KR102529179B1 - Passenger detection radar system and radar signal processing method performed thereby - Google Patents

Abstract

The present invention relates to a passenger detection radar system that detects passengers by transmitting a radar signal toward a detection target and receiving a reception signal reflected from the detection target, and a radar signal processing method performed thereby. The passenger detection radar system comprises: a data collection unit that collects ADC data received through radar reception channels; a 4D data detection unit that detects 4D data by performing primary signal processing on the collected data; a target feature detection unit that detects feature information of the target through secondary signal processing that extracts and combines new information based on the detected 4D data detection results; an artificial intelligence learning data collection unit that collects data detected by the target feature detection unit; a step-by-step artificial intelligence detection unit that performs tertiary signal processing on the results of artificial intelligence data learning; and a final target determination unit that determines the presence or absence of child passengers and the 3D spatial location thereof by using the final result of artificial intelligence learning and the result value of the 4D data detection unit. Therefore, it is possible to detect living things existing in the interior space of a vehicle, especially child passengers who are alone.

Description

Passenger detection radar system and radar signal processing method performed therein

The present invention relates to a passenger detection radar system, and more particularly, to a passenger detection radar system for detecting living beings existing in the interior space of a building or vehicle using a radar sensor, in particular, a child passenger existing alone, and a passenger detection radar system performed therein It relates to a radar signal processing method.

A radar sensor is a sensing means that transmits radio waves using microwaves and receives a partial reflection signal reflected from a target to measure distance, speed, and angle information.

Recently, radar sensors are being applied to vehicles to prevent collisions while driving and to support safe driving.

The present applicant has disclosed and registered a radar sensor technology in a number of patent documents 1 and 2 below.

Meanwhile, recently, a technology for detecting bio-signal information using a radar sensor has been developed.

In general, a bio-signal measuring device according to the prior art detects a respiratory rate or a heart rate using a contact sensor.

However, when it is difficult to use a contact sensor, such as a burn patient or an infant, a technology for detecting a biosignal using a radar sensor is being developed.

That is, among the biosignals, as the respiration signal and the heartbeat signal have periodicity, frequency modulated continuous wave (FMCW), continuous wave (CW), pulse, ultra wide band, UWB) radar sensors that use radio waves, such as radar, are used to detect periodic signal characteristics of biosignals to measure respiratory rate and heart rate.

As such, the radar sensor basically detects a target and detects biosignal information based on Doppler frequency characteristics of biosignals including respiration and heartbeat.

Meanwhile, recently, in order to prevent heat stroke accidents of children due to being left in a vehicle, CPD (Child Persense Detection), a function for detecting a child left in a vehicle, has been put into practical use.

In particular, Euro NCAP, which is in charge of vehicle evaluation in Europe, strictly regulates evaluation standards to install sensors that can directly detect the fact that children are left in cars from 2025.

A radar sensor according to the prior art generates an alarm when a creature exists within a specific range of distance, speed, and angle.

However, the radar sensor according to the prior art has a limitation in that it is impossible to distinguish when the characteristics of radar detection data for vibration or disturbance are similar to those of living organisms.

In addition, as the radar sensor according to the prior art requires a large amount of data for artificial intelligence learning in various situations, the learning time increases, and a high-performance processor for machine learning or deep learning As it has to be applied, there is a problem in that the manufacturing cost of the product increases.

Therefore, based on basic radar detection information, separate feature information is extracted and combined to create new information and used as artificial intelligence learning data to accurately grasp the situation inside the vehicle and effectively eliminate false alarms caused by vibration and disturbance. development is required.

Korean Patent Registration No. 10-1513878 (published on April 22, 2015) Republic of Korea Patent Registration No. 10-1505044 (Announced on March 24, 2015)

An object of the present invention is to solve the above problems, and a passenger detection radar system capable of detecting living things existing in the interior space of a building or vehicle, in particular, a child passenger existing alone, and a radar signal performed by the passenger detection radar system to provide a treatment method.

Another object of the present invention is a passenger detection radar system capable of detecting a child passenger by extracting and combining feature information based on basic radar detection information to create new information and using it as artificial intelligence learning data, and a radar signal performed in the passenger detection radar system to provide a treatment method.

Another object of the present invention is to provide a passenger detection radar system and a radar signal processing method performed in the passenger detection radar system capable of improving the accuracy of detection results by effectively removing false alarms caused by vibration, disturbance, and the like.

