CN103217672B - Motor message detection method and device - Google Patents

Abstract

The invention provides a kind of motor message detection method and device.Described method comprises: obtain the original echo matrix corresponding to echoed signal received; Described original echo matrix is carried out out-of-band noise to suppress to obtain out-of-band noise suppression output matrix; Described out-of-band noise is suppressed the interference outside output matrix filtering motor message band and noise, retain motor message and obtain squelch output matrix; Out-of-band noise is suppressed output matrix filtering motor message, retain the interference outside motor message band and noise, obtain noise and estimate output matrix; Described squelch output matrix and noise are estimated that output matrix calculates test statistics; The motor message in echoed signal is identified according to described test statistics and the decision threshold corresponding with the false alarm rate of setting.Adopt this method can reduce and effectively control false alarm rate.

Description

Motor message detection method and device

Technical field

The present invention relates to signal processing technology, particularly relate to a kind of motor message detection method and device.

Background technology

The detection of motor message is often applied in radar life-detection instrument, to realize the detection of motor message, and then setting movement target.Radar life-detection instrument is the instrument that one can penetrate nonmetal medium (brick wall, ruins etc.) noncontact, at a distance detection human life signal.

Radar life-detection instrument emission detection signal also receives corresponding echoed signal, by judging that whether the energy of echoed signal realizes the detection of motor message apparently higher than the signal energy in neighbor distance.If the signal energy in echoed signal corresponding to a certain distance is comparatively large, spectrum peak is concentrated and apparently higher than the signal energy in neighbor distance, then think the corresponding motor message of this signal energy this distance has moving target.

But, this traditional radar life-detection instrument normally artificially decision signal energy whether significantly higher than the signal energy in neighbor distance, and then identify motor message.Concrete, whether decision signal energy will depend on the empirical value of setting higher than the signal energy in neighbor distance significantly, namely use an empirical value arranged as decision threshold, because empirical value is just rule of thumb arranged artificially, lack theories integration, therefore cannot ensure the accuracy that motor message detects, and then the false alarm rate that result in motor message detection remains high.

Summary of the invention

Based on this, be necessary the problem that in detecting for motor message, false alarm rate is high, providing a kind of can reduce the motor message detection method also effectively controlling false alarm rate.

In addition, there is a need to provide a kind of can reduce and effective motor message pick-up unit controlling false alarm rate.

A kind of motor message detection method, comprises the steps:

Obtain the original echo matrix corresponding to echoed signal received;

Described original echo matrix is carried out out-of-band noise to suppress to obtain out-of-band noise suppression output matrix;

Described out-of-band noise is suppressed the interference outside output matrix filtering motor message band and noise, retain motor message and obtain squelch output matrix;

Out-of-band noise is suppressed output matrix filtering motor message, retain the interference outside motor message band and noise, obtain noise and estimate output matrix; Described squelch output matrix and noise are estimated that output matrix calculates test statistics;

The motor message in echoed signal is identified according to described test statistics and the decision threshold corresponding with the false alarm rate of setting.

Wherein in an embodiment, described identify the step of the motor message in echoed signal according to described test statistics and with the decision threshold corresponding to false alarm rate of setting before also comprise:

Obtain the false alarm rate preset;

In the distribution of decision threshold, extract corresponding decision threshold according to described false alarm rate, it is constant false alarm rate that the motor message that the described decision threshold of described employing realizes detects.

Wherein in an embodiment, describedly to comprise according to described test statistics and the step of motor message that identifies in echoed signal with the decision threshold corresponding to false alarm rate of setting:

Judge whether described test statistics is greater than the decision threshold corresponding with the false alarm rate of setting, if so, then judges to contain motor message in described echoed signal.

Wherein in an embodiment, described identify the step of the motor message in echoed signal according to described test statistics and with the decision threshold corresponding to false alarm rate of setting after also comprise:

The distance of moving target is estimated according to described test statistics.

A kind of motor message pick-up unit, comprising:

Echo matrix acquisition module, for obtain reception echoed signal corresponding to original echo matrix;

Out-of-band noise suppression module, suppresses to obtain out-of-band noise suppression output matrix for described original echo matrix being carried out out-of-band noise;

Filtering module, for by the interference outside described out-of-band noise suppression output matrix filtering motor message band and noise, retains motor message and obtains squelch output matrix;

Noise estimation module.For described out-of-band noise is suppressed output matrix filtering motor message, retain the interference outside motor message band and noise, obtain noise and estimate output matrix;

To described squelch output matrix and noise, test statistics computing module, for estimating that output matrix calculates test statistics;

Judging module, for identifying the motor message in echoed signal according to described test statistics and the decision threshold corresponding with the false alarm rate of setting.

