature quantity calculation unit 1. Among the frequency components of the received signal, components outside the non-target frequency band are attenuated. And the detection part 11 detects the human motion in a predetermined area based on the signal which passed the band pass filter 21, and the signal which passed the band pass filters 22-24.

  In this way, the configuration in which not only the band filter corresponding to the human operation but also the other band filter is provided, the change rate becomes a certain level in other bands in addition to the band corresponding to the human operation. In this case, it is possible to accurately detect the presence / absence of a human motion by comparing the level of the change rate in each band.

  Moreover, in the intrusion detection apparatus according to the first embodiment of the present invention, the detection unit 11 has a maximum of any of the signals that have passed through the band pass filter 21 and the signals that have passed through the band pass filters 22 to 24. The human motion in the predetermined area is detected based on whether the level is equal to or lower than the predetermined level.

  As described above, with the configuration for detecting the maximum level, it is possible to detect the presence or absence of a human action with a simple comparison process.

  In the intrusion detection device according to the first embodiment of the present invention, the detection unit 11 is a signal that has passed through the band-pass filter 21 when the signal that has passed through the band-pass filter 21 has the maximum level. When the level is equal to or higher than a predetermined threshold, it is determined that a human motion has occurred in a predetermined area.

  As described above, the presence / absence of human motion can be detected more accurately by the configuration in which the determination using the threshold value is performed in addition to the comparison processing.

  In the intrusion detection device according to the first embodiment of the present invention, the change rate calculation unit 2 performs a Fourier transform on the signal received from the spatial feature amount calculation unit 1, The integral value in the frequency band corresponding to the movement of is calculated. Then, the detection unit 11 detects a human motion in a predetermined area based on the integral value calculated by the change rate calculation unit 2.

  With such a configuration, it is possible to evaluate the temporal change of the spatial feature amount more accurately than the configuration using the bandpass filter.

  Moreover, in the intrusion detection device according to the first embodiment of the present invention, the detection unit 11 determines that a human motion has occurred in a predetermined area when the integral value is equal to or greater than a predetermined threshold.

  As described above, with the configuration in which the determination using the threshold is performed, it is possible to detect the presence or absence of a human action with a simple process.

  Further, in the intrusion detection device according to the first embodiment of the present invention, the detection target range is a range of 1 Hz or less.

  As described above, the configuration in which the range of variation of the spatial feature to be detected is set to a range corresponding to a typical human intrusion behavior can improve detection accuracy without fine adjustment. Become.

  In the intrusion detection device according to the first embodiment of the present invention, the detection target range is a range that is greater than 0 Hz and less than 1 Hz, and is less than 1 Hz.

  As described above, the detection accuracy can be further improved by removing the low-frequency fluctuation component that is often caused by another environmental factor.

  In the intrusion detection device according to the first embodiment of the present invention, the detection target range can be changed.

  With such a configuration, it is possible to set the range of fluctuation of the spatial feature value to be detected according to the radio wave propagation environment of the target space, and it is possible to perform detection operations specialized for various environments.

  In the intrusion detection apparatus according to the first embodiment of the present invention, the eigenvector is used as the spatial feature quantity. However, the present invention is not limited to this. Not only the eigenvector but also a configuration using other spatial features such as a delay profile described in Non-Patent Document 1 may be used.

  In the intrusion detection apparatus according to the first embodiment of the present invention, the one-dimensional feature value is used as the spatial feature value. However, the present invention is not limited to this. A configuration using a multidimensional feature value as the spatial feature value may be used. For example, the eigenvector itself can be used as the feature vector, or a delay profile as described in Non-Patent Document 1 can be used.

  Moreover, although the intrusion detection apparatus according to the first embodiment of the present invention is configured to perform a binary determination of the presence or absence of an intruder, the present invention is not limited to this. The structure which outputs the parameter | index which shows the level of intrusion possibility may be sufficient.

  In the intrusion detection device according to the first embodiment of the present invention, the intrusion detection device is configured to use a radio wave sensor, but the present invention is not limited to this. The configuration is not limited to the radio wave sensor, and various sensors such as a sound wave sensor such as an ultrasonic wave may be used.

  Next, another embodiment of the present invention will be described with reference to the drawings. In the drawings, the same or corresponding parts are denoted by the same reference numerals and description thereof will not be repeated.

<Second Embodiment>
The present embodiment relates to an intrusion detection device in which the configuration of the change rate calculation unit is changed compared to the intrusion detection device according to the first embodiment. The contents other than those described below are the same as those of the intrusion detection apparatus according to the first embodiment.

  FIG. 10 is a diagram showing a configuration of a change rate calculation unit in the intrusion detection device according to the second embodiment of the present invention.

