JP7848019B2 - Mobile object detection device, Doppler radar system, mobile object detection program, and mobile object detection method - Google Patents

Description

This disclosure relates to a technology for detecting moving objects using Doppler radar.

A technology for detecting moving objects such as pedestrians or vehicles using Doppler radar is disclosed in Patent Document 1, etc. The principle of the conventional moving object detection process is shown in Figures 1 and 2.

In Figure 1, a moving object such as a pedestrian or vehicle is approaching a Doppler radar device. In the upper part of Figure 1, the I signal output by the quadrature detection circuit leads the Q signal output by the quadrature detection circuit by approximately 1/4 period at the Doppler frequency corresponding to the detection target speed (approximately 1 m/s for pedestrians, approximately 10 m/s for vehicles). In the middle part of Figure 1, the Qs signal, obtained by shifting the Q signal output by the quadrature detection circuit by approximately 1/4 period at the Doppler frequency corresponding to the detection target speed, is approximately in phase with the I signal output by the quadrature detection circuit.

In the lower panel of Figure 1, the time-averaged value IQs-bar (where -bar is the time-averaged value) of the product of the Q signal output by the quadrature detection circuit (Qs, which is shifted by approximately 1/4 period) and the I signal output by the quadrature detection circuit is positive. By detecting that the time-averaged value IQs-bar is larger than a positive approach threshold, it is possible to detect that a moving object such as a pedestrian or vehicle is approaching the Doppler radar device without performing an FFT calculation.

In Figure 2, a moving object such as a pedestrian or vehicle is moving away from the Doppler radar device. In the upper part of Figure 2, the I signal output by the quadrature detection circuit is delayed by approximately 1/4 period compared to the Q signal output by the quadrature detection circuit at the Doppler frequency corresponding to the detection target speed (approximately 1 m/s for pedestrians, approximately 10 m/s for vehicles). In the middle part of Figure 2, the Qs signal, obtained by shifting the Q signal output by the quadrature detection circuit by approximately 1/4 period at the Doppler frequency corresponding to the detection target speed, is approximately out of phase compared to the I signal output by the quadrature detection circuit.

In the lower part of Figure 2, the time-averaged value IQs-bar (where -bar is the time-averaged value) of the product of the Q signal output by the quadrature detection circuit (Qs, which is shifted by approximately 1/4 period) and the I signal output by the quadrature detection circuit is negative. By detecting that the time-averaged value IQs-bar is small compared to the negative separation threshold, it is possible to detect that a moving object such as a pedestrian or vehicle is moving away from the Doppler radar device without performing an FFT calculation.

Patent No. 5996385

Figure 3 illustrates the challenges of conventional technology in the presence of ambient noise. In Figure 3, no moving objects such as pedestrians or vehicles are present near the Doppler radar device, yet ambient noise (such as rainfall or swaying vegetation) may be detected by the Doppler radar device.

The upper panel of Figure 3 shows the I signal output by the quadrature detection circuit. The middle panel of Figure 3 shows the Q signal output by the quadrature detection circuit. The lower panel of Figure 3 shows the time-averaged value IQs-bar (where -bar is the time-averaged value) of the product of the Q signal output by the quadrature detection circuit (Qs, which is shifted by approximately 1/4 period) and the I signal output by the quadrature detection circuit. A misdetection of a time-averaged value IQs-bar being small compared to a negative separation threshold can lead to a false detection that a moving object such as a pedestrian or vehicle is moving away from the Doppler radar device.

Therefore, in order to solve the aforementioned problems, this disclosure aims to prevent the false detection of environmental noise, such as the movement of rain or the swaying of vegetation, as moving objects when detecting pedestrians or vehicles using Doppler radar.

The moving object continuously approaches or moves away from the Doppler radar device over a long period of time, and the time-averaged phase difference between the I signal and Q signal output by the quadrature detection circuit is approximately ±90 degrees. On the other hand, when viewed individually, ambient noise superimposes approaches and moves away from the Doppler radar device, and the instantaneous phase difference between the I signal and Q signal output by the quadrature detection circuit is approximately ±90 degrees. However, when viewed as a whole, ambient noise repeatedly approaches and moves away from the Doppler radar device, and the time-averaged phase difference between the I signal and Q signal output by the quadrature detection circuit is approximately 0 degrees.

