KR102328065B1 - Random frequency agile radar and method of detection of multiple objects using random frequency agile radar - Google Patents

Abstract

An objective of the present invention is to provide a frequency variable radar capable of efficiently determining presence of multiple objects from a reflected signal, and a method of determining the multiple objects using the same. The frequency variable radar comprises: a transmitter for randomly changing a center frequency to emit a pulse signal; a receiver to receive a reflected signal reflected from a tracked target object among the emitted pulse signal; a preprocessor for compensating a phase between a plurality of pulses included in the reflected signal to extract a Dopple profile corresponding to a distance cell of the tracked target object; a calculator to convert the extracted Doppler profile into a probability density function and for applying a value of the probability density function to a generative model as one feature vector to calculate likelihood from the generative model; and a situation determining unit to determine whether there are objects different from the tracked target object in the distance cell using the calculated likelihood.

Description

Multi-object determination method using variable frequency radar and frequency variable radar

The embodiment relates to a multi-object determination method using a variable-frequency radar and a variable-frequency radar, and more particularly, to an apparatus and method for determining the existence of a multi-object from a reflected signal.

In order to efficiently track the tracking target, the radar needs to determine whether an object other than the tracking target exists from a signal reflected from the tracking target. In this case, the radar can determine whether multiple objects exist by extracting a Doppler profile from the reflected signal.

However, in such a conventional radar, it is common to extract a Doppler profile using a pulse signal of the same center frequency. Accordingly, when the radar randomly changes the center frequency of the pulse signal to track the target object, there is a problem in that it is difficult to extract the Doppler profile for multi-object determination.

Republic of Korea Patent Publication No. 10-1705532

The embodiment is to overcome the above-described problems, and the multi-object determination method using the variable-frequency radar and the variable-frequency radar according to the embodiment provides an apparatus and method capable of effectively determining the existence of multiple objects from a reflected signal. aim to

The technical problems to be achieved by the embodiment are not limited to the technical problems mentioned above, and other technical problems not mentioned will be clearly understood by those of ordinary skill in the art from the description of the embodiment.

The embodiment is a frequency variable radar, a transmitter that emits a pulse signal by randomly changing a center frequency, a receiver that receives a reflected signal reflected from a tracking target among the emitted pulse signals, and a plurality of pulses included in the reflected signal A preprocessor that extracts a Doppler profile corresponding to the distance cell of the tracking target by compensating for the phase between a calculation unit for calculating likelihood from the generation model by applying to .

In addition, the preprocessing unit of the frequency variable radar according to the embodiment generates a histogram based on the frequency of a plurality of center frequencies corresponding to each of the plurality of pulses included in the reflected signal, and in the generated histogram, the highest frequency is selecting a corresponding center frequency, compensating for a phase between a plurality of pulses corresponding to the selected center frequency among a plurality of pulses included in the reflected signal, and using a reflected signal corresponding to the phase-compensated pulse A Doppler profile corresponding to the distance cell of the tracking target may be extracted.

In addition, the preprocessor of the variable frequency radar according to the embodiment may compensate for the phase by removing a relative speed component between the variable frequency radar and the tracking target object.

In addition, the preprocessor of the frequency variable radar according to the embodiment may generate the histogram based on frequencies of a plurality of center frequencies corresponding to each of a plurality of pulses existing in a coherent processing interval (CPI).

In addition, the preprocessor of the variable frequency radar according to the embodiment may generate the histogram by accumulating a plurality of pulses included in the reflected signal a reference number of times or more.

Also, the calculator of the variable frequency radar according to the embodiment may normalize a signal magnitude of the extracted Doppler profile and convert the normalized Doppler profile into a probability density function.

In addition, the generation model of the variable frequency radar according to the embodiment may include a total number of a plurality of center frequencies corresponding to a plurality of pulses included in the reflected signal and a total number of a plurality of pulses corresponding to each of the plurality of center frequencies. can be learned in advance.

Also, the generation model of the frequency variable radar according to the embodiment may be a Gaussian Mixture Model (GMM).

