Vehicle occupant detector

Abstract

To enable the number of vehicle occupants to be identified with high accuracy.SOLUTION: When a plurality of occupants are detected at a first time, then at a second time immediately following the first time, a vehicle occupant detector multiplies a first ratio to the maximum peak value of one or more peaks identified for a point cloud at the second time to calculate a first threshold, and counts peaks having peak values greater than or equal to the first threshold among the one or more peaks as the number of occupants. When one occupant is detected at the first time, then at the second time, the vehicle occupant detector multiplies a second ratio higher than the first ratio to the maximum peak value of one or more peaks identified for the point cloud at the second time to calculate a second threshold, and counts peaks having peak values greater than or equal to the second threshold among the one or more peaks identified at the second time as the number of occupants.SELECTED DRAWING: Figure 7

Description

[Technical Field]

[0001] FIELD An embodiment of the present invention relates to an occupant detection device. [Background technology]

[0002] In vehicle control systems and the like, there are technologies for detecting occupants in a vehicle cabin based on data acquired by radio wave sensors and the like installed in the vehicle cabin. For example, the technology in Patent Document 1 claims that by installing radio wave sensors in appropriate positions in the vehicle cabin, occupants can be accurately detected even in the rear seats, which are used for multiple purposes. [Prior art documents] [Patent documents]

[0003] [Patent Document 1] Japanese Patent Application Publication No. 2018-202921 Summary of the Invention [Problem to be solved by the invention]

[0004] However, the technology of Patent Document 1 may erroneously detect one occupant as multiple occupants when, for example, the occupant is seated in an unsteady posture.

[0005] The present invention has been made in view of the above, and has as its object to provide an occupant detection device that can identify the number of occupants with high accuracy. [Means for solving the problem]

[0006] The occupant detection device of the embodiment includes an acquisition unit that acquires point cloud information showing one or more detection points representing positions of occupants present in the vehicle cabin as a point cloud on a three-dimensional map corresponding to a space within the vehicle cabin based on a transmission wave transmitted toward the vehicle cabin and reflected by an occupant in the vehicle cabin; a calculation unit that counts each of the point clouds that appear in a plurality of regions of interest that divide each of a plurality of seating areas adjacent to each other in the vehicle cabin by a predetermined width in a direction in which the plurality of seating areas extend within the three-dimensional map, and identifies peak positions and peak values of the point clouds within the plurality of regions of interest; and a determination unit that determines whether or not an occupant is present based on the peak positions and the peak values, and the determination unit is configured to: If multiple occupants are detected at the same time, at a second time that is one time after the first time, a first threshold is calculated by multiplying a maximum peak value of one or more peaks identified for the point cloud at the second time by a first ratio, and a peak having a peak value equal to or greater than the first threshold is counted as the number of occupants among the one or more peaks; if one occupant is detected at the first time, a second threshold is calculated at the second time by multiplying a maximum peak value of one or more peaks identified for the point cloud at the second time by a second ratio that is higher than the first ratio, and a peak having a peak value equal to or greater than the second threshold is counted as the number of occupants.

[0007] The occupant detection device of the embodiment can identify the number of occupants with high accuracy. [Brief explanation of the drawings]

[0008] [Figure 1] FIG. 1 is a perspective side view showing an example of the configuration of the interior of a vehicle in which an occupant detection system according to an embodiment is installed. [Figure 2] FIG. 2 is a perspective top view showing an example of the configuration of the interior of a vehicle in which an occupant detection system according to an embodiment is installed. [Figure 3] FIG. 3 is a block diagram illustrating an example of a hardware configuration of the occupant detection system according to the embodiment. [Figure 4] FIG. 4 is a block diagram illustrating an example of a functional configuration of the occupant detection system according to the embodiment. [Figure 5] FIG. 5 is a schematic diagram illustrating an example of an analysis method performed by the occupant detection system according to the embodiment. [Figure 6] FIG. 6 is a schematic diagram illustrating an example of an analysis result by the occupant detection system according to the embodiment. [Figure 7] FIG. 7 is a schematic diagram illustrating an example of an analysis result obtained by the occupant detection system according to the embodiment. [Figure 8] FIG. 8 is a schematic diagram illustrating an example of an analysis result by the occupant detection system according to the embodiment. [Figure 9] FIG. 9 is a schematic diagram showing an example of different seating area settings applied to the occupant detection system according to the embodiment. [Figure 10] FIG. 10 is a schematic diagram showing an example of different seating area settings applied to the occupant detection system according to the embodiment. [Figure 11] FIG. 11 is a flowchart illustrating an example of a procedure of an occupant detection process performed by the occupant detection device according to the embodiment. DETAILED DESCRIPTION OF THE INVENTION