Therefore, to solve the aforementioned problem, the first multiplication value (second multiplication value) is calculated between the I signal (Q signal) output by the quadrature detection circuit, shifted by approximately 1/4 period at the Doppler frequency corresponding to the detection target speed, and the Q signal (I signal) output by the quadrature detection circuit. Then, the difference between the time average value of the first multiplication value and the time average value of the second multiplication value is calculated.

Here, for moving objects, the difference between the time-averaged value of the first multiplication and the time-averaged value of the second multiplication is approximately twice that of either the time-averaged value of the first or second multiplication. On the other hand, for ambient noise, the difference between the time-averaged value of the first multiplication and the time-averaged value of the second multiplication is approximately zero compared to either the time-averaged value of the first or second multiplication. Therefore, moving objects can be detected while improving the signal-to-noise ratio.

Specifically, this disclosure relates to a mobile object detection device that uses Doppler radar to detect moving objects, comprising: a multiplication value calculation unit that calculates a first multiplication value (second multiplication value) of the Q signal (I signal) output by a quadrature detection circuit and a signal obtained by shifting the I signal (Q signal) output by the quadrature detection circuit by approximately 1/4 period at a Doppler frequency corresponding to the detection target speed; a difference value calculation unit that calculates the difference between the time average value of the first multiplication value and the time average value of the second multiplication value, or the time average value of the difference between the first multiplication value and the second multiplication value; and a mobile object detection unit that, based on the processing result of the difference value calculation unit, enhances the Doppler signal from the moving object compared to the Doppler signal due to ambient noise and detects the moving object.

This configuration allows for more reliable prevention, compared to conventional methods, of misidentifying ambient noise as a moving object when using Doppler radar to detect moving objects.

Furthermore, to solve the aforementioned problem, the sum of the time-averaged values of the first and second multipliers is calculated. Then, based on whether the absolute value of this sum is greater than the absolute value of the difference between the time-averaged values of the first and second multipliers, it is determined whether the Doppler signal is due to a moving object or ambient noise, and a decision is made whether to perform or cancel the detection of the moving object.

Here, for moving objects, the sum of the time-averaged values of the first and second multipliers is approximately zero compared to the time-averaged value of either the first or second multiplier. On the other hand, for ambient noise, the sum of the time-averaged values of the first and second multipliers is approximately twice the time-averaged value of either the first or second multiplier. Therefore, based on both the sum of the time-averaged values of the first and second multipliers, and the difference between them, moving objects can be detected without misidentifying signals (noise) as noise (signals).

Specifically, this disclosure relates to a mobile object detection device that uses Doppler radar to detect a moving object, comprising: a multiplication value calculation unit that calculates a first multiplication value (second multiplication value) of the Q signal (I signal) output by the quadrature detection circuit and a signal that is shifted by approximately 1/4 period at a Doppler frequency corresponding to the speed of the object to be detected; a difference value calculation unit that calculates the difference between the time average value of the first multiplication value and the time average value of the second multiplication value, or the time average value of the difference between the first multiplication value and the second multiplication value; and the time average value of the first multiplication value The moving object detection device comprises: an addition value calculation unit that calculates the time-averaged value of the sum of the second multiplication value and the first multiplication value, or the time-averaged value of the sum of the first multiplication value and the second multiplication value; and a moving object detection unit that, when the absolute value of the processing result of the addition value calculation unit is larger than the absolute value of the processing result of the difference value calculation unit, determines that the Doppler signal is due to ambient noise and does not detect the moving object, and when the absolute value of the processing result of the addition value calculation unit is smaller than the absolute value of the processing result of the difference value calculation unit, determines that the Doppler signal is due to the moving object.

This configuration makes it possible to more reliably prevent the false detection of ambient noise as a moving object when using Doppler radar to detect moving objects.

Furthermore, this disclosure provides a moving object detection device characterized in that the difference value calculation unit calculates the difference between the time-averaged value of the first multiplier and the time-averaged value of the second multiplier, or the time-averaged value of the difference between the first multiplier and the second multiplier, over a moving average period comparable to the fluctuation period between the approach and receding of the ambient noise.

With this configuration, the difference between the time-averaged value of the first multiplication value and the time-averaged value of the second multiplication value is approximately zero when compared to the time-averaged value of either the first or second multiplication value. Therefore, a device having the effects described above can be provided.

Furthermore, this disclosure provides a Doppler radar system characterized by comprising the moving object detection device described above, the quadrature detection circuit, and the radar device.

This configuration provides a system having the effects described above.

