JP4734057B2 - Moving object detection device - Google Patents

Description

  The present invention relates to an apparatus for detecting a moving object by frequency analysis of a Doppler signal. In particular, the present invention relates to an apparatus for detecting a moving object by measuring the speed and distance of a moving object having a wide speed range.

  By transmitting radio waves such as microwaves and millimeter waves to the monitoring space, receiving reflected waves from objects existing in the monitoring space, and analyzing the frequency of the reflected waves to determine the speed of the moving object and the distance to the moving object Sensors for measuring are known. These sensors are used in various fields such as intrusion detection sensors and in-vehicle radars.

  When a moving object is present in the monitoring space, the reflected wave includes a Doppler signal. The above sensor calculates the velocity of the moving object by analyzing the frequency of the reflected wave and extracting the Doppler frequency. Similarly, some sensors that analyze the frequency of Doppler signals adopt a two-frequency CW method. The two-frequency CW method is a method in which radio waves having two types of frequencies are transmitted to a monitoring area, and a distance is obtained from a phase difference between received waves corresponding to each transmission frequency.

  The frequency of the Doppler signal is proportional to the speed of the moving object, and the frequency band to be analyzed is preferably set according to the speed range of the moving object to be monitored. For example, when using sensors for security purposes to detect intruders, a wide range of speeds can be accurately detected from relatively slow moving objects such as walking people to relatively high speed moving objects such as humans who run at full speed. There is a need to. For example, when using a microwave of 24 GHz band, it is desirable that a low frequency band of several Hz to several tens Hz is used as the frequency band to be analyzed in the former and a high frequency band of several kHz is used as the frequency band in the latter.

  In addition, since the change of the Doppler signal due to the high-speed moving object is fast, it is necessary to perform frequency analysis at short time intervals (short frame length, short cycle) in order to measure the speed and distance of the high-speed moving object with high accuracy.

  On the other hand, since the change of the Doppler signal due to the low-speed moving object is slow, accurate measurement cannot be performed with a short frame length. Therefore, in order to measure the speed and distance of a low-speed moving object with high accuracy, it is necessary to perform analysis with a long frame length.

Patent Document 1 below describes a radar apparatus having a plurality of frequency analysis units having different frequency resolutions. According to this radar apparatus, the frequency analysis of the low-speed target and the high-speed target is performed by separate frequency analysis units, so that the frequency resolution of the frequency analysis unit is set to a frequency resolution suitable for each target target. Can be set.
WO02 / 019872

  As in the radar apparatus according to Patent Document 1, in the configuration of the conventional technique using a plurality of frequency analysis units, the amount of signal processing increases in proportion to the number of frequency analysis units. For this reason, a high-performance signal processing device or a plurality of signal processing devices are required, which increases the cost.

  Since the intensity of the reflected wave attenuates in proportion to the fourth power of the distance to the object, the dynamic range of the reflected wave is wide. Also, when the reflected wave is digitized and signal processing is performed, A / D conversion is performed after passing through an anti-alias filter for removing the influence of aliasing distortion due to A / D conversion.

  One way to avoid the effects of aliasing distortion while ensuring the required dynamic range is to prepare an anti-alias filter with steep band rejection characteristics. It is difficult to realize well. Therefore, in general, A / D conversion is performed at a sampling rate sufficiently higher than the highest frequency to be analyzed (oversampling). By doing so, it is possible to secure a dynamic range and remove aliasing distortion without using an anti-alias filter having such a steep band rejection characteristic.

  However, when oversampling is performed, it is necessary to perform frequency analysis including a wasteful high frequency band that is not originally required to be analyzed.

  The present invention has been made to solve the above-described problems, and enables highly accurate detection of a moving object at a low cost by suppressing a significant increase in the amount of signal processing and generation of useless processing. For the purpose.

  A moving object detection device according to the present invention includes a transmission unit that transmits a transmission wave of a predetermined frequency to a monitoring space that detects a moving object, a reception unit that receives a reflected wave with respect to the transmission wave and generates reception data, The band-specific data sequence forming means for configuring the data sequence for each band having different frequency resolution and the same number of data from the received data for a plurality of frequency bands of interest, and the data sequences for each band by the frequency shift operation. A process for associating different frequency regions that do not overlap with each other, a conversion data string generation unit that generates a conversion data string for each data string for each band, and a band synthesis data string by adding and synthesizing the plurality of conversion data strings to each other A band synthesis data sequence generation unit for generating a frequency spectrum, a frequency analysis unit for obtaining a frequency spectrum corresponding to the band synthesis data sequence, and the frequency spectrum Having a detection means for detecting the moving object based on Torr.

  In the moving object detection device according to another aspect of the invention, the frequency band of interest is set corresponding to the speed range of interest of the moving object, and the frequency resolution is set higher as the speed range of interest is lower. Is done.

  In another moving object detection apparatus according to the present invention, the reception unit samples the received reflected wave at a predetermined original sampling frequency to generate the reception data. A plurality of filter means that are provided for each frequency band of interest, receive the received data, perform frequency filtering that passes signal components of the frequency band of interest, and output a component data string; and the components of the frequency bands of interest Downsampling means for downsampling a data string to a different resampling frequency for each frequency band of interest according to the frequency resolution, and the component data after downsampling of each frequency band of interest as the data string for each band A unit that extracts a partial sequence having the number of data common to each frequency band of interest from a sequence A column extracting means, wherein the target frequency band is set corresponding to the target speed range of the moving object, and the lower the target speed range, the lower the re-sampling frequency is set. It is.

  In a preferred aspect of the present invention, at least one of the attention speed ranges is a range including a human walking speed, and at least one of the attention speed ranges is a range including a human running speed. Device.

  In another moving object detection apparatus according to the present invention, the frequency shift operation for at least one of the band-specific data strings is a sine wave having a frequency that is half the frequency of the band-specific data string and the re-sampling frequency. Is the multiplication of

  According to the present invention, signals having different frequency resolutions are generated for each of a plurality of frequency bands of interest from a reflected signal from a moving object, and these signals are combined again into one signal. At the time of synthesis, signals for each frequency band of interest have different frequency resolutions and are associated with different frequency regions that do not overlap each other. By performing frequency analysis on the signal obtained by this synthesis, it is possible to perform frequency analysis at a different frequency resolution for a plurality of frequency bands of interest at a time. That is, it is possible to realize a highly accurate moving object detection with changing frequency resolution according to the moving speed while avoiding an increase in signal processing load, and to realize a low-cost moving object detecting device. Become.