The embodiment is a multi-object determination method using a frequency variable radar, radiating a pulse signal by randomly changing a center frequency, receiving a reflected signal reflected from a tracking target object among the emitted pulse signals, extracting a Doppler profile corresponding to the distance cell of the tracking target by compensating for phases between a plurality of included pulses, converting the extracted Doppler profile into a probability density function, and converting the value of the probability density function into one feature calculating a likelihood from the generative model by applying it to the generative model as a vector, and determining whether an object different from the tracking target object exists in the distance cell using the calculated likelihood can

In addition, the step of extracting the Doppler profile of the multi-object determination method according to the embodiment includes generating a histogram based on the frequency of a plurality of center frequencies corresponding to each of a plurality of pulses included in the reflected signal, the selecting a center frequency corresponding to the highest frequency from the generated histogram, compensating for a phase between a plurality of pulses corresponding to the selected center frequency among a plurality of pulses included in the reflected signal, and the phase is compensated The method may include extracting a Doppler profile corresponding to a distance cell of the tracking target by using a reflected signal corresponding to the pulse.

In addition, the step of compensating for the phase of the multi-object determination method according to the embodiment may compensate for the phase by removing a relative velocity component between the frequency variable radar and the tracking target object.

In addition, the generating of the histogram of the multi-object determination method according to the embodiment comprises generating the histogram based on the frequency of a plurality of center frequencies corresponding to each of a plurality of pulses present in a Coherent Processing Interval (CPI). may be a step.

In addition, the generating of the histogram of the multi-object determination method according to the embodiment may include generating the histogram by accumulating a plurality of pulses included in the reflected signal more than a reference number of times.

In addition, calculating the likelihood of the multi-object determination method according to the embodiment may include normalizing the signal magnitude of the extracted Doppler profile and converting the normalized Doppler profile into a probability density function.

In addition, the generation model of the multi-object determination method according to the embodiment includes a total number of a plurality of center frequencies corresponding to a plurality of pulses included in the reflected signal and a total number of a plurality of pulses corresponding to each of the plurality of center frequencies. According to the number, it can be learned in advance.

In addition, the generation model of the multi-object determination method according to the embodiment may be a Gaussian Mixture Model (GMM).

The multi-object determination method using the variable-frequency radar and the variable-frequency radar according to the embodiment has an effect of effectively determining the existence of the multi-object from the reflected signal.

1 is a configuration diagram illustrating a schematic configuration of a frequency variable radar according to an embodiment.
2 is an example illustrating distribution of center frequencies corresponding to a plurality of pulses included in a pulse signal generated by a transmitter according to an embodiment.
3 is an example of a Doppler profile extracted by a preprocessor according to an embodiment.
4 is a flowchart illustrating a multi-object determination method using a frequency variable radar according to an embodiment.
5 is a flowchart illustrating a method of extracting a Doppler profile by a frequency variable radar according to an embodiment.

Hereinafter, the embodiments disclosed in the present specification will be described in detail with reference to the accompanying drawings, but identical or similar components are given the same and similar reference numerals, and overlapping descriptions thereof will be omitted. The suffixes "module" and "part" for the components used in the following description are given or mixed in consideration of only the ease of writing the specification, and do not have a meaning or role distinct from each other by themselves. In addition, in describing the embodiments disclosed in this specification, if it is determined that detailed descriptions of related known technologies may obscure the gist of the embodiments disclosed in this specification, the detailed description thereof will be omitted. In addition, the accompanying drawings are only for easy understanding of the embodiments disclosed in the present specification, and the technical spirit disclosed in the present specification is not limited by the accompanying drawings, and all changes included in the spirit and scope of the embodiments; It should be understood to include equivalents and substitutes.

Terms including an ordinal number, such as first, second, etc., may be used to describe various elements, but the elements are not limited by the terms. The above terms are used only for the purpose of distinguishing one component from another.

When a component is referred to as being “connected” or “connected” to another component, it is understood that the other component may be directly connected or connected to the other component, but other components may exist in between. it should be On the other hand, when it is mentioned that a certain element is "directly connected" or "directly connected" to another element, it should be understood that the other element does not exist in the middle.

The singular expression includes the plural expression unless the context clearly dictates otherwise.

In the present application, terms such as "comprises" or "have" are intended to designate that a feature, number, step, operation, component, part, or combination thereof described in the specification exists, but one or more other features It should be understood that this does not preclude the existence or addition of numbers, steps, operations, components, parts, or combinations thereof.

Hereinafter, a frequency variable radar according to an embodiment will be described with reference to FIGS. 1 to 3 .

1 is a configuration diagram illustrating a schematic configuration of a frequency variable radar according to an embodiment.