In order to achieve the above object, the passenger detection radar system according to the present invention transmits a radar signal toward a detection target and receives a received signal reflected from the detection target to detect a passenger, which is received through radar reception channels. A data collection unit that collects ADC data, a 4D data detector that detects 4D data by performing primary signal processing on the collected data, and a secondary signal processing that extracts and combines new information based on the detected 4D data detection results. A target feature detection unit detecting target feature information through a target feature detection unit, an artificial intelligence learning data collection unit collecting data detected by the target feature detection unit, and a step-by-step artificial intelligence detection unit performing tertiary signal processing of the result of artificial intelligence data learning, and It is characterized in that it includes a final target determining unit for determining the presence or absence of a child passenger and a 3D space position by using the final result of artificial intelligence learning and the resultant value of the 4D data detection unit.

In addition, in order to achieve the above object, the radar signal processing method performed in the passenger detection radar system according to the present invention detects passengers by transmitting a radar signal toward a detection target and receiving a received signal reflected from the detection target. (a) collecting ADC data received through radar reception channels in the data collection unit, (b) detecting 4D data by performing primary signal processing on the data collected by the 4D data detection unit, ( c) detecting the feature information of the target through secondary signal processing to extract and combine new information based on the 4D data detection result detected by the target feature detector, (d) the target feature in the artificial intelligence learning data collection unit Collecting data detected by the detection unit, (e) tertiary signal processing of the result of artificial intelligence data learning in the staged artificial intelligence detection unit, and (f) final result of artificial intelligence learning in the final target determination unit. and determining whether a child passenger exists and a 3D space location by utilizing the result value of the 4D data detection unit.

As described above, according to the passenger detection radar system and the radar signal processing method performed in the passenger detection radar system according to the present invention, an effect of being able to detect living organisms existing in the interior space of a vehicle, in particular, a child passenger existing alone, is obtained. .

And according to the present invention, based on the basic radar detection information, the feature information of the target is extracted and combined to create new feature information, and it is used as artificial intelligence learning data to accurately detect the presence of child passengers. .

In addition, according to the present invention, it is possible to effectively remove false alarms caused by vibration, disturbance, etc., thereby improving the accuracy of detection results.

1 is a configuration diagram of a passenger detection radar system according to a preferred embodiment of the present invention;
2 is a configuration diagram of a radar signal processing unit shown in FIG. 1;
3 is a configuration diagram of the step-by-step artificial intelligence learning unit shown in FIG. 2;
4 to 9 are diagrams for explaining a step-by-step artificial intelligence learning process, respectively;
10 is a flowchart illustrating a radar signal processing method performed in a passenger detection radar system according to a preferred embodiment of the present invention step by step.

Hereinafter, a passenger detection radar system and a radar signal processing method performed therein according to a preferred embodiment of the present invention will be described in detail with reference to the accompanying drawings.

1 is a configuration diagram of a passenger detection radar system according to a preferred embodiment of the present invention, and FIG. 2 is a configuration diagram of a radar signal processing unit shown in FIG. 1 .

In this embodiment, the biosignal may be a signal obtained by detecting breathing of the detection target, a signal detected by heartbeat, or a signal detected by simultaneously detecting breathing and heartbeat.

As shown in FIGS. 1 and 2, the radar system 10 according to a preferred embodiment of the present invention transmits a radar signal (hereinafter referred to as a 'transmission signal') toward a detection target and receives a signal reflected from the detection target. The antenna unit 11 for receiving, the radar transmitter 12 for generating the transmission signal, the radar receiver 13 for processing the data of the received signal received through the receiving antenna 22, and control to generate the transmission signal It may include a signal processing processor 14 that generates a signal and processes the received signal to detect biosignal information.

The passenger detection radar system 10 configured as described above can be applied to various radar devices using various frequency bands such as 24 GHz, 60 GHz, 77 GHz, and 79 GHz.

The antenna unit 11 may include a transmission antenna 21 for transmitting a transmission signal and a reception antenna 22 for receiving a reception signal.

The radar transmitter 12 outputs a transmission signal of a desired oscillation frequency according to the control signal of the signal processing processor 14, and the radar receiver 13 amplifies the received signal and processes the data of the received signal. there is.

The signal processing processor 14 includes a radar transmission control unit 41 that controls driving of the radar transmission unit 12, a radar reception control unit 42 that controls driving of the radar reception unit 13, and a communication unit that communicates with an external device. (43) and a radar signal processing unit 44 for detecting biosignal information through signal processing of the detected biosignal.