  Hereinafter, a moving object detection device according to an embodiment of the present invention (hereinafter referred to as an embodiment) will be described with reference to the drawings. In the present embodiment, an example will be described in which the moving object detection device is used for the purpose of detecting human intrusion. Further, an example in which a signal to be transmitted / received is a microwave and a two-frequency CW method is adopted will be described.

<Embodiment 1>
FIG. 1 and FIG. 2 are schematic block configuration diagrams of the moving object detection apparatus according to the first embodiment. FIG. 1 shows the transmitter and receiver in detail, and FIG. 2 shows the processor in detail. Has been.

The transmission unit 100 includes a rectangular wave generator 110 that generates a rectangular-wave voltage waveform, a voltage-controlled oscillator 120 that generates a microwave, and prevents microwave backflow and distributes the microwave to the transmission antenna 140 and the mixer 220. A directional coupler 130 and a transmission antenna 140 that radiates microwaves to the monitoring space are configured to transmit continuous microwaves having two different frequencies f ₁ and f ₂ to the monitoring space in a time division manner.

In the present apparatus using the two-frequency CW method, the rectangular wave generator 110 alternately generates rectangular waves having two kinds of voltage values. The voltage value of the rectangular wave controls the oscillation frequency of the voltage controlled oscillator 120, and the generation period of the rectangular wave defines the switching period between the frequencies f ₁ and f ₂ . The rectangular wave is also input to the switching unit 230 of the receiving unit 200 and used for synchronizing the sample and hold 240 and 245. The generation frequency of the rectangular wave is usually set to about several tens kHz to several hundreds kHz.

The voltage controlled oscillator 120 controls the frequency of the generated microwave based on the input voltage. That is, the voltage controlled oscillator 120 alternately generates microwaves in the 24 GHz band having the frequencies f ₁ and f ₂ according to the voltage of the input rectangular wave from the rectangular wave generator 110. The voltage controlled oscillator 120 switches synchronously with the square wave frequency of the microwaves from _{f 1} to _{f 2,} or from _{f 2} to _{f 1.}

  The microwave transmitted from the transmission unit 100 is reflected by an object existing in the monitoring space, and is received by the reception unit 200 as a reflected wave. When a moving object exists in the monitoring space, the reflected wave includes a Doppler component due to the moving object.

The receiving unit 200 receives the reflected wave, and mixes the reflected wave and the microwave from the directional coupler 130 to extract only the Doppler component included in the reflected wave, and outputs the mixer 220 as a Doppler signal. In synchronization with the generation period of the rectangular wave from the rectangular wave generator 110, the switching unit 230 and the switching unit 230 that separate the Doppler signal into _{one using} the microwave of the frequency f _{1 and one using} the microwave of the frequency f ₂ Sample hold 240, 245 (abbreviated as S / H in FIG. 1) for interpolating each continuous Doppler signal, amplifiers 250, 255 for amplifying each Doppler signal, and anti-alias filter 260 for suppressing aliasing distortion of each Doppler signal. 265, and A / D converters 270 and 275, which receive the reflected wave and are included in the reflected wave. Doppler signal extracted minute and then outputs these digital signals to. Hereinafter, the Doppler signal using the microwave having the frequency f ₁ is referred to as a first Doppler signal, and the Doppler signal using the microwave having the frequency f ₂ is referred to as a second Doppler signal.

  The anti-alias filters 260 and 265 are analog filters having relatively gentle band rejection characteristics. As the A / D converters 270 and 275, those having a sampling rate of 11.025 kHz are adopted. The Nyquist frequency at this time corresponds to 123 km / h in terms of the speed of the moving object, and is sufficiently higher than the frequency to be analyzed by intrusion detection (oversampling).

  The processing unit 300 includes a processor such as a DSP or a CPU that processes digital signals, a nonvolatile memory such as a FLASH memory that stores programs and parameters, a volatile memory such as a DRAM that temporarily stores processing results, and the like.

  The processing unit 300 will be described in detail with reference to FIG. The processing unit 300 includes buffers 310 and 315 that store digital data of the first Doppler signal and the second Doppler signal input from the receiving unit 200, the digital data of the first Doppler signal, and the second Doppler signal other than the frequency band of interest. Filter units 320, 322, 324, and 326 for removing frequency components, down-sampling units 330 and 335 for down-sampling the digital data of the first Doppler signal and the second Doppler signal by decimation, etc., digital of a plurality of Doppler signals A frequency analysis unit 350 that converts digital data synthesized by a fast Fourier transform (FFT) or the like into a complex spectrum, and synthesizes the data into one digital data so that the frequency bands of interest do not overlap each other. , 355, the first Doppler signal and The Doppler frequency extraction unit 360 that extracts the Doppler frequency based on the complex spectrum of the second Doppler signal, and the distance from the moving object detection device to the moving object based on the phase difference between the first Doppler signal and the second Doppler signal at the Doppler frequency. Distance distribution calculation unit 370 to calculate, velocity distribution calculation unit 380 to calculate the speed of the moving object, intensity distribution calculation unit 385 to calculate the reflection intensity by the moving object, intrusion based on the distance to the moving object and the speed of the moving object An intrusion determination unit 390 that determines the presence or absence of a person is included. For example, each of these units is stored in a nonvolatile memory as a program, and is read and executed by a processor when the apparatus is activated.

  Hereinafter, each unit included in the processing unit 300 will be described. FIG. 3 is a schematic diagram illustrating an example of Doppler signal data accumulated in the buffer units 310 and 315, in which the horizontal axis corresponds to time and the vertical axis corresponds to a voltage value. FIG. 4 is a schematic diagram illustrating an example of Doppler signal data in each unit in the processing unit 300. As in FIG. 3, the horizontal axis represents time and the vertical axis represents voltage value. It represents as. FIG. 5 is a schematic diagram showing a frequency spectrum corresponding to each data example of FIG. 4, where the horizontal axis corresponds to the frequency and the vertical axis corresponds to the intensity in decibel display.

  The buffer unit 310, the filter units 320 and 322, the downsampling unit 330, the synthesis unit 340, and the frequency analysis unit 350 process the first Doppler signal.

  The buffer unit 310 stores at least N latest data among the data of the first Doppler signal digitized by the A / D converter 270. In the present embodiment, a FIFO (First In First Out) buffer is used, and N is 2048.

  The filter unit 320 reads out the N latest data from the buffer unit 310 as a Doppler signal for analyzing a high-speed moving object (hereinafter, data for a high-speed moving object), and removes frequency components other than the frequency band of interest. In the present embodiment, a low-pass filter (LPF) having a cutoff frequency of 2.75625 kHz with a frequency band of interest of 0 Hz or more and 1/2 or less of the Nyquist frequency is used. This attention frequency band is equivalent to setting the attention speed range to 62 km / h or less, and is set with a sufficient margin for the speed of a human running at full speed. The frequency band removed here is a useless frequency band generated by the oversampling in the A / D converter 270 described above. FIG. 4A shows high-speed moving object data after the first Doppler signal stored in the buffer unit 310 as shown in FIG. 3 passes through the filter unit 320. FIG. 5A shows a frequency spectrum corresponding to the high-speed moving object data shown in FIG.