2 is an example illustrating distribution of center frequencies corresponding to a plurality of pulses included in a pulse signal generated by a transmitter according to an embodiment.

3 is an example of a Doppler profile extracted by a preprocessor according to an embodiment.

Referring to FIG. 1 , the variable frequency radar 100 according to the embodiment may include a transmitter 110 , a receiver 120 , a preprocessor 130 , a calculator 140 , and a situation determiner 150 . have. The variable frequency radar 100 is a kind of pulse radar, and may be a radar that emits a pulse signal and detects a tracking target using the reflected signal, but the embodiment is not limited thereto.

The transmitter 110 may radiate a pulse signal S1 to detect the tracking target object 2000 . In particular, the transmitter 110 is the center of the pulse in order to respond to the distance deception jamming of the electronic warfare side (Range Gate Pull-Off, RGPO) or distance pulling (Range Gate Pull-In, RGPI). The pulse signal S1 may be generated by randomly changing the frequency. Accordingly, each pulse included in the pulse signal S1 generated by the transmitter 110 may have a different center frequency.

내지

)중에서 하나의 중심 수파수를 무작위로 선택하여 펄스를 생성하는 경우, 동일한 중심 주파수 값에 복수의 펄스가 누적될 수 있다. 도 2에서는

에 2개의 펄스가 누적된 것을 확인할 수 있다. 이 경우, 송신부(110)가 생성한 펄스 신호(S1)에 포함된 펄스의 개수가 중심 주파수의 개수보다 많을수록 동일한 중심 주파수 값에 누적되는 펄스의 개수는 늘어날 가능성이 높아진다.In addition, referring to FIG. 2 , the transmitter 110 transmits a plurality of center frequencies (

inside

), when a pulse is generated by randomly selecting one center frequency from among, a plurality of pulses may be accumulated at the same center frequency value. In Figure 2

It can be seen that two pulses are accumulated in In this case, as the number of pulses included in the pulse signal S1 generated by the transmitter 110 is greater than the number of center frequencies, the possibility that the number of pulses accumulated at the same center frequency value increases.

The receiver 120 may receive the reflected signal S2 reflected from the tracking target object 2000 among the pulse signals S1 emitted from the transmitter 110 .

The preprocessor 130 may extract a Doppler profile corresponding to a distance cell (not shown) of the tracking target object 2000 by compensating for phases between a plurality of pulses included in the reflected signal S2 . .

The preprocessor 130 may generate a histogram based on frequencies of a plurality of center frequencies corresponding to each of a plurality of pulses included in the reflected signal S2 . In this case, the preprocessor 130 may generate a histogram based on frequencies of a plurality of center frequencies corresponding to each of a plurality of pulses existing in a coherent processing interval (CPI). That is, the preprocessor 130 may generate a histogram based on the predetermined number of pulses.

In addition, the preprocessor 130 does not generate a histogram based on the predetermined number of pulses during CPI, but accumulates pulses included in the reflected signal S2 more than a reference number of times regardless of CPI to generate a histogram. have. In this way, the preprocessor 130 may improve Doppler resolution by accumulating as many pulses as desired at the center frequency corresponding to the highest frequency on the histogram.

Also, the preprocessor 130 may select a center frequency corresponding to the highest frequency in the histogram. In addition, the preprocessor 130 may compensate a phase between a plurality of pulses corresponding to a center frequency corresponding to the highest frequency among a plurality of pulses included in the reflected signal S2 .

Due to the random frequency change of the transmitter 110 in the plurality of pulses included in the reflected signal S2 , it is highly likely that the time interval between each pulse corresponding to the most frequent center frequency is not constant. The preprocessor 130 selects the most frequent center frequency from the histogram and performs phase compensation between the corresponding pulses to uniformly generate a temporal interval between the plurality of pulses included in the reflected signal S2. have.

The equation in which the preprocessor 130 compensates for the phase of the m-th pulse in the reflected signal S2 is as shown in Equation 1 below. In Equation 1, the preprocessor 130 removes the relative velocity components of the frequency variable radar 100 and the tracking target object 2000 to generate zero Doppler.

는 중심 주파수,

는 주파수 가변 레이더(100)와 추적 대상 물체(2000) 사이의 최초 거리, v는 주파수 가변 레이더(100)와 추적 대상 물체(2000) 사이의 상대 속도, 그리고 T는 펄스 반복 주기를 의미한다.In Equation 1, θ is the phase of the transmission pulse,

is the center frequency,

is the initial distance between the variable frequency radar 100 and the tracking target 2000, v is the relative speed between the frequency variable radar 100 and the tracking target 2000, and T is the pulse repetition period.