The radar transmission controller 41 may generate a control signal to adjust the radar waveform and transmission power (Tx Power) of the transmission signal.

The radar reception control unit 42 may generate a control signal to adjust the reception gain (Rx Gain) of the received signal and to adjust the band pad filter.

The radar reception controller 42 may include an analog-to-digital converter (hereinafter referred to as 'ADC') that converts the received signal Rx in the form of an analog signal into a digital signal.

That is, the radar reception control unit 42 includes a plurality of ADCs corresponding to the number of reception antennas 22, and each ADC may ADC-process the reception signal received through each reception antenna 22.

Meanwhile, the radar sensor may detect 4D data by detecting angle information including a distance to the target, speed of the target, and azimuth and elevation information.

The respiratory signal consists of inhalation and exhalation and has a periodic characteristic. In addition, the child's respiratory signal has a relatively small size compared to the adult's respiratory signal.

Therefore, the radar sensor can detect the 4D data by detecting a biosignal having periodic characteristics such as a breathing signal.

In the present invention, the radar system 10 generates a child presence detection alarm when a living creature exists within a specific range of distance, speed, and angle, particularly when a child exists alone.

In general, when radar detection data caused by vibrations or disturbances other than living things have characteristics similar to those of living things, false alarms increase as it becomes impossible to distinguish living targets.

Therefore, in order to reduce false alarms, the present invention utilizes artificial intelligence data learning to distinguish and determine whether there is a living being, whether a child is alone, and distinguishing from vibration or disturbance.

And instead of applying a large amount of learning data and a high-performance processor for artificial intelligence learning, the present invention applies a low-end processor and uses a small amount of learning data to accurately extract life forms, vibrations, and disturbance characteristics.

To this end, the radar signal processing unit 44 includes a data collection unit 51 that collects ADC data received through radar reception channels and a 4D data detection unit that detects 4D data by performing primary signal processing on the collected data ( 52), a target feature detector 53 that detects target feature information through secondary signal processing that extracts and combines new information based on the detected 4D data detection result, and the data detected by the target feature detector 53 The collection of artificial intelligence learning data collection unit 54, the staged artificial intelligence detection unit 55 for tertiary signal processing of the result of artificial intelligence data learning, and the final result of artificial intelligence learning and 4D data detection unit 52 and a final target determining unit 56 for determining whether a child passenger exists and a 3D spatial position by using the resulting value.

In addition, the radar signal processing unit 44 may further include a step-by-step artificial intelligence learning unit 57 that classifies the target feature information in stages by using the target feature information as an artificial intelligence input and learns a result of detecting the feature information of the differentiated target. can

The data collection unit 51 may store ADC data collected through single and multiple reception channels in the form of a 3D ADC data cube having distance, speed, and angle information as parameters.

The 4D data detector 52 performs a 4D Fast Fourier Transform (FFT) to detect 4D data including distance and speed information and angle information, that is, elevation and azimuth information, Targets are detected and multi-target clustering can be performed.

Through this, the 4D data detection unit 52 can basically detect target information that can be obtained through radar signal processing.

The target feature detection unit 53 may detect feature information of the target through secondary signal processing of extracting and combining the 4D data detected by the 4D data detection unit 52 with new information based on the detected 4D data detection result. .

For example, the target feature detection unit 53 combines distance information and speed information included in the 4D data, distance information and angle information, angle information and size information, or time information and speed information. New feature information of the target may be derived by combining.

Of course, the present invention is not necessarily limited thereto, and new feature information of a target may be derived by selectively applying two or more pieces of information included in 4D data in various types of combinations as needed.

The artificial intelligence learning data collection unit 54 collects the data detected by the target feature detection unit, provides it to the staged artificial intelligence learning unit 57, learns the data in conjunction with the staged artificial intelligence learning unit 57, and learns the data for each stage. It is possible to classify and distinguish learning data required for a function.

The step-by-step artificial intelligence learning unit 57 may apply a step-by-step artificial intelligence structure and use the target feature information detection result as an input for step-by-step artificial intelligence in order to reduce the amount of learning data and reduce the influence of interference between learning data.

Further, the step-by-step AI learning unit 57 may use the previous step AI learning result and the feature information detection result of the new target as the next step AI learning input.