Wherein in an embodiment, described device also comprises:

False alarm rate acquisition module, for obtaining the false alarm rate preset;

Decision threshold computing module, for extracting corresponding decision threshold in the distribution of decision threshold according to described false alarm rate, it is constant false alarm rate that the motor message that the described decision threshold of described employing realizes detects.

Wherein in an embodiment, described identification module is also for judging whether described test statistics is greater than the decision threshold corresponding with the false alarm rate of setting, if so, then judges to contain motor message in described echoed signal.

Wherein in an embodiment, described device also comprises:

Distance estimations module, for estimating the distance of moving target according to test statistics.

Above-mentioned motor message detection method and device, obtain the original echo matrix corresponding to echoed signal received, original echo matrix is carried out out-of-band noise to suppress to obtain out-of-band noise suppression output matrix, out-of-band noise is suppressed the interference outside output matrix filtering motor message band and noise, retain motor message and obtain squelch output matrix, out-of-band noise is suppressed output matrix filtering motor message, interference outside reservation motor message band and noise obtain noise and estimate output matrix, , then squelch output matrix and noise are estimated that output matrix calculates test statistics, and then application judges test statistics with decision threshold corresponding to false alarm rate set, to identify the motor message in echoed signal, decision threshold due to application is corresponding with the false alarm rate of setting, therefore, ensure that motor message detect in false alarm rate constant, and then be conducive to the reduction and the effectively control that realize false alarm rate.

Accompanying drawing explanation

Fig. 1 is the process flow diagram of motor message detection method in an embodiment;

Fig. 2 is the echoed signal schematic diagram collected in an embodiment;

Fig. 3 is the amplitude spectrum response schematic diagram of fast temporal filtering device in an embodiment;

Fig. 4 is the shock response schematic diagram of fast temporal filtering device in an embodiment;

Fig. 5 is that in an embodiment, out-of-band noise suppresses output matrix schematic diagram;

Fig. 6 is the amplitude spectrum response schematic diagram of slow time domain bandpass filter in an embodiment;

Fig. 7 is the shock response signal of slow time domain bandpass filter in an embodiment;

Fig. 8 is squelch output matrix schematic diagram in an embodiment;

Fig. 9 is the amplitude spectrum response schematic diagram of slow time domain Hi-pass filter in an embodiment;

Figure 10 is the shock response schematic diagram of slow time domain Hi-pass filter in an embodiment;

Figure 11 is that in an embodiment, noise estimates output matrix schematic diagram;

Figure 12 is the schematic diagram of the test statistics in an embodiment;

Figure 13 is the process flow diagram of motor message detection method in another embodiment;

Figure 14 is the structural representation of motor message pick-up unit in an embodiment;

Figure 15 is the structural representation of motor message pick-up unit in another embodiment.

Embodiment

As shown in Figure 1, in one embodiment, a kind of motor message detection method, comprises the steps:

Step S110, obtains the original echo matrix corresponding to echoed signal received.

In the present embodiment, the echoed signal of reception is the radar pulse echoed signal received, and such as, can be the original signal that UWB radar collects.

Wherein, as shown in Figure 2, horizontal ordinate represents fast time domain, the data amount check n in fast time domain one frame data of the numeric representation in horizontal ordinate; Ordinate represents slow time domain, the frame number m of the slow time domain image data of the numeric representation in ordinate, and the whole section of data gathered just form an original echo matrix .

In one embodiment, for receiving the radar of echoed signal, its transmit frequency band can concentrate on 450MHz to 3.555GHz, the passband of antenna is 900MHz to 5GHz, wave cover angle is 60 °, receive by initial distance, antenna count, spatial resolution and these controling parameters of sample frequency adjust flexibly.

Wherein, initial distance refers to that radar antenna dead ahead starts the distance detected; Antenna receives to count and refers to counting of collection one frame data; Spatial resolution refers to the distance of representative between 2 in frame data; Sample frequency refers to the frame number of interior collection echoed signal per second.

Such as, initial distance can be set to 2m, antenna receives to count and is set to 2048, spatial resolution is set to 8mm, sample frequency is set to 20, then current radar investigative range is the sector region of 60 °, nearest detection range is 2m, and each frame gathers 2048 data, is referred to as 2048 points, and the distance of representative is 8mm between 2, collection 20 frame data per second, BURN-THROUGH RANGE is 2+0.008 × 2048 ≈ 18 (m), and the distance of n-th representative is s=2+0.008 × n (m), (n ∈ [1,2048]).

Step S120, carries out out-of-band noise and suppresses to obtain out-of-band noise suppression output matrix by original echo matrix.