  Referring to FIG. 10, change rate calculation unit 2 includes comb filters 61 to 64 instead of bandpass filters 21 to 24 as compared to the change rate calculation unit according to the first embodiment of the present invention. .

  The change rate calculation unit 2 calculates not a frequency component of a signal indicating the spatial feature amount received from the spatial feature amount calculation unit 1 but a periodic change of the signal as a temporal change of the spatial feature amount.

  More specifically, the comb filters 61 to 64 amplify and pass a signal having a predetermined period among signals indicating the spatial feature amount received from the spatial feature amount calculation unit 1. For example, the comb filters 61 to 64 amplify and pass signals having periods of 0.5 seconds, 1.5 seconds, 2.5 seconds, and 3.5 seconds, respectively.

  The power calculators 25 to 28 calculate the power of the signals that have passed through the comb filters 61 to 64.

  FIG. 11 is a diagram showing a configuration of a comb filter in the intrusion detection device according to the second embodiment of the present invention.

  Referring to FIG. 11, the comb filter includes a delay unit 71, a gain adjustment unit 72, and an addition unit 73.

  The delay unit 71 outputs a signal indicating the spatial feature value P (t) received from the spatial feature value calculation unit 1 with a delay of time T.

  The gain adjusting unit 72 adjusts and outputs the level of the signal received from the delay unit 71.

  The adder 73 adds the signal received from the gain adjuster 72 and the signal indicating the spatial feature P (t) received from the spatial feature calculator 1 and outputs the result.

  As a result, a signal having a period T, that is, a signal having a change rate of (1 / T) Hz can be extracted from the signal indicating the spatial feature P (t) received from the spatial feature calculator 1.

  Even with such a configuration, it is possible to detect a human motion with high accuracy in the same manner as the intrusion detection device according to the first embodiment of the present invention.

  Since other configurations and operations are the same as those of the intrusion detection device according to the first embodiment, detailed description thereof will not be repeated here.

  Next, another embodiment of the present invention will be described with reference to the drawings. In the drawings, the same or corresponding parts are denoted by the same reference numerals and description thereof will not be repeated.

<Third Embodiment>
The present embodiment relates to an intrusion detection apparatus that further performs pattern determination of a spatial feature amount as compared to the intrusion detection apparatus according to the first embodiment. The contents other than those described below are the same as those of the intrusion detection apparatus according to the first embodiment.

  FIG. 12 is a diagram showing a configuration of an intrusion detection device according to the third embodiment of the present invention.

  Referring to FIG. 12, intrusion detection device 103 further includes a rule variation component removing unit 5 as compared with the intrusion detection device according to the first embodiment of the present invention.

  The regular fluctuation component removing unit 5 is, for example, a linear prediction filter, and removes regular fluctuation of the spatial feature amount. That is, the regular fluctuation component removal unit 5 removes regular fluctuation components from the spatial feature quantity calculated by the spatial feature quantity calculation unit 1 and outputs the spatial feature quantity after the removal to the change rate calculation unit 2. .

  The change rate calculation unit 2 calculates a temporal change in the spatial feature amount after being removed by the regular fluctuation component removal unit 5.

  As described above, in the intrusion detection device according to the third embodiment of the present invention, the point (1) noted in the first embodiment of the present invention in consideration of the human intrusion behavior and the change in the state of the space. In addition to (2), note the following point (3). That is, (3) a change in the state of the space due to human movement has an irregular variation pattern. For example, there is a machine that repeats periodic motion, such as an air conditioning fan.

  FIG. 13 is a flowchart that defines an operation procedure when the intrusion detection device according to the third embodiment of the present invention performs an intrusion detection operation. The intrusion detection device 103 reads out and executes each step of the flowchart from a memory (not shown). This program can be installed externally. The installed program is distributed in a state stored in a recording medium, for example.

  The operations in steps S11 and S13 to S17 shown in FIG. 13 are the same as steps S1 to S6 in the flowchart shown in FIG.

  Referring to FIG. 13, after the spatial feature value is calculated, the detection unit 11 analyzes the variation pattern of the spatial feature value. That is, when the spatial feature amount calculated by the spatial feature amount calculation unit 1 is not irregular (NO in step S12), the detection unit 11 changes the measured spatial feature amount due to a human action. It is determined that there is no intruder (step S17). This is a determination based on the point (3).

  On the other hand, in the detection unit 11, the spatial feature amount calculated by the spatial feature amount calculation unit 1 is irregular (YES in step S12), and the change rate calculated by the change rate calculation unit 2 is within the detection target range. If the state is continuous (YES in step S14 and YES in step S15), it is determined that the variation in the measured spatial feature amount is a variation due to human action, and it is determined that there is an intruder. (Step S16).