In this way, as the step-by-step artificial intelligence structure is applied to the step-by-step artificial intelligence learning unit 57 in this embodiment, a new artificial intelligence function can be easily added if necessary.

So, the step-by-step artificial intelligence detection unit 55 may apply the artificial intelligence optimal parameters learned through the step-by-step artificial intelligence learning unit 57 .

For example, the step-by-step artificial intelligence detection unit 55 can apply the step-by-step functional structure of the step-by-step artificial intelligence learning unit 57 to the software implementation of a microcontroller unit (MCU).

Although the optimal parameters obtained through such artificial intelligence data learning may be performed in the radar signal processing unit 44, the load of the processor 14 for signal processing applied to the radar system 10 is minimized and the high-specification processor is applied. In order to reduce the cost, it can be applied after being performed in a computer terminal or server (not shown) outside the radar system 10. That is, the step-by-step artificial intelligence learning unit 57 is provided in an external computer terminal or server, and can be connected to the radar signal processing unit 44 through the communication unit 43 to enable transmission and reception of learning data.

The final target determination unit 56 may finally determine the existence of a child passenger and the position in the 3D space by utilizing the final result of the step-by-step artificial intelligence learning unit 57 and the detection result value of the 4D data detection unit 52. there is.

Next, a step-by-step artificial intelligence learning process will be described with reference to FIGS. 3 to 9 .

3 is a configuration diagram of the step-by-step artificial intelligence learning unit shown in FIG. 2, and FIGS. 4 to 9 are diagrams explaining the step-by-step artificial intelligence learning process, respectively.

4 illustrates the configuration of the function #1 artificial intelligence learning unit, FIG. 5 illustrates the function #1 artificial intelligence learning logic, and FIG. 6 illustrates the function #1 artificial intelligence learning result. In addition, FIG. 7 illustrates the configuration of the function #2 AI learning unit, FIG. 8 illustrates the function #2 AI learning logic, and FIG. 9 illustrates the function #2 AI learning result.

As shown in FIG. 3, the step-by-step artificial intelligence learning unit 57 defines the type of step-by-step function.

For example, the first function (hereafter referred to as 'function #1') is the presence of a target, function #2 is vibration or disturbance, function #3 is a living being, function #4 is a passenger, and function #5 is a child passenger. It can be defined as whether

Of course, the present invention is not necessarily limited to this, and the number of step-by-step functions and the definition of each function can be variously changed.

Here, as the AI learning result of each AI function, the AI result of each stage is used as it is, the AI result of each stage is additionally processed and used, or new feature information is derived by combining the AI result of each stage and target feature information. It can be utilized through the method used by the user.

In this embodiment, step-by-step function types and definitions and step-by-step AI learning results are used, and target feature information group mapping is described as being generated by comparing and analyzing collection results of the AI learning data collection unit 54.

For example, the step-by-step artificial intelligence learning unit 57 receives target feature information group #1 to target feature information group #N and performs artificial intelligence learning function #1 to function #N artificial intelligence learning units 571 to 571 57N) may be included.

Accordingly, the function #1 artificial intelligence learning unit 571 receives the target feature information group #1 and performs artificial intelligence learning, as shown in FIGS. 4 to 6 .

Function #2 artificial intelligence learning unit 572 to function #1 artificial intelligence learning unit 57N, as shown in FIGS. 7 to 9 , respectively, target feature information group #2 to target feature information group #N and previous Artificial intelligence learning may be performed by receiving a learning result from the artificial intelligence learning units 571 to 57N-1, that is, a function #1 artificial intelligence learning result or a function #N-1 artificial intelligence learning result.

Next, a radar signal processing method performed in a passenger detection radar system according to a preferred embodiment of the present invention will be described in detail with reference to FIG. 10 .

10 is a flowchart illustrating a radar signal processing method performed in a passenger detection radar system according to a preferred embodiment of the present invention step by step.

First, the radar transmission/reception unit 12 transmits a signal to the surroundings of a detection target, that is, a target to detect a biosignal, through the transmission antenna 21 according to the control signal of the radar transmission control unit 41 of the signal processing processor 14. and receives a signal reflected to the detection target through the receiving antenna 22.

Then, the ADC provided in the radar reception controller 42 converts the received signal in the form of an analog signal into a digital signal.

In step S10 of FIG. 4 , the data collection unit 51 collects ADC data for the received radar reflection signal.

