CN117911993A

CN117911993A - Passenger detection method and vehicle

Info

Publication number: CN117911993A
Application number: CN202410068222.0A
Authority: CN
Inventors: 袁鑫豪; 郭茂荣; 缪颖; 裴广宇
Original assignee: BYD Co Ltd
Current assignee: BYD Co Ltd
Priority date: 2024-01-16
Filing date: 2024-01-16
Publication date: 2024-04-19

Abstract

The disclosure relates to a passenger detection method and a vehicle, and relates to the technical field of vehicles, comprising the following steps: acquiring a point cloud data set of a target to be identified; determining a target center point of the target to be identified according to the point cloud data set; and determining that the target to be identified does not exist in the target seat area under the condition that the target center point of the target to be identified is positioned outside the target seat area. By using the passenger detection method and the vehicle, the occurrence of the situation that people exist in the misjudgment target seat area can be reduced, false triggering of an alarm is further reduced, and the use experience of a user is improved.

Description

Passenger detection method and vehicle

Technical Field

The present disclosure relates to the field of vehicle technologies, and in particular, to a passenger detection method and a vehicle.

Background

Along with the continuous improvement of the importance of people on environmental protection and low carbon, the development pace of the new energy automobile is obviously accelerated. The related technologies of the fields of automobile, energy, traffic, information communication and the like are combined in an accelerating way, and the development trend and trend of the automobile industry are achieved through electric technology, networking technology and intelligent technology. New energy automobiles emerge as a spring bamboo shoot after rain, for example: the application number is CN202110228793.2, and the invention is a hybrid power system, a hybrid power vehicle and a control method thereof; the application number is CN202110231482.1, and the invention is a hybrid power system, a hybrid power vehicle, a control method thereof and a whole vehicle controller; the application number is CN202110251139.3, and the invention is a hybrid power system, a hybrid power vehicle, a control method thereof and a whole vehicle controller; the application number is CN202010609385.7, and the invention name is a power generation control method and device of a vehicle and the vehicle; all describe the electricity-based hybrid technology, and have multiple advantages of rapidness, province, quietness, smoothness, greenness and the like. The application number is CN202311373582.3, and the invention is a vehicle control method, a medium and a vehicle; application number CN202311351004.X, the invention is a vehicle braking method, a brake controller, a storage medium and a vehicle; the application number is CN202311222836.1, and the invention is a parking method, a medium and a vehicle; the application number is CN202311164098.X, the invention is named vehicle control system, method and vehicle; the application number is CN202311170393.6, and the invention is a vehicle control system, a control method, a controller and a vehicle; the application number is CN202211678720.4, and the invention is a power control method, a device, a medium, a vehicle controller and a vehicle of the vehicle; the application number is CN202211469445.5, and the invention is a vehicle, a control method and device thereof, a medium and electronic equipment; the application number is CN202210182808.0, and the invention is a tire pressure identification device and a vehicle; the application number is CN202110744962.8, and the invention is an electric drive assembly, a four-wheel drive system and an automobile; the application number is CN202110474242.4, and the invention is a steering control method, an electronic control unit, a system and a vehicle for the vehicle; the application number is CN202110351399.8, and the invention is a brake control method, a brake control device, a brake control medium and electronic equipment; the four wheel motors are independently driven to serve as a core power system, so that the safety and the dynamic performance of the new energy automobile are greatly improved.

Currently, a seatbelt unbuckled reminding function ((safety belt reminder, SBR)) is provided on a vehicle for giving an alarm to warn a passenger that the passenger is belted when the passenger is seated in a cabin during running of the vehicle but the passenger is not belted.

However, the seats in the vehicle cabin are relatively compact, and if the passenger approaches the edge of the partition seat or the passenger's limb swing is large, the passenger on the partition seat is erroneously detected, and an alarm of the partition seat is erroneously triggered.

Disclosure of Invention

An object of the present disclosure is to provide a passenger detection method and a vehicle, so as to solve the above technical problems.