  On the other hand, the filter unit 322 reads the N latest data from the buffer unit 310 as a Doppler signal for analyzing a low-speed moving object (hereinafter, data for a low-speed moving object), and removes frequency components other than the frequency band of interest. . In the present embodiment, an LPF having a notable frequency band of 0 Hz or more and 1/16 or less of the Nyquist frequency and a cutoff frequency of 344.53125 Hz is used. This attention frequency band corresponds to the attention speed range being 2.14 m or less per second, which is about the walking speed of a human.

  The downsampling unit 330 downsamples a Doppler signal for analyzing a low-speed moving object. In the present embodiment, downsampling is performed by thinning out every eight pieces of data output from the filter unit 322.

  Here, the down-sampled data is N / 8, but the down-sampling unit 330 holds at least 7N / 8 data that have been down-sampled in the past, and the combined N data as the down-sampling result. Configure to output.

  The downsampled 2048 low-speed moving object data has a frame length of 1486 msec, which is longer than the frame length of 182048 msec of the 2048 high-speed moving object data, and the frequency resolution is improved. Further, by making the two pieces of data the same number, the composition by the composition unit 340 described later becomes possible.

  FIG. 4B shows low-speed moving object data output by the downsampling unit 330. FIG. 5 (b) shows the frequency spectrum of the low-speed moving object data shown in FIG. 4 (b). FIG. 5B is an enlarged form of the low frequency band of FIG. 5A, which is the frequency spectrum of the data for high-speed moving objects. For example, 50 Hz that can be observed only with a coarse resolution in FIG. It can be seen that the peak frequency appearing in the vicinity can be observed with a finer resolution in FIG. That is, the low-speed moving object data output by the downsampling unit 330 has finer resolution in the low frequency band than the high-speed moving object data.

  The down-sampling method may calculate the average value of continuous 8 data every 8 pieces, or calculate the median value of continuous 8 data every 8 pieces, and the interval is a divisor of N. If so, it may not be 8.

  The synthesizing unit 340 synthesizes the data for the high-speed moving object that has passed through the filter unit 320 and the data for the low-speed moving object output from the downsampling unit 330 into one data. Specifically, the component of the target frequency band of the low-speed moving object data is frequency-shifted to the high-frequency band of the high-speed moving object data removed by the filter unit 320, and the low-speed movement is frequency-shifted from the high-speed moving object data. The object data is added.

  In addition, it is assumed that two data are virtually sampled at the same sampling rate, and values obtained by adding the first data values, the second data values,..., The Nth data values are calculated. It is processing.

  The frequency shift is performed by multiplying the original data to be shifted by a sine wave. The frequency of the sine wave to be multiplied is called the shift frequency. Due to the frequency shift, the spectrum of the original data is translated on the frequency axis so that the position of 0 Hz becomes the position of the shift frequency while maintaining its shape. At this time, a virtual image is generated in which the spectrum shape of the original data is folded around the shift frequency. In the present embodiment, the shift frequency is set to the Nyquist frequency of 689.0625 Hz after down-sampling so that this virtual image appears in the high frequency band from which the frequency component has been removed by the filter unit 320. FIG. 4C shows the data for the low-speed moving object after the frequency shift by the synthesis unit 340, and FIG. 5C shows the virtual image that appears by frequency-shifting the frequency spectrum of FIG. 5B. Yes. As already mentioned, this device has over-sampling at a frequency higher than that required to acquire data for high-speed moving objects, so there is a gap between the frequency spectrum of the data for high-speed moving objects and the Nyquist frequency. A frequency band is generated. In this apparatus, the frequency conversion is performed so that the virtual image of the frequency spectrum of the data for the low-speed moving object is substantially contained in this empty area. That is, in the present apparatus, the virtual image is set so as not to substantially overlap the target frequency band of the data for the high-speed moving object.

  FIG. 4D shows data obtained by adding and synthesizing the high-speed moving object data in FIG. 4A and the low-speed moving object data in FIG. FIG. 5D is a frequency spectrum of the data after the addition synthesis, and a virtual image of the frequency spectrum of the data for the high-speed moving object in FIG. 5A and the frequency spectrum of the data for the low-speed moving object in FIG. It has the characteristic of combining the part.

  In the above-described configuration, it has been described that the combining unit 340 shifts the frequency of the low-speed moving object data and adds it to the high-speed moving object data, but the combining unit 340 shifts the high-speed moving object data. And may be added to the data for low-speed moving objects.

  The frequency analysis unit 350 performs window processing using a Hamming window on the data from the synthesis unit 340 and performs frequency analysis on the windowed data using FFT to calculate a complex spectrum. The data from the synthesizing unit 340 includes both the high-speed moving object data and the low-speed moving object data, and the analysis regarding the two frequency bands of interest can be executed simultaneously by one frequency analysis.

  A Hanning window or the like may be used as a window function for window processing, or frequency analysis may be performed by a discrete Fourier transform (DFT) or the like.

  The configuration of processing the first Doppler signal in the processing unit 300 has been described above. The configuration of processing the second Doppler signal includes the buffer unit 315, the filter unit 324, the filter unit 326, the downsampling unit 335, and the synthesis unit. 345, performed by the frequency analysis unit 355. Here, the buffer unit 315 is a buffer unit 310, the filter unit 324 is a filter unit 320, the filter unit 326 is a filter unit 322, the downsampling unit 335 is a downsampling unit 330, the synthesis unit 345 is a synthesis unit 340, Since the frequency analysis unit 355 is the same as the frequency analysis unit 350, description thereof will be omitted.

  The Doppler frequency extraction unit 360 calculates the intensities of the complex spectrum of the first Doppler signal input from the frequency analysis unit 350 and the complex spectrum of the second Doppler signal input from the frequency analysis unit 355, A frequency whose intensity difference is smaller than a predetermined threshold is extracted as a Doppler frequency.

Such processing utilizes the fact that the reflected wave from the same moving object is similarly received even in the case of either f ₁ or f ₂ frequency, and excludes noise components other than the moving object from the intrusion determination. Process. In other words, Doppler frequencies having similar intensities are extracted for both the first Doppler signal and the second Doppler signal.