In the present embodiment, the original echo matrix corresponding to echoed signal received comprises much noise and interference usually, wherein, the baseline wander including clutter that background environment produces, system thermonoise, the DC baseline of instability introduced owing to sampling and introduce due to system amplitude instability, therefore, need to carry out squelch to original echo matrix.

Concrete, the out-of-band noise being realized original echo matrix by fast time domain bandpass filtering is suppressed.Bandwidth and the radar pulse frequency band of the frequency response of fast time domain bandpass filter match.In a preferred embodiment, transmit frequency band concentrates on 450MHz to 3.555GHz, the passband of antenna is 900MHz to 5GHz, accordingly, the frequency response of fast time domain bandpass filter keeps as far as possible being approximately 1 on 900MHz to 3.555GHz, and in other frequency range, for restraint speckle, should as far as possible close to 0, to guarantee that radar pulse there will not be the situation of distortion on amplitude spectrum, and good filter out-band external noise.

Fast time domain bandpass filter is designed to linear phase, to keep the time domain waveform of echo-pulse.

Out-of-band noise suppresses output matrix to be obtained by following formulae discovery, that is:

$r_1[n,m]=\tilde{r}[n,m]{*}_n{h}_1\left[j\right]$

Wherein, r1 [m, n] is the output matrix of fast time domain bandpass filter, and namely out-of-band noise suppresses output matrix, for original echo matrix, * _nfor the convolution of fast time domain calculates, h ₁[j] is the unit impulse response of fast time domain bandpass filter.

In one embodiment, as shown in Figure 3 and Figure 4, the exponent number of this fast time domain bandpass filter is 82 to the parameter that fast time domain bandpass filter uses, and passband is 0.9GHz to 3.555GHz.Such as, out-of-band noise suppresses output matrix as shown in Figure 5.

Step S130, suppresses the interference outside output matrix filtering motor message band and noise by out-of-band noise, retain motor message and obtain squelch output matrix.

In the present embodiment, output matrix is suppressed to realize the filtering of interference outside motor message band and noise by slow time domain bandpass filter to out-of-band noise, this slow time domain bandpass filter is linear phase finite impulse response digital filter, to ensure the time domain waveform of signal.

Consider that motor message mainly concentrates on low-frequency range, and original echo matrix also exists stronger low frequency spur, therefore, in order to filtering interfering and noise, retain motor message, improve signal to noise ratio (S/N ratio), need to suppress output matrix to carry out slow time domain bandpass filtering to out-of-band noise.

Further, the DC baseline of the instability introduced due to sampling system and the baseline wander introduced due to system amplitude instability usually concentrate on extremely low frequency section in slow time domain, thus noise will be made to be effectively suppressed by slow time domain bandpass filter.

Such as, the 40 rank linear-phase filters that the slow time domain bandpass filter used is 0.2Hz to 6Hz for passband.

Concrete, the squelch output matrix that slow time domain bandpass filtering exports is obtained by following formulae discovery, that is:

r ₂[n,m]＝r ₁[n,m]* _mh ₂[j]

Wherein, r ₂[n, m] is squelch output matrix, r ₁[n, m] is out-of-band noise suppression output matrix, * _mfor the convolution of slow time domain calculates, h ₂[j] is the unit impulse response of slow time domain bandpass filter.

In one embodiment, as shown in Figure 6 and Figure 7, the exponent number of this slow time domain bandpass filter is 40 to the parameter that slow time domain bandpass filter uses, and passband is 0.2Hz to 6Hz, and such as, squelch output matrix as shown in Figure 8.

Step S140, suppresses output matrix filtering motor message by out-of-band noise, retain the interference outside motor message band and noise, obtains noise and estimates output matrix.

In the present embodiment, out-of-band noise suppression output matrix is carried out to the filtering of motor message by slow time domain Hi-pass filter, to obtain the signal that contains only system thermonoise.In slow time domain Hi-pass filter, cutoff frequency should be enough high, with the DC baseline of the instability of filtering motor message, noise signal, sampling system introducing as far as possible and the baseline wander due to the introducing of system amplitude instability.

Concrete, the noise obtained by slow time domain high-pass filtering estimates the calculating of output matrix as shown by the following formula:

r ₃[n,m]＝r ₁[n,m]* _mh ₃[j]

Wherein, r ₃[n, m] is noise estimation output matrix, r ₁[n, m] is out-of-band noise suppression output matrix, * _mfor the convolution of slow time domain calculates, h ₃[j] is the unit impulse response of slow time domain Hi-pass filter.

Realize parameter that slow time domain Hi-pass filter uses in one embodiment as shown in Figure 9 and Figure 10, this filter order is 40, and cutoff frequency is 8Hz.Such as, noise estimates output matrix as shown in figure 11.