To achieve the above object, a first aspect of the present disclosure provides a passenger detection method, including:

Acquiring a point cloud data set of a target to be identified;

Determining a target center point of the target to be identified according to the point cloud data set;

and determining that the target to be identified does not exist in the target seat area under the condition that the target center point of the target to be identified is positioned outside the target seat area.

Optionally, the method further comprises:

And under the condition that the target center point is positioned in the target seat area, detecting the object in the target seat area according to the relation between the average height of the point clouds of the point cloud data set and the height threshold value and the relation between the average density of the point clouds of the point cloud data set and the density threshold value.

Optionally, the density threshold comprises a first density threshold; and performing object detection on the target seat area according to the relation between the average height of the point cloud data set and a height threshold value and the relation between the average density of the point cloud data set and a density threshold value under the condition that the target center point is located in the target seat area, wherein the object detection comprises:

And determining that the target to be identified exists in the target seat area under the condition that the average height of the point cloud is larger than the height threshold value and the average density of the point cloud is larger than the first density threshold value.

Optionally, the density threshold comprises a second density threshold; and performing object detection on the target seat area according to the relation between the average height of the point cloud data set and a height threshold value and the relation between the average density of the point cloud data set and a density threshold value under the condition that the target center point is located in the target seat area, wherein the object detection comprises:

and determining that the target to be identified exists in the target seat area under the condition that the average height of the point cloud is smaller than the height threshold value and the average density of the point cloud is larger than the second density threshold value.

Optionally, the density threshold includes a first density threshold and a second density threshold; and performing object detection on the target seat area according to the relation between the average height of the point cloud data set and a height threshold value and the relation between the average density of the point cloud data set and a density threshold value under the condition that the target center point is located in the target seat area, wherein the object detection comprises:

And determining that the target to be identified does not exist in the target seat area when the average height of the point cloud is larger than the height threshold value and the average density of the point cloud is smaller than the first density threshold value and/or when the average height of the point cloud is smaller than the height threshold value and the average density of the point cloud is smaller than the second density threshold value.

Optionally, the density threshold is determined by:

Determining a point cloud average density of the historical point cloud dataset;

Correcting the average density of the point cloud by adopting a correction coefficient to obtain the density threshold; the density threshold is less than the point cloud average density.

Optionally, the determining, according to the point cloud data set, a target center point of the target to be identified includes:

projecting the point cloud data set on a two-dimensional plane to obtain a two-dimensional area;

and taking the center point of the two-dimensional area as the target center point.

Optionally, the acquiring the point cloud data set of the object to be identified includes:

Determining a point cloud data set of the target object;

Screening out a target point cloud closest to a seat center point of the target seat area from a point cloud data set of the target object;

and taking the point cloud data set of the target point cloud as the point cloud data set of the target to be identified.

Optionally, the determining the point cloud data set of the target object includes:

detecting point cloud data of a target object in the vehicle cabin through a radar detection device;

and clustering the plurality of point cloud data to obtain a point cloud data set of the target object.

In order to achieve the above object, a second aspect of the present disclosure provides a vehicle provided with a controller for performing the steps of the occupant monitoring method provided in the first aspect of the present disclosure.

According to the technical scheme, although the situation that the swing amplitude of the limb of the target to be identified is large or the target to be identified touches the edge of the adjacent target seat area possibly exists, the center point of the target to be identified is positioned outside the target seat area, so that the fact that the target to be identified does not exist in the target seat area can be determined, the alarm of the target seat area can not be triggered by mistake, and the user experience is improved.

Additional features and advantages of the present disclosure will be set forth in the detailed description which follows.

Drawings

The accompanying drawings are included to provide a further understanding of the disclosure, and are incorporated in and constitute a part of this specification, illustrate the disclosure and together with the description serve to explain, but do not limit the disclosure. In the drawings:

fig. 1 is a flowchart illustrating steps of a passenger detection method according to an exemplary embodiment.