Furthermore, this disclosure is a mobile object detection program that causes a computer to execute each processing step performed by each processing unit of the mobile object detection device described above.

This configuration allows for the provision of a program having the effects described above.

Furthermore, this disclosure provides a method for detecting moving objects, characterized by executing each processing step performed by each processing unit of the moving object detection device described above.

This configuration allows for the execution of procedures that have the effects described above.

Thus, this disclosure makes it possible to prevent the false detection of environmental noise, such as the movement of rain or the swaying of vegetation, as moving objects when detecting pedestrians or vehicles using Doppler radar.

This diagram shows the principle of conventional moving object detection processing. This diagram shows the principle of conventional moving object detection processing. This figure shows the challenges that conventional technologies face when environmental noise is present. This figure shows the configuration of the Doppler radar system disclosed herein. This figure shows the principle of the moving object detection process of this disclosure. This figure shows the procedure for the moving object detection process described herein. This figure shows the simulation results of the moving object detection process described herein. This figure shows the simulation results of the moving object detection process described herein. This figure shows the actual detection results of the moving object detection process of this disclosure. This figure shows the resolution results when ambient noise is present.

Embodiments of this disclosure will be described with reference to the attached drawings. The embodiments described below are examples of the implementation of this disclosure, and this disclosure is not limited to these embodiments.

(Configuration of the Doppler radar system in this disclosure)
Figure 4 shows the configuration of the Doppler radar system of this disclosure. The Doppler radar system S includes a Doppler radar device R and a moving object detection device D to detect moving objects M such as pedestrians or vehicles using microwaves or the like. However, the Doppler radar system S may detect environmental noise E such as the movement of rain or the swaying of vegetation.

The Doppler radar system R comprises a local oscillator 2, an amplifier 3, a transmitting antenna 4, a receiving antenna 5, an amplifier 6, a multiplier 7, a π/2 phase shifter 8, a multiplier 9, a high-pass filter 10, a high-pass filter 11, an amplifier 12, and an amplifier 13. The moving object detection system D comprises a D/A converter 1, an A/D converter 14, an A/D converter 15, a π/2 phase shifter 16, a multiplier 17, a π/2 phase shifter 18, a multiplier 19, a time averager 20, a time averager 21, a subtractor 22, an adder 23, and a moving object detection unit 24. The program shown in Figure 6 is installed on the computer.

The D/A converter 1 outputs a control voltage for the local oscillator 2. The local oscillator 2 generates a radar transmission signal. The amplifier 3 amplifies the radar transmission signal. The transmitting antenna 4 emits the radar transmission signal. The receiving antenna 5 receives the radar reflection signal. The amplifier 6 amplifies the radar reflection signal. The multiplier 7 outputs the I signal using the local oscillator 2. The multiplier 9 outputs the Q signal using the local oscillator 2 and the π/2 phase shifter 8.

High-pass filter 10 removes the DC component of the I signal. High-pass filter 11 removes the DC component of the Q signal. Amplifier 12 amplifies the I signal (it may also remove the DC component of the I signal as an AC amplifier). Amplifier 13 amplifies the Q signal (it may also remove the DC component of the Q signal as an AC amplifier). A/D converter 14 quantizes the I signal at a predetermined sampling rate and stores it as time-series data. A/D converter 15 quantizes the Q signal at a predetermined sampling rate and stores it as time-series data.

The π/2 phase shifter 16, multiplier 17, π/2 phase shifter 18, multiplier 19, time averager 20, time averager 21, subtractor 22, adder 23, and moving object detection unit 24, when using Doppler radar to detect moving objects M such as pedestrians or vehicles, prevent the misdetection of environmental noise E, such as rainfall movement or swaying vegetation, as moving objects M such as pedestrians or vehicles.

(Principles and procedures of the mobile object detection process described herein)
The principle of the moving object detection process of this disclosure is shown in Figure 5. In Figure 5, the phase difference between the I signal and the Q signal output by multipliers 7 and 9 is calculated as the phase of the peak frequency component of the STFT analysis after the I signal and Q signal output by multipliers 7 and 9 are subjected to STFT analysis. The frequency of occurrence of the phase difference between the I signal and the Q signal output by multipliers 7 and 9 is then displayed in a histogram.