In step S12, the 4D data detector 52 detects the 4D radar signal by performing primary signal processing. In this case, the 4D data detector 52 may detect multiple targets through 4D FFT and detect a basic radar target detection signal including range, speed, azimuth, elevation information and size information of the target.

In step S14, the target feature detector 53 detects target feature information through secondary signal processing to extract and combine new information based on the detected 4D data detection result. At this time, the target feature detector 53 combines distance information and speed information included in the 4D data, distance information and angle information, angle information and size information, or time information and speed information. and processing to detect new feature information of the target. Through this, the target feature detection unit 53 may collect various data for each situation, such as whether a target exists, whether there is vibration or disturbance, whether there is a living creature, whether there is a child passenger, etc., and extracts a feature signal of the target.

In step S16, the artificial intelligence learning data collection unit 54 collects and stores primary and secondary signal processing results, that is, artificial intelligence learning data.

In step S18, the step-by-step artificial intelligence learning unit 54 defines step-by-step artificial intelligence functions, creates target feature information groups, and extracts artificial intelligence optimal parameters.

Here, the step-by-step artificial intelligence learning unit 54 is provided in the signal processing processor 14 of the radar system 10 or in a computer terminal or server communicatively connected to the radar system 10, in step S16. By performing AI learning on the learning data collected in , it is possible to define AI functions for each stage, create a target feature information group, and derive AI optimal parameters.

In step S20, the staged AI detection unit 55 detects the staged AI signal through tertiary signal processing. Here, the staged artificial intelligence detection unit 55 applies the artificial intelligence optimal parameters derived through the staged artificial intelligence learning unit 57.

In step S22, the artificial intelligence detection unit 55 classifies and separates for various situations through a combination of the final result of artificial intelligence for each step and the primary signal processing result in step S12. At this time, the artificial intelligence detection unit 55 for each stage can determine the presence or absence of a target, presence or absence of vibration or disturbance, presence or absence of a living creature, presence or absence of a person, whether an adult or child passenger, etc. based on the results of the first and third signal processing. .

So, in step S24, the final target determining unit 56 checks whether there is a child passenger.

If, as a result of the inspection in step S24, there is a child passenger, the final target determining unit 56 determines the 3D location information of the vehicle seat, and generates an alarm detecting the presence of the child passenger in step S26.

After performing step S26, or when there is no child passenger as a result of the check in step S24, the radar signal processing unit 44 stops driving each device and ends the operation.

Through the process as described above, the present invention can detect living organisms existing in the interior space of a vehicle, especially child passengers existing alone.

In addition, the present invention extracts and combines target feature information based on basic radar detection information to create new feature information, and uses it as artificial intelligence learning data to accurately detect the presence of child passengers.

In addition, the present invention can improve the precision of detection results by effectively removing false alarms caused by vibration, disturbance, and the like.

On the other hand, in the above embodiment, in order to satisfy the CPD regulation for preventing heat stroke in children left in the vehicle, it has been described that a child passenger present alone in the interior space of the vehicle is detected, but the present invention is necessarily limited thereto. It is not.

For example, the present invention may be modified to detect an adult or an elderly person having different vital signal characteristics from a child passenger, or a companion animal such as a dog or cat.

In addition, the present invention may be applied not only to the interior space of a vehicle but also to the interior space of a building to detect various detection targets.

Although the invention made by the present inventors has been specifically described according to the above embodiments, the present invention is not limited to the above embodiments, and various changes can be made without departing from the gist of the present invention.

The present invention is applied to a passenger detection radar system and a radar signal processing method technology performed in the passenger detection radar system for detecting living beings existing in the interior space of a vehicle, in particular, a child passenger existing alone.

10: Radar system
11: antenna unit 12: radar transmission unit
13: radar receiver 14: processor for signal processing
21: transmit antenna 22: receive antenna
41: radar transmission control unit 42: radar reception control unit
43: communication unit 44: radar signal processing unit
51: data collection unit 52: 4D data detection unit
53: target feature detection unit 54: artificial intelligence learning data collection unit
55: step-by-step artificial intelligence detection unit 56: final target determination unit
57: step-by-step artificial intelligence learning unit
571 to 57N: function #1 to function #N artificial intelligence learning unit

Claims (10)