To squelch output matrix and noise, step S150, estimates that output matrix calculates test statistics.

In the present embodiment, consider motor message continuation over time and space, by the accumulation done echoed signal on Time and place, to improve the probability of detection of motor message in testing process.In a preferred embodiment, the accumulated time parameter of employing is 41 × 0.05s ≈ 2s, and spatial summation parameter is 19 × 8mm ≈ 15cm, respectively corresponding 41, front and back and left and right 19 point.

Definition test statistics λ [N, M] is as follows:

$\mathrm{\λ}[N,M]=\frac{\frac{{\mathrm{\Σ}}_{n=N-9}^{N+9}{\mathrm{\Σ}}_{m=M-20}^{M+20}{\left(r_2\right[n,m\left]\right)}^2}{{\mathrm{\Σ}}_{j=0}^{40}{\left(h_2\right[j\left]\right)}^2}}{\frac{{41\×\max }_{m\∈[M-20,M+20]}\left\{{\mathrm{\Σ}}_{n=N-9}^{N+9}{\left(r_3\right[n,m\left]\right)}^2\right\}}{{\mathrm{\Σ}}_{j=0}^{40}{\left(h_3\right[j\left]\right)}^2}}$

Wherein, r ₂[n, m] is squelch output matrix, h ₂[j] is the unit impulse response of slow time domain bandpass filter; r ₃[n, m] is noise estimation output matrix; h ₃[j] for the unit impulse response of slow time domain Hi-pass filter, N, M be constant, N represents the space center in current " accumulation space ", and M represents the time centre of current " accumulated time ".

Such as, the test statistics λ [N, M] obtained in an embodiment as shown in figure 12.

Step S170, identifies the motor message in echoed signal according to test statistics and the decision threshold corresponding with the false alarm rate of setting.

In the present embodiment, preset false alarm rate, can obtain decision threshold corresponding with it according to the false alarm rate of setting, this decision threshold for judging whether there is motor message in original echoed signals, and ensures the constant false alarm rate of decision process.

As shown in figure 13, in one embodiment, also comprise before above-mentioned steps S170:

Step S210, obtains the false alarm rate preset.

Step S230, calculates corresponding decision threshold according to false alarm rate.

In the present embodiment, it is constant false alarm rate that the motor message adopting this decision threshold to realize detects.After filtering clutter, due to the irregular repetition frequency mechanism of radar front end emission coefficient, air interference is greatly restrained, and therefore, filtered echoed signal can be similar to regard as only has white Gaussian noise and motor message.

Due to design judgement index and noise power have nothing to do, in echoed signal, only have white Gaussian noise and motor message ideally, Monte Carlo can be adopted to emulate, obtain the decision threshold that different false alarm rate is corresponding.For same false alarm rate, the decision threshold under white Gaussian noise is constant, and therefore, the motor message realizing constant false alarm rate based on this principle detects.In one embodiment, conventional false alarm rate value and the decision threshold of correspondence as shown in the table:

False alarm rate P f	10e-3	10e-4	10e-5	10e-6
					Decision threshold η	0.8565	0.9274	0.9943	1.0495

In one embodiment, the detailed process of above-mentioned steps S170 is: judge whether test statistics is greater than the decision threshold corresponding with the false alarm rate of setting, if so, then judges to contain motor message in echoed signal, if not, then terminates.

In the present embodiment, the dualism hypothesis that whether there is motor message according to the decision threshold corresponding with the false alarm rate of setting is adjudicated, if determine test statistics to be greater than decision threshold, then illustrate to there is motor message, if determine test statistics to be less than decision threshold, then account for motion signal does not exist.

Such as, false alarm rate P is set _f=10e-5, decision threshold η=0.9943, is successfully detecting that output matrix that motor message obtains as shown in figure 15.

In one embodiment, further comprises the step of the distance estimating moving target according to test statistics after above-mentioned steps S170.

In the present embodiment, after determining in echoed signal and containing motor message, estimate to the distance of moving target.Concrete, to each frame data, will find λ [N] maximal value, the fast time domain variable n of its correspondence uses represent, the distance of moving target obtains by following formulae discovery:

$d=d_0+\mathrm{\Δd}\×(\hat{n}-\mathrm{delay})$

Wherein, d ₀for initial distance, Δ d is spatial resolution, and delay is the group delay of fast time-bands bandpass filter.

As shown in figure 14, in one embodiment, a kind of motor message pick-up unit, comprises echo matrix acquisition module 110, out-of-band noise suppression module 120, filtering module 130, noise estimation module 140, test statistics computing module 150 and judging module 170.