FIG. 2 is a schematic logic diagram illustrating a detection of a passenger according to an example embodiment.

Fig. 3 is a schematic diagram illustrating preprocessing of point cloud data monitored by a radar detection apparatus according to an exemplary embodiment.

Fig. 4 is a schematic diagram showing various seats in a vehicle cabin and a vehicle body coordinate system according to an exemplary embodiment.

FIG. 5 is a schematic diagram illustrating a density threshold to obtain a point cloud, according to an example embodiment.

Fig. 6 is a logic diagram illustrating a passenger detection method according to an example embodiment.

Fig. 7 is a block diagram illustrating a passenger detection device according to an example embodiment.

Detailed Description

Specific embodiments of the present disclosure are described in detail below with reference to the accompanying drawings. It should be understood that the detailed description and specific examples, while indicating and illustrating the disclosure, are not intended to limit the disclosure.

In the related art, referring to fig. 2, the workflow of the seat belt unbuckled reminding function is as follows: firstly, detecting whether the vehicle speed is greater than 0, and under the condition that the vehicle speed is greater than 0, determining whether passengers exist on the current seat, if so, determining whether a safety belt bolt on the current seat is disconnected, if so, giving an alarm, and if not, giving an alarm.

In this solution, since the seats in the cabin are relatively compact, for example, as shown in fig. 4, three seats in the rear row are adjacent, and if the middle passenger approaches the edges of the left and right seats or the swing amplitude of the limbs of the middle passenger is large, the left and right seats may be caused to trigger an alarm by mistake.

Based on this, the present disclosure proposes a passenger detection method for use in a vehicle, comprising the steps of:

In step S21, a point cloud data set of the object to be identified is acquired.

The radar detection device can be used for detecting the point cloud data of the target object in the vehicle cabin, and clustering is carried out on the plurality of point cloud data to obtain a point cloud data set of the target object. The radar detection device can be a 4D millimeter wave radar, the millimeter wave radar can work all weather and all day long, and the millimeter wave radar can not cause the privacy leakage of a user like a camera, so that the radar detection device is widely applied in the automotive field, for example, the millimeter wave radar can be applied to the functions of a child legacy monitoring (CHILD PRESENCE detection (CPD), a safety belt unblinding reminding, an automobile intrusion warning (intruder alert, IA), an airbag limiting opening (FMVSS), a breath and heartbeat monitoring function and the like.

Referring to fig. 3, the flow of obtaining the point cloud data of the target object is as follows:

(1) Under the receiving and transmitting of a frame of the radar detection device, the receiving antenna on the radar detection device collects original ADC (analog-to-Digital Converter, conversion from analog signal to digital signal) data on each channel, and the ADC data is data of a digital signal and is represented by a series of numbers.

(2) The ADC data is compressed in the distance and speed dimensions through 2D-FFF (two-dimensional fast Fourier transform), which can be understood as converting the ADC data on each channel into the frequency domain, and obtaining the spectrum information of the ADC data in the distance dimension and the speed dimension through Fourier transform.

(3) Incoherent accumulation is carried out in the channel dimension, and the target object can be detected after the accumulated data matrix is detected by a Constant False Alarm (CFAR) algorithm. Uncorrelated accumulation refers to the incoherent superposition of multiple received signals.

(4) And carrying out two-dimensional angle estimation (2D-DOA) on the targets to obtain the horizontal angle and the pitching angle of the targets, and forming the point cloud data of the targets by the distance, the speed, the horizontal angle and the pitching angle of the targets. The two-dimensional angle estimate is used to determine the horizontal angle and the pitch angle of the target, the horizontal angle being the horizontal angle of the target relative to the receiving antenna.

After the point cloud data of the target object is obtained, the point cloud data of the target object is detected by the radar detection device, so that the point cloud data of the target object is positioned under a radar coordinate system where the radar detection device is positioned; the position of the seat is obtained based on the vehicle body coordinate system, so that coordinate system conversion is required to be carried out on the point cloud data of the target object, and the point cloud data of the target object is converted into the vehicle body coordinate system from the radar coordinate system, so that the position of the seat and the point cloud data of the target object are located under the same vehicle body coordinate system.