  The distance distribution calculation unit 370 generates the first Doppler signal and the second Doppler signal based on the complex spectrum of the first Doppler signal extracted from the frequency analysis unit 350 and the second Doppler signal extracted from the frequency analysis unit 355 for each Doppler frequency. The phase difference from the signal is obtained, and the distance to the moving object is calculated using the phase difference. From this calculation result, it can be seen how far the moving object is distributed from the moving object detection device.

  Based on the complex spectrum of the first Doppler signal extracted from the frequency analysis unit 350 and the second Doppler signal extracted from the frequency analysis unit 355 for each Doppler frequency, the intensity distribution calculation unit 385 determines the intensity of the first Doppler signal and the first Doppler signal. An average value with the intensity of the two Doppler signal is obtained, and the obtained average value is associated with the distance calculated by the distance distribution calculation unit 370. From the calculation result here, it can be seen how much the moving object of the reflection intensity is distributed at what distance from the moving object detection device.

  As shown in FIG. 5D, the velocity distribution calculation unit 380 uses a left half as a frequency axis for high-speed moving object detection and a right half as a frequency axis for low-speed moving object detection for each Doppler frequency. And the obtained speed is associated with the distance calculated by the distance distribution calculation unit 370 based on the Doppler frequency. From the calculation result here, it can be understood how much moving objects are distributed at what distance from the moving object detection device.

  The intrusion determining unit 390 determines whether the moving object is a human based on the intensity distribution calculated by the intensity distribution calculating unit 385 and the velocity distribution of the moving object calculated by the velocity distribution calculating unit 380. When it is determined, it is determined that there is an intrusion.

  The output unit 400 outputs a detection signal to the outside of the moving object detection device when the intrusion determination unit 390 determines that there is an intrusion. The detection signal is input to a controller or the like (not shown) of the security system, and the controller performs display indicating intrusion detection or reporting to the security center via the communication line.

  Next, the operation of this apparatus will be described.

When the apparatus is activated, continuous microwaves having two different frequencies f ₁ and f ₂ are alternately transmitted from the transmitter 100 to the monitoring space in a time division manner. The receiving unit 200 receives a reflected wave from an object in the monitoring space, converts a Doppler signal obtained by extracting the Doppler component generated by the moving object from the reflected wave into digital data, and converts the Doppler signal by the transmission wave of the frequency f ₁ . Digital data (first Doppler signal) and digital data (second Doppler signal) of a Doppler signal generated by a transmission wave having a frequency f ₂ are input to the processing unit 300.

  Hereinafter, the flow of processing by the processing unit 300 will be described with reference to the flowcharts of FIGS. 6, 7, and 8. The first Doppler signal and the second Doppler signal are subjected to frequency analysis in the first Doppler signal processing S10 and the second Doppler signal processing S12, respectively, and those results are input to the comprehensive analysis processing S14. FIG. 7 is a flowchart showing the first Doppler signal processing S10 and the second Doppler signal processing S12 in more detail, and FIG. 8 is a flowchart showing the comprehensive analysis processing S14 in more detail.

  Hereinafter, the first Doppler signal processing S10 will be described with reference to FIG. The first Doppler signal is digitized by the A / D converter 270 (step S400) and sequentially stored by the buffer unit 310 (step S402). At this time, the buffer unit 310 counts the number of accumulated data. The operations S400 and S402 of the buffer unit 310 are executed independently of the operations of the other units and repeatedly.

  Filter unit 320 refers to the count value of the number of stored data in buffer unit 310, and when the counter reaches N (YES in step S404), N data stored in buffer unit 310 is used for high-speed moving objects. Data is read out, LPF with a cutoff frequency of 2.75625 kHz is applied to the read data (step S406), and the processing result is temporarily stored (step 408).

  The filter unit 322 also refers to the count value of the buffer unit 310, and when the count value reaches N, the filter unit 322 uses the N pieces of data accumulated in the buffer unit 310 as low-speed moving object data. The read and read data are subjected to LPF with a cutoff frequency of 344.53125 Hz and input to the downsampling unit 330 (step S410).

  The down-sampling unit 330 performs down-sampling by thinning-out processing that extracts every 8 pieces of data from the input N pieces of data (step S412). In addition, the downsampling unit 330 deletes the oldest N / 8 of the N past downsampling results accumulated, and adds the N / 8 data generated in step S412 (step S412). S416).

  The synthesizer 340 reads N pieces of low-speed moving object data stored in the downsampling unit 330, multiplies the read data by a sine wave having a frequency of 689.0625 Hz, and performs frequency shift (step S418). The low-speed moving object data and the high-speed moving object data temporarily stored in the filter unit 320 are added to synthesize the two, and the combined data is temporarily stored (step S420). In the above, an example has been shown in which the processing related to the data for high-speed moving objects (steps S406 and S408) and the processing related to the data for low-speed moving objects (steps S410, S412, S416, and S418) are performed in parallel. The processing order in that case can be designed as appropriate.

  The frequency analysis unit 350 reads the synthesized data temporarily stored in the synthesis unit 340, performs window processing by multiplying the read synthesized data by a Hamming window (step S422), and performs FFT on the data subjected to the window processing. Analysis is performed to calculate the complex spectrum of the first Doppler signal, and the calculated complex spectrum is temporarily stored (step S424). As described above, according to the present invention, by generating synthesized data, a plurality of frequency analyzes with different resolutions can be performed by one FFT analysis. The amount of processing related to frequency shift and addition in the combining unit 340 is very small compared to the amount of processing related to FFT analysis, and the cost reduction effect is high.

  The processing S10 related to the first Doppler signal has been described in detail above, but the processing S12 related to the second Doppler signal is also executed in the same manner as the first Doppler signal, and the complex spectrum of the second Doppler signal as the processing result is obtained by the frequency analysis unit 355. Temporarily stored.

  Specifically, in the second Doppler signal processing S12, the buffer unit 315 performs steps S400, S402, and S404, the filter unit 324 performs steps S406 and S408, and the filter unit 326 performs step S410. The downsampling unit 335 performs the processing of steps S412 and S416, the synthesizing unit 345 performs the processing of steps S418 and S420, and the frequency analysis unit 355 performs the processing of steps S422 and S424.

  Next, the comprehensive analysis process S14 will be described with reference to FIG. The Doppler frequency extraction unit 360 extracts a frequency including the Doppler shift caused by the moving object from the complex spectrum of the first Doppler signal calculated in Process S10 and the complex spectrum of the second Doppler signal calculated in Process S12. (Step S460). For this purpose, first, the Doppler frequency extraction unit 360 calculates the frequency spectrum of the first Doppler signal and the frequency spectrum of the second Doppler signal, and obtains the difference in spectral intensity between the two for each frequency. Next, the Doppler frequency extraction unit 360 extracts a frequency whose difference is equal to or less than a predetermined threshold as a Doppler frequency and temporarily stores it.