The moving object M continuously approaches or moves away from the Doppler radar device R over a long period of time, and the time-averaged phase difference between the I signal and Q signal output by multipliers 7 and 9 is approximately ±90 degrees (see upper panel of Figure 5). On the other hand, when viewed individually, the ambient noise E superimposes the approaches and moves away from the Doppler radar device R, and the instantaneous value of the phase difference between the I signal and Q signal output by multipliers 7 and 9 is approximately ±90 degrees (see lower panel of Figure 5). However, when viewed as a whole, the ambient noise E repeatedly approaches and moves away from the Doppler radar device R, and the time-averaged phase difference between the I signal and Q signal output by multipliers 7 and 9 is approximately 0 degrees (see lower panel of Figure 5).

The procedure for the moving object detection process described in this disclosure is shown in Figure 6. Based on the principle of the moving object detection process shown in Figure 5, Figure 6 illustrates two types of moving object detection methods.

The multiplier 17 calculates the multiplication value QIs of the I signal output by the multiplier 7 and the high-pass filter 10, which is shifted by approximately 1/4 period using the π/2 phase shifter 16 at a Doppler frequency corresponding to the target speed (approximately 1 m/s for pedestrians, approximately 10 m/s for vehicles), and the Q signal output by the multiplier 9 and the high-pass filter 11 (Step S1).

The multiplier 19 calculates the multiplication value IQs of the Q signal, which is obtained by shifting the Q signal output by the multiplier 9 and the high-pass filter 11 by approximately 1/4 period using the π/2 phase shifter 18 at a Doppler frequency corresponding to the target speed (approximately 1 m/s for pedestrians, approximately 10 m/s for vehicles), and the I signal output by the multiplier 7 and the high-pass filter 10 (step S2).

The time averaging units 20, 21, and subtractor 22 calculate the difference D1 between the time average value of the multiplied value QIs and the time average value of the multiplied value IQs (step S3). Alternatively, a similar result can be obtained by calculating the time average value of the difference between the multiplied value QIs and the multiplied value IQs.

The time averaging units 20 and 21, and the adder 23 calculate the sum D2 of the time average values of the multiplicative values QIs and IQs (step S4). Alternatively, a similar result can be obtained by calculating the time average value of the sum of the multiplicative values QIs and IQs.

First, we will explain the first of the two methods for detecting moving objects. The moving object detection unit 24, based on the difference value D1, enhances the Doppler signal from the moving object M by comparing it with the Doppler signal from the ambient noise E, and detects the moving object M (step S5).

Here, for the moving object M, the difference value D1 is approximately twice the time-averaged value of the multiplied value QIs or the time-averaged value of the multiplied value IQs. On the other hand, for the ambient noise E, the difference value D1 is approximately zero compared to the time-averaged value of the multiplied value QIs or the time-averaged value of the multiplied value IQs. Therefore, the signal-to-noise ratio can be improved, and the moving object M can be detected. In other words, when detecting a moving object M using Doppler radar, it is possible to reliably prevent the misdetection of ambient noise E as the moving object M compared to conventional methods.

Next, the second of the two methods for detecting moving objects will be explained. The moving object detection unit 24 determines that the Doppler signal is due to ambient noise E when the absolute value of the summation value D2 is greater than the absolute value of the difference value D1 (YES in step S6), and stops detecting the moving object M (step S7). On the other hand, when the absolute value of the summation value D2 is smaller than the absolute value of the difference value D1 (NO in step S6), it determines that the Doppler signal is due to the moving object M, and performs detection of the moving object M (step S8).

Here, for the moving object M, the added value D2 is approximately zero compared to the time-averaged value of the multiplied value QIs or the time-averaged value of the multiplied value IQs. On the other hand, for the ambient noise E, the added value D2 is approximately twice the time-averaged value of the multiplied value QIs or the time-averaged value of the multiplied value IQs. Therefore, the moving object M can be detected based on both the added value D2 and the difference value D1, rather than just one of them, without misidentifying the signal (noise) as noise (signal). In other words, when detecting a moving object M using Doppler radar, it is possible to more reliably prevent the misidentification of ambient noise E as the moving object M.

Here, it is desirable that the time-averaging unit 20, the time-averaging unit 21, and the subtractor 22 calculate the difference value D1 over a moving average period that is comparable to the fluctuation period between the approach and departure of the ambient noise E. Then, for the ambient noise E, the difference value D1 will be approximately zero compared to the time-averaged value of the multiplied value QIs or the time-averaged value of the multiplied value IQs, thus achieving the effects described above.