Please refer to fig. 4, it is assumed that the vehicle is a five-seat passenger car, the radar detection device may be mounted on a roof skylight side pillar, the radar detection device irradiates toward the interior of the cabin, and of course, the mounting position of the radar detection device may be disposed at other positions in the cabin.

The vehicle body coordinate system takes the center of the vehicle as the origin of coordinates, takes the vertical ground as the Z axis, takes the direction of the first central axis of the vehicle pointing to the vehicle head as the X axis, and takes the direction of the second central axis of the vehicle pointing to the side of the vehicle as the Y axis. Under the vehicle body coordinate system, five seat areas as shown in fig. 4 can be constructed according to the size and the position of each seat in the vehicle cabin, and the seat center of each seat area can be obtained. And calibrating the coordinate conversion relation between the vehicle body coordinate system and the radar coordinate system to obtain the coordinate conversion relation between the vehicle body coordinate system and the radar coordinate system. And finally, according to the coordinate conversion relation and the point cloud data under the radar coordinate system, obtaining the point cloud data under the vehicle body coordinate system. It will be appreciated that there are a variety of ways to calibrate the coordinate conversion relationship between the vehicle body coordinate system and the radar coordinate system, and this disclosure is not limited in this regard.

The target object refers to a plurality of objects which are detected by the radar detection device and are positioned in the vehicle cabin, the target object to be identified is contained in the target object, and the target object to be identified refers to a target object which is positioned in the vehicle cabin and is close to a target seat area.

After the point cloud data of the target objects are obtained, the point cloud data of at least one target object can be clustered, the point cloud data of the same target object can be clustered into one cluster, and if N clusters are obtained, N target objects are corresponding, and the point cloud data sets of the target objects comprise the point cloud data sets of the target to be identified.

The clustering method aims at clustering point cloud data into different clusters, wherein the point cloud data of the different clusters are divided into different targets, and the point cloud data of the same cluster belong to the same target. The clustering may be a DBSCAN (density clustering algorithm) clustering method, which is not limited by the present disclosure. And after clustering, edge point clouds which are far away from the average value of the point cloud data in the cluster can be screened out, so that the accuracy of the cluster is improved, and then the data of the point cloud density average value and the point cloud height average value, which are obtained by subsequent calculation based on the point cloud data set in the cluster, can be more accurate.

In step S22, a target center point of the target to be identified is determined according to the point cloud data set.

After the point cloud data set of the target to be identified is obtained, a target center point of the target to be identified can be found out from the plurality of point cloud data, and the target center point represents the actual position of the target to be identified.

It will be appreciated that if the object to be identified is a passenger, then the center point of the object is the center point of the trunk of the passenger, and not the center points of the limbs of the passenger.

In step S23, in the case where the target center point of the target to be identified is located outside the target seating area, it is determined that the target to be identified is not present in the target seating area.

In the case that the target center point of the target to be identified is located outside the target seat area, it is determined that the target to be identified does not exist in the target seat area.

Referring to fig. 4, assuming that the middle seat is the current position of the passenger, the seats on the left and right sides are the target seat areas, when the passenger touches the edges of the left and right seats of the partition wall or the swing amplitude of the limbs of the passenger is large, since the target center point of the passenger is located inside the middle seat and outside the target seat areas on the left and right sides, it can be determined that the passenger is not present on the seats on the left and right sides, and an alarm is not triggered by mistake.

According to the technical scheme, although the limb swing amplitude of the object to be identified is large or the object to be identified touches the edge of the adjacent object seat area, the center point of the object to be identified is positioned outside the object seat area, so that the fact that the object to be identified does not exist in the object seat area can be determined, the alarm of the object seat area can not be triggered by mistake, and the use experience of a user is improved.