From Equation 1, the preprocessor 130 may compensate for a phase between a plurality of pulses corresponding to the most frequent center frequency selected from the histogram among a plurality of pulses included in the reflected signal S2 . Also, the preprocessor 130 may extract a Doppler profile corresponding to the distance cell of the tracking target 2000 by using the reflected signal S2 corresponding to the phase-compensated pulse.

Referring to FIG. 3 , the number of pulses is 200, the number of center frequencies used for the pulses is 10, the number of a plurality of pulses corresponding to the most frequent center frequency is 28, the frequency variable radar 100 and the tracking target object (2000) is shown the Doppler profile for a relative velocity v between 200 m/s. In this case, the x-axis is a Doppler cell representing a Doppler frequency, and the y-axis is the power of the Doppler cell.

The calculator 140 may convert the Doppler profile extracted by the preprocessor 130 into a probability density function. In this case, the calculator 140 may normalize the signal magnitude of the Doppler profile and convert the normalized Doppler profile into a probability density function.

Also, the calculator 140 may apply the value of the probability density function converted from the Doppler profile as one feature vector to the generative model. For example, if it is assumed that the number of x-axis indices of the probability density function is 10, the total sum of probability values corresponding to 10 x-axis should be 1. That is, the value can be used as a feature vector as it is. That is, if the probability is distributed like this [0.19 0.05 0.1 0.15 0.01 0.1 0.1 0.1 0.1 0.1], the y-side value of the probability density function can be used as a feature vector as it is.
The generation model is a probability distribution model trained for the case where only the tracking target 2000 exists in the distance cell of the tracking target 2000 and the case where an object different from the tracking target 2000 exists. Specifically, the calculator 140 may use a Gaussian mixture model (GMM) as a generation model.

The probability density function converted from the Doppler profile is a probability distribution according to a total number of a plurality of center frequencies corresponding to a plurality of pulses included in the reflected signal S2 and a total number of a plurality of pulses corresponding to each of the plurality of center frequencies. may be formed differently. Accordingly, the calculation unit 140 pre-generates the generation model according to the total number of the plurality of center frequencies corresponding to the plurality of pulses included in the reflected signal S2 and the total number of the plurality of pulses corresponding to each of the plurality of center frequencies. After training, the value of the probability density function converted from the Doppler profile can be applied to the generative model as a feature vector. That is, the calculator 140 uses the value of the probability density function converted from the Doppler profile as one feature vector, applies the feature vector and the feature vectors having the same dimension to the generative model previously learned, and tracks it. In the distance cell of the target object 2000 , the likelihood of a case in which only the tracking target 2000 exists and a case in which an object different from the tracking target 2000 exists may be calculated.

The situation determination unit 150 may determine whether an object different from the tracking target object 2000 exists in the distance cell of the tracking target object 2000 using the likelihood calculated by the calculation unit 140 . Specifically, the likelihood of a case in which an object different from the tracking target 2000 exists in the distance cell of the tracking target 2000 is a case in which only the tracking target 200 exists in the distance cell of the tracking target 2000. If it is higher than the likelihood, the situation determination unit 150 may determine that an object different from the tracking target object 2000 exists in the distance cell of the tracking target object 2000 .

Hereinafter, a method of determining multiple objects using the variable frequency radar 100 according to an embodiment will be described with reference to FIG. 4 .

4 is a flowchart illustrating a multi-object determination method using a frequency variable radar according to an embodiment.

Referring to FIG. 4 , in step S100 , the frequency variable radar 100 randomly changes the center frequency to emit a pulse signal S1 , and a reflection reflected from the tracking target among the emitted pulse signals S1 . A signal S2 may be received. In particular, the frequency variable radar 100 corresponds to the distance deception of the electronic warfare side in step S100 (Range Gate Pull-Off, RGPO) or the distance deception of the (Range Gate Pull-In, RGPI) jamming. To this end, the pulse signal S1 may be generated by randomly changing the center frequency of the pulse. Accordingly, each pulse included in the pulse signal S1 generated in step S100 may have a different center frequency.