Echo matrix acquisition module 110, for obtain reception echoed signal corresponding to original echo matrix.

In the present embodiment, the echoed signal of reception is the radar pulse echoed signal received, and such as, can be the original signal that UWB radar collects.

Echo matrix acquisition module 110 forms an original echo matrix by the data that whole section gathers .

In one embodiment, for receiving the radar of echoed signal, its transmit frequency band can concentrate on 450MHz to 3.555GHz, the passband of antenna is 900MHz to 5GHz, wave cover angle is 60 °, receive by initial distance, antenna count, spatial resolution and these controling parameters of sample frequency adjust, flexibly to facilitate the reception of echoed signal.

Wherein, initial distance refers to that radar antenna dead ahead starts the distance detected; Antenna receives to count and refers to counting of collection one frame data; Spatial resolution refers to the distance of representative between 2 in frame data; Sample frequency refers to the frame number of interior collection echoed signal per second.

Such as, initial distance can be set to 2m, antenna receives to count and is set to 2048, spatial resolution is set to 8mm, sample frequency is set to 20, then current radar investigative range is the sector region of 60 °, nearest detection range is 2m, and each frame gathers 2048 data, is referred to as 2048 points, and the distance of representative is 8mm between 2, collection 20 frame data per second, BURN-THROUGH RANGE is 2+0.008 × 2048 ≈ 18 (m), and the distance of n-th representative is s=2+0.008 × n (m), (n ∈ [1,2048]).

Out-of-band noise suppression module 120, suppresses to obtain out-of-band noise suppression output matrix for original echo matrix being carried out out-of-band noise.

In the present embodiment, the original echo matrix corresponding to echoed signal received comprises much noise and interference usually, wherein, the baseline wander including clutter that background environment produces, system thermonoise, the DC baseline of instability introduced owing to sampling and introduce due to system amplitude instability, therefore, out-of-band noise suppression module 120 pairs of original echo matrixes are needed to carry out squelch.

Concrete, the out-of-band noise that out-of-band noise suppression module 120 realizes original echo matrix by fast time domain bandpass filter suppresses.Bandwidth and the radar pulse frequency band of the frequency response of fast time domain bandpass filter match.In a preferred embodiment, transmit frequency band concentrates on 450MHz to 3.555GHz, the passband of antenna is 900MHz to 5GHz, accordingly, the frequency response of fast time domain bandpass filter keeps as far as possible being approximately 1 on 900MHz to 3.555GHz, and in other frequency range, for restraint speckle, should as far as possible close to 0, to guarantee that radar pulse there will not be the situation of distortion on amplitude spectrum, and good filter out-band external noise.

Fast time domain bandpass filter is designed to linear phase, to keep the time domain waveform of echo-pulse.

Out-of-band noise suppresses output matrix to be obtained by following formulae discovery, that is:

$r_1[n,m]=\tilde{r}[n,m]{*}_n{h}_1\left[j\right]$

Wherein, r ₁[m, n] is the output matrix of fast time domain bandpass filter, and namely out-of-band noise suppresses output matrix, for original echo matrix, * _nfor the convolution of fast time domain calculates, h ₁[j] is the unit impulse response of fast time domain bandpass filter.

Filtering module 130, for out-of-band noise being suppressed interference outside output matrix filtering motor message band and noise, retaining motor message and obtaining squelch output matrix.

In the present embodiment, filtering module 130 suppresses output matrix to realize the filtering of interference outside motor message band and noise by slow time domain bandpass filter to out-of-band noise, this slow time domain bandpass filter is linear phase finite impulse response (FIR) data filter, to ensure the time domain waveform of signal.

Consider that motor message mainly concentrates on low-frequency range, and original echo matrix also exists stronger low frequency spur, therefore, in order to filtering interfering and noise, retain motor message, improve signal to noise ratio (S/N ratio), need filtering module 130 pairs of out-of-band noises to suppress output matrix to carry out slow time domain bandpass filtering.

Further, the DC baseline of the instability introduced due to sampling system and the baseline wander introduced due to system amplitude instability usually concentrate on extremely low frequency section in slow time domain, thus noise will be made to be effectively suppressed by slow time domain bandpass filter.

Such as, the slow time domain bandpass filter that filtering module 130 uses is for passband is between the 40 rank linear-phase filters of 0.2Hz to 6Hz.

Concrete, the squelch output matrix that slow time domain bandpass filtering exports is obtained by following formulae discovery, that is:

r ₂[n,m]＝r ₁[n,m]* _mh ₂[j]

Wherein, r ₂[n, m] is squelch output matrix, r ₁[n, m] is out-of-band noise suppression output matrix, * _mfor the convolution of slow time domain calculates, h ₂[j] is the unit impulse response of slow time domain bandpass filter.