A specific embodiment related to the above step S11 is described below, which is used to explain how to obtain the point cloud data set of the object to be identified.

(1) A point cloud dataset of the target is determined.

The point cloud data of at least one target object in the vehicle cabin can be detected through the radar detection device, and the point cloud data are clustered after being preprocessed, so that a point cloud data set of the at least one target object is obtained. The preprocessing step includes the ADC conversion, 2D-FFT processing, incoherent accumulation, and the like shown in fig. 3.

The object is an object in the vehicle cabin detected by the radar detection device, and the object may be a passenger, an object placed by the passenger, a pet, or the like.

(2) And screening out the target point cloud closest to the seat center point of the target seat area from the point cloud data set of the target object.

The seat center point is the intersection between the vertical line half the length and half the width of the target seat area.

There may be multiple targets in the cabin, and thus the cloud of target points closest to the seat center point needs to be screened from the multiple targets.

For example, referring to fig. 4, if the target seat area is the middle seat in fig. 4, the target point cloud closest to the seat center point of the middle seat needs to be screened from the plurality of point cloud data detected by the radar detection device.

(3) And taking the point cloud data set of the target point cloud as the point cloud data set of the target to be identified.

The point cloud data set where the target point cloud is located represents the target object closest to the seat center point of the target seat area, and thus the point cloud data set closest to the seat center point can be taken as the point cloud data set of the target to be identified to be compared in the present disclosure.

After the target to be identified closest to the target seat area is determined from the plurality of targets, the target center point of the target to be identified can be compared with the target seat area without comparing the center points of all the targets with the target seat area, so that whether passengers exist in the target seat area can be rapidly judged.

According to the technical scheme, the point cloud data set closest to the seat center point of the target seat area can be screened out from the plurality of point cloud data sets detected by the radar detection device, so that whether the target to be identified is located in the target seat area or not can be judged conveniently. It will be appreciated that since the object to be identified is closest to the seat center point, the distance between the remaining objects and the seat center is greater than the distance between the object to be identified and the seat center, if the object to be identified is located outside the target seat area, the remaining objects other than the object to be identified should also be located outside the target seat area.

A specific embodiment of how to obtain the target center point of the target to be identified will be described below with reference to the above step S12.

(1) And projecting the point cloud data set on a two-dimensional plane to obtain a two-dimensional area.

Referring to fig. 4, a point cloud data set of an object to be identified may be projected onto a two-dimensional plane of X-Y to obtain a horizontal two-dimensional area corresponding to the object to be identified.

It will be appreciated that the target seating area is in fact a horizontal two-dimensional area, so the present disclosure requires that the point cloud data set also be projected onto the horizontal two-dimensional plane of X-Y in order to compare the point cloud data set to the target seating area in the same dimension.

(2) And taking the center point of the two-dimensional area as the target center point.

An intersection point between a center point vertical line of the length of the two-dimensional region and a center point vertical line of the width may be taken as a center point of the two-dimensional region.

After the target center point of the target to be identified is obtained, it may be determined whether the target center point is located within the target seating area, thereby determining whether a passenger is present on the target seating area.

Through the technical scheme, the center point of the projection of the point cloud data set of the target to be identified on the two-dimensional plane can be used as the target center point of the target to be identified, so that the target to be identified and the target seat area are placed on the same plane for comparison, and whether the target to be identified is located in the target seat area or not can be judged more accurately.

An alternative embodiment of the present disclosure is described below, with reference to fig. 6, which is used to explain how to accurately determine whether an object to be identified within a target seating area is a person or an item, in the case where the center point of the object to be identified is located within the target seating area.

Before object detection is performed on the target seat area according to the relation between the point cloud average height and the height threshold value and the relation between the point cloud average density and the density threshold value, statistical analysis processing is required to be performed on the point cloud data to obtain parameters such as the height threshold value, the density threshold value and the like.