In step S200 , the frequency variable radar 100 extracts a Doppler profile corresponding to a distance cell (not shown) of the tracking target object 2000 by compensating for phases between a plurality of pulses included in the reflected signal S2 . can do.

In step S300 , the frequency variable radar 100 may convert the Doppler profile extracted in step S200 into a probability density function. In this case, the frequency variable radar 100 may normalize the signal magnitude of the Doppler profile and convert the normalized Doppler profile into a probability density function.

Also, in step S300 , the frequency tunable radar 100 may apply the value of the probability density function converted from the Doppler profile to the generation model as one feature vector. The generation model is a probability distribution model trained for the case where only the tracking target 2000 exists in the distance cell of the tracking target 2000 and the case where an object different from the tracking target 2000 exists. Specifically, the frequency variable radar 100 may use a Gaussian mixture model (GMM) as a generation model in step S300 .

The probability density function converted from the Doppler profile is a probability distribution according to a total number of a plurality of center frequencies corresponding to a plurality of pulses included in the reflected signal S2 and a total number of a plurality of pulses corresponding to each of the plurality of center frequencies. may be formed differently. Accordingly, in step S300 , the frequency variable radar 100 may include a total number of a plurality of center frequencies corresponding to a plurality of pulses included in the reflected signal S2 and a total number of a plurality of pulses corresponding to each of the plurality of center frequencies. A generative model may be trained in advance according to the number, and the probability density function value converted from the Doppler profile may be applied to the generative model as one feature vector. That is, in step S300 , the frequency tunable radar 100 uses the value of the probability density function converted from the Doppler profile as one feature vector, and is pre-learned with feature vectors having the same dimension as the feature vector. By applying the generative model, it is possible to calculate the likelihood of a case in which only the tracking target 2000 exists and a case in which an object different from the tracking target 2000 exists in the distance cell of the tracking target 2000 .

In step S400 , the variable frequency radar 100 determines whether an object different from the tracking target 2000 exists in the distance cell of the tracking target 2000 using the likelihood calculated in step S300 . can do. Specifically, the likelihood of a case in which an object different from the tracking target 2000 exists in the distance cell of the tracking target 2000 is a case in which only the tracking target 200 exists in the distance cell of the tracking target 2000. If it is higher than the likelihood, the frequency variable radar 100 may determine that an object different from the tracking target 2000 exists in the distance cell of the tracking target 2000 .

Hereinafter, the step S200 according to the embodiment will be described in detail with reference to FIG. 5 .

5 is a flowchart illustrating a method of extracting a Doppler profile by a frequency variable radar according to an embodiment.

Referring to FIG. 5 , in step S210 , the frequency variable radar 100 may generate a histogram based on frequencies of a plurality of center frequencies corresponding to each of a plurality of pulses included in the reflected signal S2 . .

In this case, the frequency variable radar 100 may generate a histogram based on frequencies of a plurality of center frequencies corresponding to each of a plurality of pulses existing in a coherent processing interval (CPI). That is, the frequency variable radar 100 may generate a histogram based on the predetermined number of pulses.

In addition, the frequency variable radar 100 does not generate a histogram based on the predetermined number of pulses during CPI, but accumulates pulses included in the reflected signal S2 more than a reference number of times regardless of CPI to generate a histogram. can In this way, the variable frequency radar 100 may improve Doppler resolution by accumulating as many pulses as desired at the center frequency corresponding to the highest frequency on the histogram.

In step S220, the frequency variable radar 100 may select a center frequency corresponding to the highest frequency in the histogram.

In step S230 , the frequency variable radar 100 may compensate a phase between a plurality of pulses corresponding to a center frequency corresponding to the highest frequency among a plurality of pulses included in the reflected signal S2 .

The plurality of pulses included in the reflected signal S2 is highly likely to have a non-uniform time interval between each pulse corresponding to the most frequent center frequency due to the random frequency change in step S100 . The frequency-variable radar 100 selects the most frequent center frequency from the histogram in step S220 and performs phase compensation between each pulse corresponding thereto in step S230. It is possible to create a uniform temporal interval between pulses.

In step S230, the equation in which the frequency variable radar 100 compensates for the phase of the m-th pulse in the reflected signal S2 is the same as Equation 1. In Equation 1, the variable frequency radar 100 removes the relative velocity component of the frequency variable radar 100 and the tracking target object 2000 in order to generate zero Doppler. The frequency variable radar 100 may compensate for a phase between a plurality of pulses corresponding to the most frequent center frequency selected from the histogram among a plurality of pulses included in the reflected signal S2 from Equation (1).