Noise estimation module 140, for out-of-band noise being suppressed output matrix filtering motor message, retains the interference outside motor message band and noise, obtains noise and estimates output matrix.

In the present embodiment, noise estimation module 140 pairs of out-of-band noises suppress output matrix to be carried out the filtering of motor message by slow time domain Hi-pass filter, to obtain the signal that contains only system thermonoise.In slow time domain Hi-pass filter, cutoff frequency should be enough high, with the DC baseline of the instability of filtering motor message, noise signal, sampling system introducing as far as possible and the baseline wander due to the introducing of system amplitude instability.

Concrete, the noise that noise estimation module 140 is obtained by slow time domain Hi-pass filter estimates the calculating of output matrix as shown by the following formula:

r ₃[n,m]＝r ₁[n,m]* _mh ₃[j]

Wherein, r ₃[n, m] is noise estimation output matrix, r ₁[n, m] is out-of-band noise suppression output matrix, * _mfor the convolution of slow time domain calculates, h ₃[j] is the unit impulse response of slow time domain Hi-pass filter.

To squelch output matrix and noise, test statistics computing module 150, for estimating that output matrix calculates test statistics.

In the present embodiment, consider motor message continuation over time and space, the accumulation that test statistics computing module 150 will do echoed signal on Time and place, to improve the probability of detection of motor message in testing process.In a preferred embodiment, the accumulated time parameter of employing is 41 × 0.05s ≈ 2s, and spatial summation parameter is 19 × 8mm ≈ 15cm, respectively corresponding 41, front and back and left and right 19 point.

It is as follows that test statistics computing module 150 defines test statistics λ [N, M]:

$\mathrm{\λ}[N,M]=\frac{\frac{{\mathrm{\Σ}}_{n=N-9}^{N+9}{\mathrm{\Σ}}_{m=M-20}^{M+20}{\left(r_2\right[n,m\left]\right)}^2}{{\mathrm{\Σ}}_{j=0}^{40}{\left(h_2\right[j\left]\right)}^2}}{\frac{{41\×\max }_{m\∈[M-20,M+20]}\left\{{\mathrm{\Σ}}_{n=N-9}^{N+9}{\left(r_3\right[n,m\left]\right)}^2\right\}}{{\mathrm{\Σ}}_{j=0}^{40}{\left(h_3\right[j\left]\right)}^2}}$

Wherein, r ₂[n, m] is squelch output matrix, h ₂[j] is the unit impulse response of slow time domain bandpass filter; r ₃[n, m] is noise estimation output matrix; h ₃[j] for the unit impulse response of slow time domain Hi-pass filter, N, M be constant, N represents the space center in current " accumulation space ", and M represents the time centre of current " accumulated time ".

Judging module 170, for identifying the motor message in echoed signal according to test statistics and the decision threshold corresponding with the false alarm rate of setting.

In the present embodiment, preset false alarm rate, judgement identification module 170 can obtain decision threshold corresponding with it according to the false alarm rate of this setting, and this decision threshold for judging whether there is motor message in original echoed signals, and ensures the constant false alarm rate of decision process.

As shown in figure 15, in one embodiment, above-mentioned motor message pick-up unit also comprises false alarm rate acquisition module 210 and decision threshold computing module 230.

False alarm rate acquisition module 210, for obtaining the false alarm rate preset.

Decision threshold computing module 230, for calculating corresponding decision threshold according to false alarm rate.

In the present embodiment, it is constant false alarm rate that the motor message adopting this decision threshold to realize detects.After filtering clutter, due to the irregular repetition frequency mechanism of radar front end emission coefficient, air interference is greatly restrained, and therefore, filtered echoed signal can be similar to regard as only has white Gaussian noise and motor message.

Due to design judgement index and noise power have nothing to do, in echoed signal, only have white Gaussian noise and motor message ideally, Monte Carlo can be adopted to emulate, obtain the decision threshold that different false alarm rate is corresponding.For same false alarm rate, the decision threshold under white Gaussian noise is constant, and therefore, the motor message realizing constant false alarm rate based on this principle detects.In one embodiment, conventional false alarm rate value and the decision threshold of correspondence as shown in the table:

False alarm rate P f	10e-3	10e-4	10e-5	10e-6
					Decision threshold η	0.8565	0.9274	0.9943	1.0495

In another embodiment, above-mentioned judging module 170, also for judging whether test statistics is greater than the decision threshold corresponding with the false alarm rate of setting, if so, then judging to contain motor message in echoed signal, if not, then stopping performing.