On the other hand, it is desirable that the time-averaging unit 20, the time-averaging unit 21, and the adder 23 calculate the summation value D2 over a moving average period that is roughly the same as the fluctuation period between the approach and receding of the ambient noise E. Then, for the ambient noise E, the summation value D2 will be approximately twice the time-averaged value of the multiplicative value QIs or the time-averaged value of the multiplicative value IQs, thus achieving the effects described above.

Regarding rainfall patterns, the fluctuation period between approach and receding rainfall is relatively short, so a relatively short moving average period is appropriate. Conversely, regarding vegetation swaying, the fluctuation period between approach and receding rainfall is relatively long, so a relatively long moving average period is appropriate.

(Specific example results of the moving object detection process of this disclosure)
The simulation results of the moving object detection process of this disclosure are shown in Figures 7 and 8. In Figures 7 and 8, first, the control voltage of the local oscillator 2 is discretely changed, generating spike noise. Next, the control voltage of the local oscillator 2 is set to a constant value, and the moving object M repeatedly approaches and moves away.

Here, to facilitate the simulation of the moving object detection process, instead of using ambient noise E, which consists of various superimposed components, we adopted short-term, roughly square-shaped spike noise. This spike noise is the noise component of the Doppler signal output by high-pass filters 10 and 11, resulting from the DC offset fluctuation of the Doppler signal output by multipliers 7 and 9 during discrete changes in the control voltage of local oscillator 2. Therefore, the time-averaged phase difference between the I signal and Q signal output by multipliers 7 and 9 is approximately 0 degrees or approximately 180 degrees, depending on the discrete change in radar frequency and the reflection phase change from nearby fixed objects or the feedback phase change between transmission and reception.

The upper left column of Figure 7 shows the I signal. The lower left column of Figure 7 shows the Q signal. For spike noise, the time-averaged phase difference between the I signal and the Q signal is approximately 180 degrees. For the moving object M, the time-averaged phase difference between the I signal and the Q signal is approximately ±90 degrees.

The upper right column of Figure 7 shows the multiplied values QIs and IQs. The lower right column of Figure 7 shows the time-averaged values QIs-bar and IQs-bar (-bar represents the time-averaged value). For spike noise, the time-averaged values QIs-bar and IQs-bar have the same sign. For the moving object M, the time-averaged values QIs-bar and IQs-bar have opposite signs.

The upper left column of Figure 8 shows the difference value D1. The lower left column of Figure 8 shows the sum value D2. For spike noise, the difference value D1 is approximately 0, and the sum value D2 is amplified by approximately 2 times. For the moving object M, the difference value D1 is amplified by approximately 2 times, and the sum value D2 is approximately 0.

The upper right panel of Figure 8 shows the absolute values of the difference value D1 and the sum value D2. For spike noise, the absolute value of the sum value D2 is larger than the absolute value of the difference value D1. For the moving object M, the absolute value of the difference value D1 is larger than the absolute value of the sum value D2. In the lower right panel of Figure 8, the difference value D1 is suppressed to 0 during periods when the absolute value of the sum value D2 is larger than the absolute value of the difference value D1—that is, during periods when spike noise is detected. Therefore, based on the difference value D1, the spike noise is suppressed, and only the moving object M is detected.

Figure 9 shows the actual detection results of the moving object detection process described in this disclosure. In Figure 9, first, ambient noise E is generated, then the ambient noise E decreases, and the moving object M repeatedly approaches and moves away.

The upper left column of Figure 9 shows the difference value D1. The lower left column of Figure 9 shows the sum value D2. For ambient noise E, the difference value D1 is approximately 0, and the sum value D2 is approximately twice as high. For the moving object M, the difference value D1 is approximately twice as high, and the sum value D2 is approximately 0.

The upper right panel of Figure 9 shows the absolute values of the difference value D1 and the sum value D2. For ambient noise E, the absolute value of the sum value D2 is larger than the absolute value of the difference value D1. For the moving object M, the absolute value of the difference value D1 is larger than the absolute value of the sum value D2. In the lower right panel of Figure 9, the difference value D1 is suppressed to 0 during periods when the absolute value of the sum value D2 is larger than the absolute value of the difference value D1—that is, during periods when ambient noise E is detected. Therefore, in actual detection, ambient noise E is suppressed based on the difference value D1, and only the moving object M is detected.

Figure 10 shows the resolution results in the presence of ambient noise according to this disclosure. In Figure 10, similar to Figure 3, it is confirmed whether a moving object M, such as a pedestrian or vehicle, is not present near the Doppler radar device R, and whether ambient noise E is detected by the Doppler radar device R.