  As described above, in the intrusion detection device according to the third exemplary embodiment of the present invention, the regular variation component removal unit 5 extracts regular variation components from among the spatial feature amounts calculated by the spatial feature amount calculation unit 1. Remove. Then, the change rate calculation unit 2 calculates a temporal change in the spatial feature amount after being removed by the regular fluctuation component removal unit 5.

  That is, when the variation of the spatial feature amount is periodic, the detection unit 11 determines that the variation is due to a mechanical motion or the like of the air conditioning fan, and determines that there is no intruder.

  This makes it possible to remove fluctuation components that are much more dominant than fluctuations due to intruders, such as regular fluctuations in spatial feature amounts due to some device, and can stably obtain high detection accuracy.

  Since other configurations and operations are the same as those of the intrusion detection device according to the first embodiment, detailed description thereof will not be repeated here.

  Next, another embodiment of the present invention will be described with reference to the drawings. In the drawings, the same or corresponding parts are denoted by the same reference numerals and description thereof will not be repeated.

<Fourth embodiment>
The present embodiment relates to an intrusion detection device in which the configuration of the change rate calculation unit is simplified compared to the intrusion detection device according to the first embodiment. The contents other than those described below are the same as those of the intrusion detection apparatus according to the first embodiment.

  FIG. 14 is a diagram showing a configuration of an intrusion detection device according to the fourth embodiment of the present invention.

  Referring to FIG. 14, intrusion detection apparatus 104 includes a spatial feature amount calculation unit 51, a change rate calculation unit 91, and a detection unit 92. The change rate calculation unit 91 includes a band pass filter 52, a power calculation unit 53, and a threshold determination unit 54. The detection unit 92 includes an integration filter 55.

  The spatial feature quantity calculation unit 51 extracts a spatial feature quantity using the arrival angle distribution in the same manner as the spatial feature quantity calculation unit 1 in the intrusion detection apparatus 101.

  The change rate calculation unit 91 calculates the change rate of the spatial feature amount. In the change rate calculation unit 91, the band pass filter 52 is similar to the band pass filter 21 in the intrusion detection device 101, among the frequency components of the signal indicating the feature amount P (t) received from the spatial feature amount calculation unit 51. Components other than the target frequency band corresponding to human motion, for example, 0.1 to 1.0 Hz are attenuated.

  The detection unit 92 detects a human motion in a predetermined area based on the signal that has passed through the band pass filter 52. For example, when the level of the signal that has passed through the band pass filter 52 is equal to or greater than a predetermined threshold, the detection unit 92 determines that a human action has occurred in a predetermined area.

  More specifically, the power calculation unit 53 calculates the power of the signal that has passed through the band-pass filter 52, similarly to the power calculation unit 25 in the intrusion detection device 101.

  Similar to the threshold determination unit 29 in the intrusion detection device 101, the threshold determination unit 54 compares the power calculated by the power calculation unit 53 with a predetermined threshold and outputs the comparison result to the detection unit 92.

  The detection unit 92 determines that no intruder exists when the pass signal of the band pass filter 52 does not exceed a predetermined threshold.

  In the detection unit 92, the integration filter 55 determines whether or not the variation of the detection target range continues for a predetermined time based on the comparison result of the threshold determination unit 54, similarly to the change duration calculation unit 4 in the intrusion detection device 101. Monitor. The integration filter 55 is, for example, a first-order low-pass filter. When the state in which the pass signal of the band-pass filter 52 exceeds a predetermined threshold value continues for a predetermined time, a signal having a logic level indicating that an intruder is present is obtained. Output.

  Note that the threshold determination unit 54 may be provided at a subsequent stage of the integration filter 55 and compare the power that has passed through the integration filter 55 with a predetermined threshold.

  Thus, in the intrusion detection apparatus according to the fourth embodiment of the present invention, only the band pass filter corresponding to the target band is provided without providing a plurality of band pass filters. Even with such a simplified configuration, it is possible to achieve the object of the present invention to detect human motion with high accuracy.

  Since other configurations and operations are the same as those of the intrusion detection device according to the first embodiment, detailed description thereof will not be repeated here.

  The above embodiment should be considered as illustrative in all points and not restrictive. The scope of the present invention is defined by the terms of the claims, rather than the description above, and is intended to include any modifications within the scope and meaning equivalent to the terms of the claims.