In the present embodiment, the dualism hypothesis whether judging module 170 exists motor message according to the decision threshold corresponding with the false alarm rate of setting is adjudicated, if determine test statistics to be greater than decision threshold, then illustrate to there is motor message, if determine test statistics to be less than decision threshold, then account for motion signal does not exist.

In one embodiment, above-mentioned motor message pick-up unit also comprises distance estimations module.This distance estimations module is used for the distance estimating moving target according to test statistics.

In the present embodiment, after determining in echoed signal and containing motor message, identification module 170 is estimated to the distance of moving target.Concrete, to each frame data, will find λ [N] maximal value, the fast time domain variable n of its correspondence uses represent, the distance of moving target obtains by following formulae discovery:

$d=d_0+\mathrm{\Δd}\×(\hat{n}-\mathrm{delay})$

Wherein, d ₀for initial distance, Δ d is spatial resolution, and delay is the group delay of fast time-bands bandpass filter.

Above-mentioned motor message detection method and device will be applied in life-detection instrument, be applicable to earthquake, will cave in, ruins rescue under building collapsing, fire-fighting, municipal administration, mine rescue and ocean patrol administration, customs, border, harbour, security staff detect laminated container and whether steal into another country personnel.

Such as, when using radar life-detection instrument search and finding survivor in ruins, use the motor message that radar life-detection instrument detection indicator of trapped personnel is faint, and then realize detection to moving target, location and tracking.

By above-mentioned motor message detection method and device, the motor message achieving constant false alarm rate detects, when external interference Strength Changes, can automatically adjust its sensitivity, false alarm rate is remained unchanged, meet quick detection and the location requirement of moving target in the application scenarioss such as actual search and rescue, and then be conducive to the widespread use promoting radar life-detection instrument, improve in actual application confidence level of reporting to the police.

Above-mentioned motor message detection method and device, obtain the original echo matrix corresponding to echoed signal received, original echo matrix is carried out out-of-band noise to suppress to obtain out-of-band noise suppression output matrix, out-of-band noise is suppressed the interference outside output matrix filtering motor message band and noise, retain motor message and obtain squelch output matrix, out-of-band noise is suppressed output matrix filtering motor message, interference outside reservation motor message band and noise obtain noise and estimate output matrix, then squelch output matrix and noise are estimated that output matrix calculates test statistics, and then application judges test statistics with decision threshold corresponding to false alarm rate set, to identify the motor message in echoed signal, decision threshold due to application is corresponding with the false alarm rate of setting, therefore, ensure that motor message detect in false alarm rate constant, and then be conducive to the reduction and the effectively control that realize false alarm rate.

One of ordinary skill in the art will appreciate that all or part of flow process realized in above-described embodiment method, that the hardware that can carry out instruction relevant by computer program has come, described program can be stored in a computer read/write memory medium, this program, when performing, can comprise the flow process of the embodiment as above-mentioned each side method.Wherein, described storage medium can be magnetic disc, CD, read-only store-memory body (Read-Only Memory, ROM) or random store-memory body (Random Access Memory, RAM) etc.

The above embodiment only have expressed several embodiment of the present invention, and it describes comparatively concrete and detailed, but therefore can not be interpreted as the restriction to the scope of the claims of the present invention.It should be pointed out that for the person of ordinary skill of the art, without departing from the inventive concept of the premise, can also make some distortion and improvement, these all belong to protection scope of the present invention.Therefore, the protection domain of patent of the present invention should be as the criterion with claims.

Claims (8)

1. a motor message detection method, comprises the steps:

Obtain the original echo matrix corresponding to echoed signal received;

Described original echo matrix is carried out out-of-band noise to suppress to obtain out-of-band noise suppression output matrix, concrete, the out-of-band noise being realized original echo matrix by fast time domain bandpass filtering is suppressed, bandwidth and the radar pulse frequency band of the frequency response of fast time domain bandpass filter match, and fast time domain bandpass filter is designed to linear phase;

Described out-of-band noise is suppressed the interference outside output matrix filtering motor message band and noise, retain motor message and obtain squelch output matrix;

Out-of-band noise is suppressed output matrix filtering motor message, retain the interference outside motor message band and noise, obtain noise and estimate output matrix;

Estimate that output matrix calculates test statistics to described squelch output matrix and noise, concrete, echoed signal is done to the accumulation on Time and place, definition test statistics λ [N, M] is as follows:

$\mathrm{\λ}[N,M]=\frac{\frac{{\mathrm{\Σ}}_{n=N-9}^{N+9}{\mathrm{\Σ}}_{m=M-20}^{M+20}{\left(r_2\right[n,m\left]\right)}^2}{{\mathrm{\Σ}}_{j=0}^{40}{\left(h_2\right[j\left]\right)}^2}}{\frac{41\×{\max }_{m\∈[M-20,M+20]}\left\{{\mathrm{\Σ}}_{n=N-9}^{N+9}{\left(r_3\right[n,m\left]\right)}^2\right\}}{{\mathrm{\Σ}}_{j=0}^{40}{\left(h_3\right[j\left]\right)}^2}}$