In a passenger detection radar system that detects a passenger by transmitting a radar signal toward a detection target and receiving a reception signal reflected from the detection target,
A data collection unit for collecting ADC data received through radar reception channels;
A 4D data detector detecting 4D data by performing primary signal processing on the collected data;
A target feature detector detecting target feature information through secondary signal processing that extracts and combines new information based on the detected 4D data detection result;
An artificial intelligence learning data collection unit for collecting data detected by the target feature detection unit;
A step-by-step artificial intelligence detection unit that processes the result of artificial intelligence data learning into a tertiary signal, and
And a final target determination unit for determining the presence or absence of a child passenger and a 3D spatial location using the final result of artificial intelligence learning and the resultant value of the 4D data detection unit,
The 4D data detection unit detects multiple targets by performing 4D fast Fourier transform to detect 4D data including distance and speed information and elevation and azimuth information, and performs multiple target clustering to obtain target information through radar signal processing detect,
The passenger detection radar system, characterized in that the target feature detector derives new feature information of the target by selectively combining two or more pieces of information included in the 4D data. According to claim 1,
Further comprising a step-by-step artificial intelligence learning unit that classifies the target feature information in stages by using the target feature information as an artificial intelligence input and learns a result of detecting the feature information of the distinguished target;
The artificial intelligence learning data collection unit learns data in conjunction with the step-by-step artificial intelligence learning unit, and classifies and distinguishes the learning data required for each step function. According to claim 2,
The step-by-step AI learning unit applies a step-by-step AI structure and uses the target feature information detection result as an input for step-by-step AI, and inputs the previous step AI learning result and the new target feature information detection result to the next step AI learning input. Passenger detection radar system, characterized in that for use as. delete According to claim 2,
The step-by-step artificial intelligence detection unit applies the parameters learned through the step-by-step artificial intelligence learning unit to determine whether there is a target, whether there is vibration or disturbance, whether there is a living being, whether there is a person, whether an adult or child passenger is present,
The passenger detection radar system, characterized in that the final target determiner determines the three-dimensional position information of the vehicle seat and generates an alarm detecting the presence of a child passenger when it is determined that a child passenger exists. A radar signal processing method performed in a passenger detection radar system for detecting a passenger by transmitting a radar signal toward a detection target and receiving a reception signal reflected from the detection target,
(a) collecting ADC data received through radar reception channels in a data collection unit;
(b) detecting 4D data by performing primary signal processing on the data collected by the 4D data detector;
(c) detecting target feature information through secondary signal processing to extract and combine new information based on the 4D data detection result detected by the target feature detector;
(d) collecting data detected by the target feature detection unit in an artificial intelligence learning data collection unit;
(e) tertiary signal processing of the result of artificial intelligence data learning in the staged artificial intelligence detection unit, and
(f) determining whether a child passenger exists and a 3D spatial location by using the final result of artificial intelligence learning performed by the final target determination unit and the resultant value of the 4D data detection unit;
In step (b), the 4D data detection unit detects multiple targets by performing 4D fast Fourier transform to detect 4D data including distance and speed information and elevation and azimuth information, and performs multiple target clustering to obtain radar signals. detect target information obtained through processing;
In the step (c), the target feature detection unit selectively combines two or more pieces of information included in the 4D data to derive new feature information of the target. Radar signal processing method performed in a passenger detection radar system. According to claim 6,
(g) further comprising the step of using the target feature information as an artificial intelligence input in a step-by-step artificial intelligence learning unit to classify the target feature information in stages, and learning a result of detecting the feature information of the differentiated target;
In the step (d), the artificial intelligence learning data collection unit learns data in conjunction with the step-by-step artificial intelligence learning unit, and classifies and distinguishes the learning data required for each step function. Radar signal processing method. According to claim 7,
In the step (g), the step-by-step artificial intelligence learning unit applies a step-by-step artificial intelligence structure and uses the result of detecting feature information of the target as an input of step-by-step artificial intelligence to obtain the result of AI learning at the previous step and the feature information detection result of the new target. A radar signal processing method performed in a passenger detection radar system, characterized in that it is used as a next step artificial intelligence learning input. delete According to claim 7,
In the step (d), the step-by-step artificial intelligence detection unit applies the parameters learned through the step-by-step artificial intelligence learning unit to determine whether there is a target, whether there is vibration or disturbance, whether there is a living being, whether there is a person, whether an adult or child passenger is present, ,
In the step (f), when the final target determining unit determines that a child passenger exists, the passenger detection radar system determines the 3D location information of the vehicle seat and generates an alarm detecting the presence of the child passenger. Radar signal processing method performed in.

Images