The upper part of Figure 10 shows the I signal output by multiplier 7. The middle part of Figure 10 shows the Q signal output by multiplier 9. The lower part of Figure 10 shows the difference value D1 after suppressing ambient noise E. By detecting that the difference value D1 is smaller than the positive proximity threshold, and larger than the negative separation threshold, it is possible to detect that a moving object M, such as a pedestrian or vehicle, is not present in the vicinity of the Doppler radar device R.

The mobile object detection device, Doppler radar system, mobile object detection program, and mobile object detection method disclosed herein can be applied to outdoor applications such as streetlights, vending machines, and security sensors, and can reduce computational and storage costs.

S: Doppler radar system R: Doppler radar device D: Moving object detection device M: Moving object E: Ambient noise 1: D/A converter 2: Local oscillator 3: Amplifier 4: Transmitting antenna 5: Receiving antenna 6: Amplifier 7: Multiplier 8: π/2 phase shifter 9: Multiplier 10: High-pass filter 11: High-pass filter 12: Amplifier 13: Amplifier 14: A/D converter 15: A/D converter 16: π/2 phase shifter 17: Multiplier 18: π/2 phase shifter 19: Multiplier 20: Time averager 21: Time averager 22: Subtractor 23: Adder 24: Moving object detection unit

Claims (6)

A mobile object detection device that uses Doppler radar to detect moving objects,
A first multiplication value calculation unit calculates a first multiplication value of the Q signal output by the quadrature detection circuit, which is obtained by shifting the I signal output by the quadrature detection circuit by approximately 1/4 period at a Doppler frequency corresponding to the detection target speed.
A second multiplication value calculation unit calculates a second multiplication value of the Q signal output by the quadrature detection circuit, shifted by approximately 1/4 period at a Doppler frequency corresponding to the detection target speed, and the I signal output by the quadrature detection circuit.
A difference value calculation unit that calculates the difference between the time average value of the first multiplication value and the time average value of the second multiplication value, or the time average value of the difference between the first multiplication value and the second multiplication value,
A moving object detection unit determines that the Doppler signal is due to ambient noise when the processing result of the difference value calculation unit is smaller than a positive approach threshold and larger than a negative separation threshold, and determines that the Doppler signal is due to the moving object when the processing result of the difference value calculation unit is larger than the positive approach threshold or smaller than the negative separation threshold .
A mobile object detection device characterized by comprising the following features. A mobile object detection device that uses Doppler radar to detect moving objects,
A first multiplication value calculation unit calculates a first multiplication value of the Q signal output by the quadrature detection circuit, which is obtained by shifting the I signal output by the quadrature detection circuit by approximately 1/4 period at a Doppler frequency corresponding to the detection target speed.
A second multiplication value calculation unit calculates a second multiplication value of the Q signal output by the quadrature detection circuit, shifted by approximately 1/4 period at a Doppler frequency corresponding to the detection target speed, and the I signal output by the quadrature detection circuit.
A difference value calculation unit that calculates the difference between the time average value of the first multiplication value and the time average value of the second multiplication value, or the time average value of the difference between the first multiplication value and the second multiplication value,
An addition value calculation unit that calculates the sum of the time average values of the first multiplication value and the second multiplication value, or the time average value of the sum of the first multiplication value and the second multiplication value,
A moving object detection unit determines that the Doppler signal is due to ambient noise and does not detect the moving object when the absolute value of the processing result of the summation value calculation unit is larger than the absolute value of the processing result of the difference value calculation unit, and determines that the Doppler signal is due to the moving object when the absolute value of the processing result of the summation value calculation unit is smaller than the absolute value of the processing result of the difference value calculation unit.
A mobile object detection device characterized by comprising the following features. The moving object detection device according to claim 1 or 2, characterized in that the difference value calculation unit calculates the difference between the time average value of the first multiplier and the time average value of the second multiplier, or the time average value of the difference between the first multiplier and the second multiplier, over a moving average period of a similar magnitude to the fluctuation period between the approach and receding of the ambient noise. A Doppler radar system comprising a moving object detection device according to any one of claims 1 to 3, the quadrature detection circuit, and a radar device. A mobile object detection program for causing a computer to execute each processing step performed by each processing unit of the mobile object detection device according to any one of claims 1 to 3. A method for detecting a moving object, characterized by performing each processing step performed by each processing unit of the moving object detection device described in any one of claims 1 to 3.