  The distance distribution calculation unit 370 reads the complex spectrum of the first Doppler signal and the complex spectrum of the second Doppler signal for each Doppler frequency extracted in Step S460, and obtains the phase difference between the two complex spectra (Step S464). ), The distance to the moving object is calculated based on the phase difference (step S466). When the phase difference is Δφ and the speed of light is c, the distance R to the moving object is obtained by the following equation.

  The intensity distribution calculation unit 385 calculates the intensity distribution with respect to the distance to the moving object (step S468). Therefore, the intensity distribution calculation unit 385 first reads the complex spectrum of the first Doppler signal and the complex spectrum of the second Doppler signal, calculates the intensity of each read complex spectrum, and calculates the calculated intensity of the first Doppler signal. And the intensity of the second Doppler signal is defined as the intensity at each Doppler frequency. Note that the maximum value may be used instead of the average value. Next, the intensity distribution calculation unit 385 reads the distance for each Doppler frequency calculated in step S466, and associates the distance with the intensity.

The velocity distribution calculation unit 380 calculates the velocity distribution of the moving object with respect to the distance to the moving object (step S470). Therefore, the velocity distribution calculation unit 380 first converts the Doppler frequency extracted in step S460 into a velocity. When the Doppler frequency is f _d, the speed of the moving object is determined by the following equation.

  Next, the velocity distribution calculation unit 380 reads the distance for each Doppler frequency calculated in step S466, and associates the distance with the velocity.

  Here, the complex spectrum of the first Doppler signal calculated in the process S10 and the complex spectrum of the second Doppler signal calculated in the process S12 are respectively a region relating to the data for the high-speed moving object and the data for the low-speed moving object having different frequency axes. Therefore, it is necessary to obtain the value of the Doppler frequency in consideration of this. For example, in FIG. 5D, the left half is for high-speed moving object detection and the right half is for low-speed moving object detection. The former frequency axis increases to the right with the left end at 0 Hz, and the maximum value is 2 While the latter is .75625 kHz, the latter increases to the left with the right end at 0 Hz, and the maximum value is 344.53125 Hz.

  The intrusion determination unit 390 refers to the intensity distribution calculated in step S468 and the velocity distribution calculated in step S470, regards those that are close to each other as the same moving object, and determines the intensity distribution and velocity distribution for each moving object. If the intensity distribution and the velocity distribution are within a range that can be regarded as a human, it is determined that there is an intrusion (steps S472 and S474).

  If it is determined in step S474 that there is an intrusion, the output unit 400 generates an intrusion detection signal and outputs the intrusion detection signal to an external device such as a controller of the security system (step S476).

  The operation of this apparatus has been described above. As described above, in this apparatus, the processing unit 300 effectively uses a free space generated in a frequency band by oversampling a data string reconfigured so that the frequency resolution differs for each of a plurality of speed ranges of interest. Thus, they are combined into one data string. At that time, for example, the shift frequency is set to be equal to or lower than the Nyquist frequency after downsampling, and the arrangement of discrete points on the frequency axis of each of the shifted frequency spectrum and the non-shifted frequency spectrum is suitable for FFT. Is set.

The shift frequency f _shift is set to the Nyquist frequency f _srd / 2 after down-sampling the data of the speed range of interest to be frequency-shifted (data for the low-speed moving object in the configuration of the apparatus described above). It is preferable to do so. FIG. 9 is a schematic diagram of a frequency spectrum for explaining this, and schematically shows the state of aliasing that occurs when frequency shifting is performed. In FIG. 9, a spectrum 600 is a frequency spectrum of data before frequency shift, a spectrum 602 is a mapping of the frequency spectrum by the frequency shift, a spectrum 604 is a virtual image of the frequency spectrum caused by the frequency shift, and a dotted line is an aliasing of the spectra 602 and 604. Represents the frequency components respectively. The frequency component due to aliasing has a shape obtained by folding back the spectra 602 and 604, which are frequency components generated by frequency shift around the Nyquist frequency. In FIG. 9A, the entire spectrum 604, which is a virtual image, is within the high frequency band without overlapping the aliasing component. On the other hand, in FIG. 9B, aliasing components overlap in a part of the spectrum 604 that is a virtual image, and frequency analysis of this part becomes difficult. FIG. 9C is a schematic diagram in the case where the shift frequency described above is matched with the Nyquist frequency f _srd / 2 after downsampling. In this case, since the shape of the frequency component due to aliasing completely matches the shape of the mapping and virtual image due to the frequency shift, it is possible to eliminate adverse effects such as accuracy degradation that the aliasing component has on frequency analysis. In FIG. 9A and FIG. 9B, a wasteful band is generated between the shift frequency f _shift and the Nyquist frequency f _srd / 2. However, if it is as shown in FIG. . In this case, the above-described frequency shift operation can be realized by a simple process of inverting the sign of every other data.

  Note that the frequency spectrum after the data corresponding to the target frequency band is down-sampled to 1 / m is expanded to a frequency range before the down-sampling, that is, m times the bandwidth of the target frequency band. Therefore, when the frequency spectrum of each frequency band of interest is arranged without duplication using the empty frequency region generated in oversampling, the arrangement is designed in consideration of this point.

<Embodiment 2>
In the first embodiment, an example of a moving object detection apparatus that performs frequency analysis of two types of frequency bands, that is, a frequency band related to a high-speed moving object and a frequency band related to a low-speed moving object, simultaneously using one frequency analysis unit has been described. According to the above, it is also possible to simultaneously analyze the frequency of M types (M is an integer of 2 or more) with one frequency analysis unit.

  In a second embodiment below, a configuration example of a moving object detection device when M = 3 is shown. The second embodiment is an example in which frequency analysis of a frequency band related to a medium-speed moving object is added to the high-speed moving object and the low-speed moving object of the first embodiment.

  FIG. 10 is a block diagram illustrating a schematic configuration of the moving object detection device according to the second embodiment. 11 and 12 are schematic frequency spectrum diagrams for explaining the frequency conversion operation for each frequency band of interest in this apparatus. Since the configuration of the transmission unit 100, the reception unit 200, and the output unit 400 of the present apparatus is the same as that of the moving object detection apparatus of the first embodiment, description thereof is omitted. In addition, regarding the configuration of the processing unit 300, the buffer units 310 and 315, the filter units 320, 322, 324, and 326, the frequency analysis units 350 and 355, the Doppler frequency extraction unit 360, the distance distribution calculation unit 370, and the velocity distribution calculation unit 380 The intensity distribution calculation unit 385 and the intrusion determination unit 390 have the same configuration as that of the first embodiment, and thus description thereof is omitted.