DESCRIPTION OF SYMBOLS 1 Spatial feature-value calculation part 2 Change rate calculation part 3 Change rate determination part 4 Change duration calculation part 5 Rule fluctuation component removal part 11 Detection part 21-24 Band pass filter 25-28 Power calculation part 29-32 Threshold determination part 33 Maximum value determination unit 51 Spatial feature amount calculation unit 52 Band pass filter 53 Power calculation unit 54 Threshold determination unit 55 Integration filter 61 to 64 Comb filter 71 Delay unit 72 Gain adjustment unit 73 Addition unit 81 DFT operation unit 82 to 85 Integral operation Unit 91 Change rate calculation unit 92 Detection unit 101, 102, 103, 104 Intrusion detection device

Claims (17)

A spatial feature amount calculation unit for calculating a spatial feature amount indicating a state of a predetermined area;
A change rate calculation unit for calculating a temporal change in the spatial feature amount calculated by the spatial feature amount calculation unit;
An intrusion detection device comprising: a detection unit for detecting a human motion in the predetermined area based on the temporal change calculated by the change rate calculation unit.   The intrusion detection apparatus according to claim 1, wherein the detection unit determines that a human motion has occurred in the predetermined area when the temporal change is within a predetermined range corresponding to a human motion.   The intrusion according to claim 2, wherein the detection unit determines that a human motion has occurred in the predetermined area when the state in which the temporal change is within the predetermined range continues for a predetermined time or longer. Detection device. The spatial feature amount calculation unit outputs a signal indicating the calculated spatial feature amount,
The change rate calculation unit calculates a frequency component of the signal received from the spatial feature amount calculation unit as the temporal change,
The intrusion detection device according to any one of claims 1 to 3, wherein the detection unit detects a human motion in the predetermined area based on the frequency component calculated by the change rate calculation unit. The change rate calculation unit includes a target filter for attenuating a component outside the target frequency band among the frequency components of the signal received from the spatial feature amount calculation unit for a target frequency band corresponding to human motion. Including
The intrusion detection device according to claim 4, wherein the detection unit detects a human motion in the predetermined area based on the signal that has passed through the target filter.   The intrusion detection apparatus according to claim 5, wherein the detection unit determines that a human motion has occurred in the predetermined area when a level of a signal that has passed through the target filter is equal to or greater than a predetermined threshold. The change rate calculation unit further includes, for a non-target frequency band that is a band other than the target frequency band, out of the frequency components of the signal received from the spatial feature amount calculation unit, a component outside the non-target frequency band. Including one or more asymmetric filters to attenuate,
The intrusion detection device according to claim 5, wherein the detection unit detects a human motion in the predetermined area based on the signal that has passed through the target filter and the signal that has passed through each of the non-target filters.   The detection unit detects a human motion in the predetermined area based on which signal has the maximum level among the signal that has passed through the target filter and the signal that has passed through each of the non-target filters. The intrusion detection device according to claim 7.   When the signal that has passed through the target filter has a maximum level and the level of the signal that has passed through the target filter is equal to or greater than a predetermined threshold, human motion occurs in the predetermined area. The intrusion detection device according to claim 8, wherein the intrusion detection device is determined to have performed. The change rate calculation unit performs a Fourier transform on the signal received from the spatial feature amount calculation unit, and calculates an integral value in a frequency band corresponding to a human motion in the result of the Fourier transform,
The intrusion detection device according to claim 4, wherein the detection unit detects a human motion in the predetermined area based on the integral value calculated by the change rate calculation unit.   The intrusion detection device according to claim 10, wherein the detection unit determines that a human motion has occurred in the predetermined area when the integrated value is equal to or greater than a predetermined threshold.   The intrusion detection device according to claim 2 or 3, wherein the predetermined range is a range of 1 Hz or less.   The intrusion detection apparatus according to any one of claims 2, 3, and 12, wherein the predetermined range is a range that is greater than 0 Hz and less than 1 Hz and greater than or equal to 1 Hz.   The intrusion detection device according to any one of claims 2, 3, 12, and 13, wherein the predetermined range can be changed. The intrusion detection device further includes:
A regular fluctuation component removing unit for removing a regular fluctuation component of the spatial feature quantity calculated by the spatial feature quantity calculation unit;
The intrusion detection device according to any one of claims 1 to 14, wherein the change rate calculation unit calculates a temporal change in the spatial feature amount after being removed by the regular fluctuation component removal unit. Calculating a spatial feature amount indicating a state of a predetermined area;
Calculating the temporal change of the calculated spatial feature amount;
Detecting an action of a human in the predetermined area based on the calculated temporal change. On the computer,
Calculating a spatial feature amount indicating a state of a predetermined area;
Calculating the temporal change of the calculated spatial feature amount;
An intrusion detection program for executing a step of detecting a human motion in the predetermined area based on the calculated temporal change.

Images