For the density threshold, referring to fig. 5, the average density of the point clouds of the historical point cloud dataset may be determined; correcting the average density of the point cloud by adopting a correction coefficient to obtain the density threshold; the density threshold is less than the point cloud average density.

Passengers of different age groups correspond to different density thresholds, passengers of a first age group correspond to a first density threshold, and passengers of a second age group correspond to a second density threshold. The first age group is greater than the second age group, e.g., the first age group is an adult and the second age group is a child; the first density threshold is greater than the second density threshold.

A large amount of historical point cloud data can be acquired through the radar detection device, and the average density of the point clouds of the passengers in the first age group is counted to be S1, and the average density of the point clouds of the passengers in the second age group is counted to be S2. Then the first density threshold may be set to be m times the average density of the point cloud of the passenger in the first age group, the second density threshold may be set to be m times the average density of the point cloud of the passenger in the second age group, and m may be less than 0.5, or may be other values, which the present disclosure is not limited to. The calculation formulas of the first density threshold and the second density threshold are as follows:

S3＝m*S1

S4＝m*S2

Wherein S3 is a first density threshold, S1 is a point cloud average density of the passenger in the first age group; s4 is a second density threshold, S2 is the average density of the point cloud of the passenger of the second age group, and m is a correction coefficient.

For example, it may be counted that the average density of the point cloud of the adult is S1, and then m times the average density of the point cloud of the adult S1 is used as the first density threshold value to compare with the average density of the point cloud of the subsequent adult; the average density of the point clouds of the child can be counted to be S2, and then m times of the average density S2 of the point clouds of the child is used as a second density threshold value to be compared with the average density of the point clouds of the subsequent child.

The reason why m times the average density of the point cloud is used as the density threshold is that: in an actual scene, some passengers such as a slim adult or a child exist, and the average density of the point clouds of the passengers is lower than the average density of the point clouds obtained by statistics in the past, so if the average density of the point clouds of the past history point cloud data set is set as a density threshold value, the passengers with the smaller point cloud density cannot exceed the density threshold value, and people in the target seat area cannot be misjudged. Therefore, the density threshold value can be set to be lower than the average density of the point clouds of the historical point cloud data set obtained through statistics in the past, and because m is smaller than 0.5, the density threshold value can be caused to be smaller than the average density of the point clouds of the historical point cloud data set, so that passengers with smaller density can be detected, and the occurrence of unmanned situations on a misjudgment target seat area is reduced as much as possible.

It will be appreciated that the above-described distinction between passengers of a first age group and passengers of a second age group may be made in any of the following ways: a, dividing according to heights, taking passengers with the length of 140cm or more as passengers in a first age group, and taking passengers with the length of 140cm or less as passengers in a second age group; and B, taking the passengers with the age of more than 13 as passengers in a first age group and taking the passengers with the age of less than or equal to 13 as passengers in a second age group according to the age classification.

For the height threshold, referring to fig. 5, the average height of the point cloud of the passenger in the first age group and the average height of the point cloud of the passenger in the second age group may be determined first, and then the average value of the average height of the point cloud of the passenger in the first age group and the average height of the point cloud of the passenger in the second age group is used as the height threshold.

And detecting the object in the target seat area according to the relation between the average height of the point cloud data set and the height threshold value and the relation between the average density of the point cloud data set and the density threshold value, including the following situations, please refer to fig. 6.

In the first case, determining that the target to be identified exists in the target seat area when the average height of the point cloud is greater than the height threshold and the average density of the point cloud is greater than the first density threshold.

And under the condition that the average height of the point cloud of the current object to be identified is larger than the height threshold value, the object to be identified is a passenger or a higher object in the first age group. In this case, whether the average density of the point cloud is greater than the first density threshold may be further differentiated, and if the average density of the point cloud is greater than the first density threshold, which indicates that the target to be identified is determined to be a passenger of the first age group, and not a higher item, it is determined that the passenger of the first age group exists in the target seat area.