  The filter unit 323 reads the N latest data from the buffer unit 310 and removes frequency components other than the target frequency band related to the medium-speed moving object. For example, when the original spectrum 620 has a shape as shown in FIG. 11A, the low frequency side region and the high frequency side region are filtered as shown in FIG. The intermediate frequency component 622 that is the frequency band of interest is removed and extracted. In the present embodiment, the frequency band of interest is 1/16 (344.53125 Hz) or more of the Nyquist frequency and 3/16 (1033.59375 Hz) or less of the Nyquist frequency, and the filter unit 323 passes 344.53125 to 10395375 Hz. A band pass filter (BPF) is used. In this case, a Doppler signal from a moving object having a speed of 2.14 m / s or more and 6.42 m / s or less can be detected from the data that has passed through the filter unit 323.

  The frequency conversion unit 331 has the function of the downsampling unit 330 of the first embodiment and the frequency shift function of the synthesis unit 340. Similarly, the frequency conversion unit 332 provided corresponding to the medium-speed moving object is also downsampled. And a function of frequency shift.

  First, the frequency conversion unit 332 performs frequency shift such that the target frequency band of the data that has passed through the filter unit 323 is arranged in a region near 0 Hz. For example, the spectrum 622 before the frequency shift shown in FIG. 11B is converted into the spectrum 624 shown in FIG. 11C by the frequency shift. In the present embodiment, a virtual image spectrum 626 is caused to appear in a frequency band from 0 Hz to 103939575 Hz by multiplying by a sine wave having a shift frequency of 1033.59375 Hz. In this apparatus, since this virtual image spectrum 626 is used for frequency analysis, the spectrum 624 on the higher frequency side than the shift frequency is suppressed. Therefore, the frequency conversion unit 332 next applies LPF with the shift frequency as the cutoff frequency to the data after the frequency shift (FIG. 11D). Subsequently, the frequency conversion unit 332 performs a downsampling process on the data that has passed through the LPF (FIG. 12A). In this embodiment, downsampling is performed by thinning out data every two data. The bandwidth of the spectrum 650 after downsampling is expanded to twice that of the spectrum 626 before downsampling. Here, the downsampled data is N / 2, but the frequency conversion unit 332 holds at least N / 2 data downsampled in the past, and N data obtained by combining these data is used as the downsampling result. Pass to the next process. The frequency conversion unit 332 multiplies the down-sampled medium-speed moving object data by a sine wave, shifts the frequency to a frequency band that does not overlap with the high-speed moving object data, and outputs the result (FIG. 12B). In FIG. 12B, a map 652 of the spectrum 650 appears on the higher frequency side than the shift frequency, and a virtual image 654 obtained by inverting the shape of the map 652 appears on the lower frequency side.

  The frequency conversion unit 331 that performs downsampling and frequency shift of the data for the low-speed moving object is different from the setting in the first embodiment, and performs downsampling by thinning out data every four pieces. The frequency conversion unit 331 holds 3N / 4 past downsampling results and outputs N downsampling results combined with the newly generated N / 4 data. The frequency conversion unit 331 multiplies the downsampled low-speed moving object data by a sine wave, shifts the frequency to a frequency band that does not overlap with the high-speed moving object data and the medium-speed moving object data, and outputs the result.

  The combining unit 343 combines the data for the high-speed moving object from the filter unit 320, the data for the medium-speed moving object from the frequency conversion unit 332, and the data for the low-speed moving object from the frequency conversion unit 331 into one data. . Specifically, the synthesizing unit 343 virtually combines the high-speed moving object data output from the filter unit 320 and the frequency-shifted medium-speed moving object data output from the frequency conversion unit 332 at the same sampling rate. Add the data as sampled data. Further, the synthesizing unit 343 performs LPF processing on the frequency-shifted data and suppresses the mapping 652 that appears on the higher frequency band side than the shift frequency (FIG. 12C). Further, the synthesizing unit 343 adds the low-speed moving object data having the spectrum 656 from the frequency conversion unit 331 to the data obtained by adding the high-speed moving object data and the medium-speed moving object data (FIG. 12D). ).

  In the present embodiment, the shift frequency and the LPF cutoff frequency relating to the medium-speed moving object data are set to 2067.1875 Hz (3/4 of the Nyquist frequency after downsampling performed during the processing relating to the medium-speed moving object). Further, the shift frequency related to the low-speed moving object data is set to 1378.125 Hz (the Nyquist frequency after downsampling performed during the processing related to the low-speed moving object).

  Next, the operation of the moving object detection device according to the second embodiment will be described with reference to the flowcharts of FIGS. The operations that are basically the same as those in the first embodiment are denoted by the same reference numerals as those in the first embodiment, and the description is simplified. The first Doppler signal and the second Doppler signal are subjected to frequency analysis in the first Doppler signal processing S20 and the second Doppler signal processing S22, respectively, and the results are input to the comprehensive analysis processing S14. FIG. 14 is a flowchart showing in more detail each of the first Doppler signal processing S20 and the second Doppler signal processing S22.

  As in the case of the first embodiment, when the apparatus is activated, the transmission unit 100 transmits a microwave to the monitoring space, the reception unit 200 receives a reflected wave from an object in the monitoring space, and receives the first Doppler signal and the first Doppler signal. A two Doppler signal is input to the processing unit 300.

  Hereinafter, the first Doppler signal processing S20 will be described with reference to FIG. The first Doppler signal is digitized by the A / D converter 270 (step S400), sequentially accumulated by the buffer unit 310 (step S402), and when N pieces of data are accumulated, N pieces of data are accumulated by the filter unit 320. Are read out as high-speed moving object data. The filter unit 320 applies LPF with a cutoff frequency of 2.75625 kHz to the high-speed moving object data (step S406), and temporarily stores the processing result (step S408).

  On the other hand, when the number of accumulated data counters by the buffer unit 310 reaches N, the processing by the filter unit 322 is also executed. The filter unit 322 reads out N pieces of data stored in the buffer unit 310 as low-speed moving object data, and applies LPF with a cutoff frequency of 344.53125 Hz to the read data (step S410). Subsequently, every four pieces of data subjected to the LPF are thinned out by the frequency conversion unit 331 (step S412). The frequency conversion unit 331 deletes the oldest N / 4 of the N past downsampling results accumulated, and adds N / 4 data that is the result of thinning out in step S416 ( Step S416).