Wherein, r ₂[n, m] is squelch output matrix, h ₂[j] is the unit impulse response of slow time domain bandpass filter; r ₃[n, m] is noise estimation output matrix; h ₃[j] for the unit impulse response of slow time domain Hi-pass filter, N, M be constant, N represents the space center in current " accumulation space ", and M represents the time centre of current " accumulated time ";

The motor message in echoed signal is identified according to described test statistics and the decision threshold corresponding with the false alarm rate of setting.

2. motor message detection method according to claim 1, is characterized in that, described identify the step of the motor message in echoed signal according to described test statistics and with the decision threshold corresponding to false alarm rate of setting before also comprise:

Obtain the false alarm rate preset;

Calculate corresponding decision threshold according to described false alarm rate, it is constant false alarm rate that the motor message that the described decision threshold of described employing realizes detects.

3. motor message detection method according to claim 1, is characterized in that, describedly comprises according to described test statistics and the step of motor message that identifies in echoed signal with the decision threshold corresponding to false alarm rate of setting:

Judge whether described test statistics is greater than the decision threshold corresponding with the false alarm rate of setting, if so, then judges to contain motor message in described echoed signal.

4. motor message detection method according to claim 3, is characterized in that, described identify the step of the motor message in echoed signal according to described test statistics and with the decision threshold corresponding to false alarm rate of setting after also comprise:

The distance of moving target is estimated according to described test statistics.

5. a motor message pick-up unit, is characterized in that, comprising:

Echo matrix acquisition module, for obtain reception echoed signal corresponding to original echo matrix;

Out-of-band noise suppression module, suppress to obtain out-of-band noise suppression output matrix for described original echo matrix being carried out out-of-band noise, concrete, the out-of-band noise being realized original echo matrix by fast time domain bandpass filtering is suppressed, bandwidth and the radar pulse frequency band of the frequency response of fast time domain bandpass filter match, and fast time domain bandpass filter is designed to linear phase;

Filtering module, for by the interference outside described out-of-band noise suppression output matrix filtering motor message band and noise, retains motor message and obtains squelch output matrix;

Noise estimation module, for described out-of-band noise is suppressed output matrix filtering motor message, retains the interference outside motor message band and noise, obtains noise and estimates output matrix;

Test statistics computing module, for estimating that output matrix calculates test statistics to described squelch output matrix and noise, concrete, echoed signal is done to the accumulation on Time and place, definition test statistics λ [N, M] is as follows:

$\mathrm{\λ}[N,M]=\frac{\frac{{\mathrm{\Σ}}_{n=N-9}^{N+9}{\mathrm{\Σ}}_{m=M-20}^{M+20}{\left(r_2\right[n,m\left]\right)}^2}{{\mathrm{\Σ}}_{j=0}^{40}{\left(h_2\right[j\left]\right)}^2}}{\frac{41\×{\max }_{m\∈[M-20,M+20]}\left\{{\mathrm{\Σ}}_{n=N-9}^{N+9}{\left(r_3\right[n,m\left]\right)}^2\right\}}{{\mathrm{\Σ}}_{j=0}^{40}{\left(h_3\right[j\left]\right)}^2}}$

Wherein, r ₂[n, m] is squelch output matrix, h ₂[j] is the unit impulse response of slow time domain bandpass filter; r ₃[n, m] is noise estimation output matrix; h ₃[j] for the unit impulse response of slow time domain Hi-pass filter, N, M be constant, N represents the space center in current " accumulation space ", and M represents the time centre of current " accumulated time ";

Judging module, for identifying the motor message in echoed signal according to described test statistics and the decision threshold corresponding with the false alarm rate of setting.

6. motor message pick-up unit according to claim 5, is characterized in that, described device also comprises:

False alarm rate acquisition module, for obtaining the false alarm rate preset;

Decision threshold computing module, for calculating corresponding decision threshold according to described false alarm rate, it is constant false alarm rate that the motor message that the described decision threshold of described employing realizes detects.

7. motor message pick-up unit according to claim 5, it is characterized in that, described judging module is also for judging whether described test statistics is greater than the decision threshold corresponding with the false alarm rate of setting, if so, then judges to contain motor message in described echoed signal.

8. motor message pick-up unit according to claim 5, is characterized in that, described device also comprises:

Distance estimations module, for estimating the distance of moving target according to test statistics.