For example, when the average height of the point cloud of the current object to be identified is greater than the height threshold and the average density of the point cloud is greater than the first density threshold, indicating that an adult is on the target seating area, the vehicle may issue an alarm to alert the occupant to belting.

In the second case, determining that the target to be identified exists in the target seat area when the average height of the point cloud is smaller than the height threshold and the average density of the point cloud is larger than the second density threshold.

And under the condition that the average height of the point cloud of the current object to be identified is smaller than the height threshold value, the object to be identified is a passenger or a lower object in the second age group. In this case, whether the average density of the point cloud is greater than the second density threshold may be further differentiated, and if the average density of the point cloud is greater than the second density threshold, which indicates that the target to be identified is determined to be a passenger of the second age group, and not a lower item, it is determined that a passenger of the second age group exists in the target seat area.

For example, when the average height of the point cloud of the current target to be identified is less than the height threshold and the average density of the point cloud is greater than the second density threshold, indicating that a child is on the target seating area, the vehicle may issue an alarm to alert the occupant to belting.

In a third case, determining that the target to be identified does not exist in the target seat area when the average height of the point cloud is greater than the height threshold and the average density of the point cloud is less than the first density threshold, and/or when the average height of the point cloud is less than the height threshold and the average density of the point cloud is less than the second density threshold.

And if the average point cloud density is smaller than the first density threshold, indicating that no passenger exists in the target seat area.

And under the condition that the average height of the point cloud is smaller than the height threshold value, the point cloud average density is larger than the second density threshold value, the fact that the object to be identified is the passenger can be indicated, and if the point cloud average density is smaller than the second density threshold value, the fact that no passenger exists in the target seat area is indicated.

In the related art, the pressure sensor is arranged under the saddle, the number of the pressure sensors is the number of seats in the cabin, the pressure sensor monitors the weight on the seat, and the alarm is triggered under the condition that the safety belt bolt is disconnected, however, the pressure sensor cannot identify whether the weight on the seat is a person or sundries, and false alarm can be easily caused.

Through the technical scheme, the method for comparing the average height of the point cloud with the height threshold value divides the object to be identified into the passenger in the first age group and the passenger in the second age group, compares the passenger in the first age group with the first density threshold value corresponding to the first age group, compares the passenger in the second age group with the second density threshold value corresponding to the second age group, and compares the passenger in the different age groups with the density threshold value in the different age groups can more accurately determine whether the passenger or other articles exist in the object seat area, so that the passenger and the articles are accurately distinguished, and the false triggering of the alarm function is avoided.

Fig. 7 is a block diagram of a passenger detection device according to an exemplary embodiment of the present disclosure, the passenger detection device 700 including: the acquisition module 710, the target center point determination module 720, and the identification module 730.

An acquisition module 710 configured to acquire a point cloud dataset of an object to be identified;

A target center point determination module 720 configured to determine a target center point of the target to be identified according to the point cloud data set;

and an identification module 730 configured to determine that the target to be identified is not present in the target seating area if the target center point of the target to be identified is located outside the target seating area.

Optionally, the passenger detection device 700 includes:

And the detection module is configured to detect an object in the target seat area according to the relation between the average height of the point clouds of the point cloud data set and a height threshold value and the relation between the average density of the point clouds of the point cloud data set and a density threshold value under the condition that the target center point is located in the target seat area.

Optionally, the density threshold comprises a first density threshold; the detection module comprises:

A first detection sub-module configured to determine that the target to be identified exists in the target seat area if the point cloud average height is greater than the height threshold and the point cloud average density is greater than the first density threshold.

Optionally, the density threshold comprises a second density threshold; the detection module comprises:

And the second detection submodule is configured to determine that the target to be identified exists in the target seat area when the average height of the point cloud is smaller than the height threshold value and the average density of the point cloud is larger than the second density threshold value.