  On the other hand, when the number of stored data counters by the buffer unit 310 reaches N, the processing by the filter unit 323 is also executed. The filter unit 323 reads the N pieces of data stored in the buffer unit 310 as medium-speed moving object data, performs BPF that passes 344.315 to 1033.593 Hz on the read data, and inputs the BPF to the frequency conversion unit 332. (Step S500).

  The frequency conversion unit 332 multiplies the frequency band of the data by multiplying by a sine wave of 1033.59375 Hz (step S505), and then performs LPF of the cut-off frequency 1033.59375 Hz on the frequency-shifted data. The generated mapping is suppressed (step S510). The frequency conversion unit 332 performs down-sampling by thinning-out processing that extracts every second data from the N pieces of data that have passed through the LPF (step S515). Thereafter, the frequency conversion unit 332 deletes the oldest N / 2 of the N past downsampling results accumulated and adds the N / 2 data generated in step S515 (step S515). S520). The latest N pieces of accumulated data are read out by the synthesis unit 343.

  The synthesizer 343 multiplies the medium-speed moving object data stored in the frequency converter 332 by a sine wave having a frequency of 2067.1875 Hz, and shifts the frequency band of the data to the high frequency side by 2067.1875 Hz (step). S525). Subsequently, the synthesizing unit 343 reads out the data for the high-speed moving object temporarily stored in the filter unit 320, and adds the read-out data and the data for the medium-speed moving object that has been frequency-shifted in Step S525 (Step S530). Further, the synthesizer 343 applies an LPF with a cutoff frequency of 4134.375 Hz (frequency based on the data for high-speed moving objects. 2067.1875 Hz when the data for medium-speed moving objects is used as a reference) to the added data, The mapping generated by the frequency shift in step S525 is suppressed (step S535).

  The synthesizing unit 343 multiplies the low-speed moving object data accumulated in the frequency conversion unit 331 by a sine wave having a frequency of 1378.125 Hz, and shifts the frequency band of the data to the high frequency side by 1378.125 Hz ( Step S418). Further, the synthesizer 343 adds the data subjected to the LPF and the data for the low-speed moving object shifted in step S418 (step S540).

  Through the above processing, the data for the high speed moving object, the data for the medium speed moving object, and the data for the low speed moving object are combined into one data so that the frequency bands of interest do not overlap each other. The synthesized data is subjected to window processing by the frequency analysis unit 350 (step S422), and FFT analysis is executed to convert it into a complex spectrum of the first Doppler signal (step S424).

  The processing S20 related to the first Doppler signal has been described in detail above, but the processing S22 related to the second Doppler signal is also executed in the same manner as the first Doppler signal, and the complex spectrum of the second Doppler signal as the processing result is obtained by the frequency analysis unit 355. Temporarily stored.

  Specifically, in the second Doppler signal processing S22, the buffer unit 315 performs steps S400, S402, and S404, the filter unit 324 performs steps S406 and S408, and the filter unit 326 performs step S410. The frequency conversion unit 336 performs steps S412, S416, and S418, the filter unit 327 performs step S500, and the frequency conversion unit 337 performs steps S505, S510, S515, S520, and S525. Are processed, the synthesizing unit 348 performs steps S530, S535, and S540, and the frequency analyzing unit 355 performs steps S422 and S424.

  Based on the complex spectrum of the first Doppler signal and the complex spectrum of the second Doppler signal calculated by the above processing, the comprehensive analysis processing S14 described in the first embodiment is executed. That is, the Doppler frequency extraction unit 360 extracts the Doppler frequency (Step S460), the distance distribution calculation unit 370 calculates the relationship between the Doppler frequency and the distance to the moving object (Steps S464 and S466), and the intensity distribution calculation unit 385. The relationship between the distance to the moving object and the intensity is calculated (step S468), the relationship between the distance to the moving object and the speed of the moving object is calculated by the velocity distribution calculation unit 380 (step S470), and the intrusion determination unit 390 monitors the monitoring space. The presence or absence of a human intrusion is determined (step S472). If it is determined that there is an intrusion (step S474), the output unit 400 outputs an intrusion detection signal to the external device (step S476). In the above-described configuration, an example in which downsampling is performed in order to make the number of pieces of data provided to the synthesis units 340, 345, 343, and 348 the same has been described, but upsampling (data interpolation, The number of data may be the same by inserting.

  In the above configuration, low-speed moving object data and high-speed moving object data, or low-speed moving object data, medium-speed moving object data, and high-speed moving object are converted from A / D converted at the same sampling rate. By reading out, filtering and down-sampling as data for each frequency band, a frequency-dependent data sequence is configured by frequency shift operation by configuring a frequency-dependent data sequence with different frequency resolution and the same number of data. However, it goes without saying that data sequences having different frequency resolutions can be generated using A / D converters having different sampling rates.

  For example, in the case of the first embodiment, the output of the amplifier 250 is input to the filter units 320 and 322 of the processing unit 300, and the filter units 320 and 322 are configured with analog LPFs. Further, the outputs of the filter units 320 and 322 are connected to the anti-alias filter, the output of the anti-alias filter is connected to the A / D converter, and the output of the A / D converter is connected to the synthesis unit 340. Preferably, the sampling rate of the A / D converter that processes the signal from the filter unit 322 is set lower than the sampling rate of the A / D converter that processes the signal from the filter unit 320. The output of the amplifier 255 can be similarly configured.

  When configured in this way, the output of each A / D converter is digital data corresponding to the band-specific data string. Also in the second embodiment, the filter units 320, 322, 323, 324, 326, and 327 can be configured by analog filters, and a similar configuration using A / D converters with different sampling rates can be employed.

  For example, in the case of the first embodiment, the output of the amplifier 250 is input to the filter units 320 and 322 of the processing unit 300, and the filter units 320 and 322 are configured by analog LPFs. Further, the output of the filter unit 320 is input to a multiplier, an oscillator that outputs a sine wave of a shift frequency is connected to the multiplier, the output of the multiplier is converted to an anti-alias filter, and the output of the anti-alias filter is A / D converted. The output of the A / D converter is connected to the combiner 340. Preferably, the sampling rate of the A / D converter that processes the signal from the filter unit 322 is set lower than the sampling rate of the A / D converter that processes the signal from the filter unit 320. The output of the amplifier 255 can be similarly configured.

  The output of each A / D converter when configured in this way is digital data corresponding to the band-specific data string after the frequency shift operation described above. In the second embodiment as well, the filter units 320, 322, 323, 324, 326, and 327 are analog filters, the frequency shift operation is configured by an oscillator and a multiplier, and A / D converters having different sampling rates are used. The configuration can be taken.