In operation S240 , the frequency variable radar 100 may extract a Doppler profile corresponding to the distance cell of the tracking target object 2000 by using the reflected signal S2 corresponding to the phase-compensated pulse.

Although the embodiment has been described in detail above, the scope of the embodiment is not limited thereto, and various modifications and improvements by those skilled in the art using the basic concept of the embodiment defined in the claims below are also included in the scope of the embodiment.

Accordingly, the foregoing detailed description should not be construed as limiting in all respects but as examples. The scope of the embodiments should be determined by a reasonable interpretation of the appended claims, and all modifications within the equivalent scope of the embodiments are included in the scope of the embodiments.

100: frequency variable radar
110: transmitter
120: receiver
130: preprocessor
140: calculator
150: situation judgment unit
2000: object to be tracked

Claims (16)

A transmitter that randomly changes the center frequency to emit a pulse signal,
a receiver for receiving a reflected signal reflected from the tracking target object among the emitted pulse signals;
a preprocessor for extracting a Doppler profile corresponding to a distance cell of the tracking target;
a calculator that converts the extracted Doppler profile into a probability density function and calculates likelihood from the generative model by applying the value of the probability density function to the generative model; and
A situation determination unit that determines whether an object different from the tracking target object exists in the distance cell using the calculated likelihood
Frequency tunable radar comprising a. According to claim 1,
The preprocessor is
generating a histogram based on the frequency of a plurality of center frequencies corresponding to each of a plurality of pulses included in the reflected signal;
selecting a center frequency corresponding to the highest frequency from the generated histogram,
Compensating for a phase between a plurality of pulses corresponding to the selected center frequency among a plurality of pulses included in the reflected signal,
extracting a Doppler profile corresponding to a distance cell of the tracking target object using a reflected signal corresponding to the phase-compensated pulse;
Frequency tunable radar. delete 3. The method of claim 2,
The preprocessor is
A frequency variable radar for generating the histogram based on the frequency of a plurality of center frequencies corresponding to each of a plurality of pulses present in a Coherent Processing Interval (CPI). 3. The method of claim 2,
The preprocessor is
A frequency variable radar for generating the histogram by accumulating a plurality of pulses included in the reflected signal more than a reference number of times. According to claim 1,
The calculator is
A frequency variable radar that normalizes the signal magnitude of the extracted Doppler profile and converts the normalized Doppler profile into a probability density function. delete According to claim 1,
The generative model is
A Gaussian Mixture Model (GMM), a frequency-tunable radar. A multi-object determination method using a frequency variable radar, comprising:
Radiating a pulse signal by randomly changing a center frequency, and receiving a reflected signal reflected from a tracking target object among the emitted pulse signal;
extracting a Doppler profile corresponding to a distance cell of the tracking target;
converting the extracted Doppler profile into a probability density function, and calculating a likelihood from the generative model by applying the value of the probability density function to the generative model; and
determining whether an object different from the tracking target object exists in the distance cell using the calculated likelihood
A multi-object determination method comprising a. 10. The method of claim 9,
Extracting the Doppler profile comprises:
generating a histogram based on the frequency of a plurality of center frequencies corresponding to each of a plurality of pulses included in the reflected signal;
selecting a center frequency corresponding to the highest frequency from the generated histogram;
Compensating for a phase between a plurality of pulses corresponding to the selected center frequency among a plurality of pulses included in the reflected signal, and
extracting a Doppler profile corresponding to a distance cell of the tracking target object using a reflected signal corresponding to the phase-compensated pulse
A multi-object determination method comprising a. delete 11. The method of claim 10,
The step of generating the histogram comprises:
and generating the histogram based on the frequency of a plurality of center frequencies corresponding to each of a plurality of pulses present in a Coherent Processing Interval (CPI). 11. The method of claim 10,
The step of generating the histogram comprises:
generating the histogram by accumulating a plurality of pulses included in the reflected signal more than a reference number of times. 10. The method of claim 9,
The step of calculating the likelihood is:
Normalizing the signal magnitude of the extracted Doppler profile and converting the normalized Doppler profile into a probability density function
A multi-object determination method comprising a. delete 10. The method of claim 9,
The generative model is
A Gaussian Mixture Model (GMM), a multi-object judgment method.

Images