Optionally, the density threshold includes a first density threshold and a second density threshold; the detection module comprises:

and a third detection sub-module configured to determine that the target to be identified does not exist in the target seat area when the average height of the point cloud is greater than the height threshold and the average density of the point cloud is less than the first density threshold, and/or when the average height of the point cloud is less than the height threshold and the average density of the point cloud is less than the second density threshold.

Optionally, the passenger detection device 700 includes:

a point cloud average density determination module configured to determine a point cloud average density of the historical point cloud dataset;

the correction module is configured to correct the point cloud average density by adopting a correction coefficient to obtain the density threshold; the density threshold is less than the point cloud average density.

Optionally, the target center point determination module 720 includes:

A projection sub-module configured to project the point cloud dataset on a two-dimensional plane to obtain a two-dimensional area;

a target center point determination submodule configured to take a center point of the two-dimensional region as the target center point.

Optionally, the acquiring module 710 includes:

A point cloud data set determination sub-module configured to determine a point cloud data set of the target object;

A screening sub-module configured to screen out a target point cloud closest to a seat center point of the target seat area from a point cloud dataset of the target object;

and the acquisition sub-module is configured to take the point cloud data set where the target point cloud is located as the point cloud data set of the target to be identified.

Optionally, the point cloud data set determination submodule includes:

the monitoring sub-module is configured to detect point cloud data of a target object in the vehicle cabin through the radar detection device;

And the clustering sub-module is configured to cluster the plurality of point cloud data to obtain a point cloud data set of the target object.

Based on the same inventive concept, the present disclosure also proposes a vehicle on which a controller, a seat, and a radar detection device are disposed; the radar monitoring device is used for detecting point cloud data sets of a plurality of targets in the cabin to be provided for processing by a controller, and the controller is used for executing the steps of the passenger detection method. And judging whether passengers exist in the seat.

The preferred embodiments of the present disclosure have been described in detail above with reference to the accompanying drawings, but the present disclosure is not limited to the specific details of the embodiments described above, and various simple modifications may be made to the technical solutions of the present disclosure within the scope of the technical concept of the present disclosure, and all the simple modifications belong to the protection scope of the present disclosure.

In addition, the specific features described in the foregoing embodiments may be combined in any suitable manner, and in order to avoid unnecessary repetition, the present disclosure does not further describe various possible combinations.

Moreover, any combination between the various embodiments of the present disclosure is possible as long as it does not depart from the spirit of the present disclosure, which should also be construed as the disclosure of the present disclosure.

Claims

1. A passenger detection method, comprising:

Acquiring a point cloud data set of a target to be identified;

2. The method according to claim 1, wherein the method further comprises:

3. The method of claim 2, wherein the density threshold comprises a first density threshold; and performing object detection on the target seat area according to the relation between the average height of the point cloud data set and a height threshold value and the relation between the average density of the point cloud data set and a density threshold value under the condition that the target center point is located in the target seat area, wherein the object detection comprises:

4. The method of claim 2, wherein the density threshold comprises a second density threshold; and performing object detection on the target seat area according to the relation between the average height of the point cloud data set and a height threshold value and the relation between the average density of the point cloud data set and a density threshold value under the condition that the target center point is located in the target seat area, wherein the object detection comprises:

5. The method of claim 2, wherein the density threshold comprises a first density threshold and a second density threshold; and performing object detection on the target seat area according to the relation between the average height of the point cloud data set and a height threshold value and the relation between the average density of the point cloud data set and a density threshold value under the condition that the target center point is located in the target seat area, wherein the object detection comprises:

6. The method according to any one of claims 2 to 5, wherein the density threshold is determined by:

7. The method of claim 1, wherein the determining the target center point of the target to be identified from the point cloud dataset comprises:

8. The method of claim 1, wherein the acquiring the point cloud dataset of the object to be identified comprises:

Determining a point cloud data set of the target object;

9. The method of claim 8, wherein the determining the point cloud dataset of the target object comprises:

10. A vehicle, characterized in that the vehicle is provided with a controller for performing the steps of the method according to any one of claims 1-9.