  In the above configuration, the moving object detection device has been described as an example of a case where the moving object detection device is used for detecting human intrusion. However, the moving object detection device according to the present invention is also applicable to a moving object such as a vehicle. Applicable. The attention speed range is not limited to the above, and is set as appropriate according to the moving object to be detected.

  In the above-described configuration, the signal transmitted and received by the moving object detection device has been described as a microwave. However, the moving object detection device according to the present invention is a radio wave such as a millimeter wave, a sound wave such as an ultrasonic wave, a light such as a laser. It is also applicable when using. The attention frequency band corresponding to the attention speed range and each exemplified frequency are appropriately set according to the signal to be transmitted and received. The illustrated microwave frequency is also an example, and the present invention is not limited to this.

  In the above configuration, the example of calculating the speed and distance of the moving object by the two-frequency CW method has been described. However, the moving object detection device according to the present invention transmits and receives radio waves, sound waves, and light of one kind of frequency. The present invention is also applicable to the case where the Doppler frequency is extracted and the speed and intensity of the moving object are calculated from the extracted Doppler frequency.

  As described above, the moving object detection device according to the present invention described using each embodiment has components of different frequency bands of interest from reflected signals from moving objects, has different frequency resolutions, and the same number of data. Reconstructing a plurality of data strings, frequency-shifting the reconfigured data strings to eliminate duplication of the frequency bands of interest, adding the frequency-shifted data strings to synthesize one data string, Since the synthesized data string is converted into frequency data, a plurality of frequency analyzes with different frequency resolutions can be performed at a time, and the detection accuracy of a moving object can be improved at low cost. Further, it is possible to effectively use the frequency band that has been wasted due to oversampling.

  In addition, the target frequency band is set to correspond to the target speed range of the moving object, and the lower the target speed range, the lower the sampling frequency, so that the number of data is the same. Can perform frequency analysis.

  At least one of the attention speed ranges is a range that includes human walking speed, and at least one of the attention speed ranges is a range that includes human running speed, so that humans with a wide speed range can be detected with high accuracy. it can.

  Since one of the shift frequencies is set to one half of the frequency after downsampling, the frequency-shifted data is not subjected to distortion due to aliasing, and an accurate frequency analysis is possible and the frequency band is used without waste. Can do.

1 is a schematic block configuration diagram of a moving object detection device according to Embodiment 1. FIG. FIG. 2 is a schematic block configuration diagram illustrating a configuration of a processing unit of the moving object detection device according to the first embodiment. It is a schematic diagram which shows the example of data of the Doppler signal accumulate | stored in the buffer part. It is a schematic diagram which shows the data example of the Doppler signal in each part in a process part. It is a schematic diagram which shows the frequency spectrum corresponding to each data example of FIG. 3 is a schematic flowchart illustrating processing contents of a processing unit of the moving object detection device according to the first embodiment. 6 is a schematic flowchart illustrating processing contents for a first Doppler signal and a second Doppler signal of a processing unit of the moving object detection device according to the first embodiment. It is a general | schematic flowchart which shows the content of the moving object detection process based on the frequency analysis result of each of a 1st Doppler signal and a 2nd Doppler signal. It is a schematic diagram of the frequency spectrum which shows the relationship between the spectrum and frequency aliasing which were frequency-shifted. FIG. 5 is a schematic block configuration diagram of a moving object detection device according to a second embodiment. FIG. 10 is a schematic frequency spectrum diagram for explaining a frequency conversion operation for each frequency band of interest in the moving object detection device according to the second embodiment. FIG. 10 is a schematic frequency spectrum diagram for explaining a frequency conversion operation for each frequency band of interest in the moving object detection device according to the second embodiment. 6 is a schematic flowchart illustrating processing contents of a processing unit of the moving object detection device according to the second embodiment. 6 is a schematic flowchart illustrating processing contents for a first Doppler signal and a second Doppler signal of a processing unit of the moving object detection device according to the second embodiment.

Explanation of symbols

  100 transmission unit, 200 reception unit, 300 processing unit, 310, 315 buffer unit, 320, 322, 323, 324, 326, 327 filter unit, 330, 335 downsampling unit, 331, 332, 336, 337 frequency conversion unit, 340, 343, 345, 348 Synthesizer, 350, 355 Frequency analyzer, 360 Doppler frequency extractor, 370 Distance distribution calculator, 380 Velocity distribution calculator, 385 Intensity distribution calculator, 390 Intrusion determination unit, 400 Output unit.

Claims (5)

Transmitting means for transmitting a transmission wave of a predetermined frequency to a monitoring space for detecting a moving object;
Receiving means for receiving a reflected wave with respect to the transmission wave, and generating received data by oversampling the received reflected wave at a sampling frequency that generates a free frequency region between the frequency band of interest and the Nyquist frequency ;
A plurality of the target frequency band by from the received data, and the band-by-band data stream structure means constituting the band-by-band data sequence is different from frequency resolution each other and the number of data is the same,
Wherein performs processing to associate the different frequency regions do not overlap with each other each band-by-band data string by frequency shift operation to the vacant frequency domain, converts the data string generator for generating conversion data string before each per Symbol - band data sequence Means,
Band synthesized data string generating means for adding and synthesizing the plurality of converted data strings to generate a band synthesized data string;
A frequency analysis means for obtaining a frequency spectrum corresponding to the band synthesis data string;
Detecting means for detecting the moving object based on the frequency spectrum;
A moving object detection apparatus comprising: The moving object detection device according to claim 1,
The target frequency band is set corresponding to the target speed range of the moving object,
The lower the target speed range, the higher the frequency resolution,
A moving object detection device characterized by the above. The moving object detection device according to claim 1 ,
The band-specific data string configuration means includes:
Provided before each Kichu th frequency band is inputted to the reception data, performs frequency filtering for passing a signal component of the target frequency bands, a plurality of filter means for outputting the component data string,
A down sampling means for the said component data string of each target frequency band, down-sampling to re-sampling frequency that is different for each pre-Kichu th frequency band according to the frequency resolution,
Subsequence extraction means for extracting a subsequence having the number of data common to each frequency band of interest from the component data sequence after downsampling of each frequency band of interest as the data sequence by band;
Have
The target frequency band is set corresponding to the target speed range of the moving object,
The lower the speed range of interest, the lower the resampling frequency is set;
A moving object detection device characterized by the above. The moving object detection device according to claim 3 ,
The frequency shift operation for at least one of the data sequences for each band is a multiplication of the data sequence for each band and a sine wave having a frequency that is half the re-sampling frequency. Object detection device. In the moving object detection device according to any one of claims 2 to 4 ,
At least one of the attention speed ranges is a range including a human walking speed,
At least one of the noted speed ranges is a range including a human traveling speed;
A moving object detection device characterized by the above.

Images