[0009] Exemplary embodiments of the present invention are disclosed below. The configurations of the embodiments described below, as well as the actions, results, and advantages brought about by the configurations, are merely examples. The present invention can be realized using configurations other than those disclosed in the following embodiments, and it is possible to obtain at least one of the various advantages and derivative advantages based on the basic configurations.

[0010] (Configuration example of an occupant detection system) Fig. 1 is a perspective side view showing an example of the configuration inside a passenger compartment R of a vehicle C equipped with an occupant detection system 1 according to an embodiment. Fig. 2 is a perspective top view showing an example of the configuration inside a rear part of the passenger compartment R of a vehicle C equipped with an occupant detection system 1 according to an embodiment.

[0011] In the figure, the +X direction is the direction from the center of the left-right direction of vehicle C toward the right, and the -X direction is the direction from the center of the left-right direction of vehicle C toward the left. The left-right direction of vehicle C is the direction as seen from the driver's seat. The +Y direction is the direction from the center of the front-rear direction of vehicle C toward the front, and the -Y direction is the direction from the center of the front-rear direction of vehicle C toward the rear. The +Z direction is the direction downward from the ceiling 5 of vehicle C.

[0012] The vehicle C may be, for example, an automobile powered by an internal combustion engine or a motor, or a hybrid vehicle powered by both of these.

[0013] As shown in Figures 1 and 2, vehicle C of this embodiment is a so-called three-row seat vehicle, and includes, from the front of vehicle C, seat rows SR1, SR2, and SR3. Seat row SR1 includes a driver's seat and a passenger seat. Seat rows SR2 and SR3 each include a plurality of seats adjacent to each other.

[0014] In the example of Figure 2, seat row SR2 includes three seats adjacent to each other, with occupants P seated on either side of the center seat. Seat row SR3 includes two seats adjacent to each other, with occupants P seated on the seat on the right side of vehicle C.

[0015] The occupant detection system 1 has a function of detecting an occupant P present in the vehicle compartment R and the seating position of the occupant P. The detection result of the occupant P by the occupant detection system 1 is used, for example, for a seat belt reminder. The seat belt reminder is a function of notifying the vehicle compartment R of a seat position where the seat belt is not fastened even though the occupant P has been detected. In the embodiment, a case will be described in which an occupant P present mainly in seat row SR2 out of seat rows SR1 to SR3 is detected.

[0016] The occupant detection system 1 includes a sensor device 2 and an occupant detection device 3.

[0017] The sensor device 2 is installed, for example, on the ceiling 5 near the center of the vehicle C in the ±X directions, transmits a transmission wave toward the interior of the vehicle compartment R, and receives a reflected wave generated when the transmission wave is reflected by an object present in the vehicle compartment R. The sensor device 2 of the embodiment is installed behind the seat backs of the seat row SR1 when viewed from the ceiling 5. The occupant detection device 3 is installed, for example, in the dashboard, and is connected to the sensor device 2 via a network such as a CAN (Controller Area Network).

[0018] The installation positions of the sensor device 2 and the occupant detection device 3 are not limited to those described above. Also, the number of sensor devices 2 installed in the vehicle interior R is not limited to the example shown in FIG.

[0019] FIG. 3 is a block diagram showing an example of a hardware configuration of the occupant detection system 1 according to the embodiment.

[0020] As shown in FIG. 3, the sensor device 2 included in the occupant detection system 1 includes a transmitter 21, a receiver 22, a sensor ECU (Electronic Control Unit) 23, and an input / output unit 24.

[0021] The transmitter 21 transmits (irradiates) an electromagnetic wave of a predetermined frequency, such as 60 GHz to 65 GHz, as a transmission wave into the vehicle interior R. The receiver 22 receives a reflected wave generated when the transmission wave is reflected by an object present in the vehicle interior R, and generates an electrical signal indicating the intensity of the reflected wave. The transmitter 21 and the receiver 22 may be configured using, for example, an oscillator circuit, a piezoelectric element, an AD converter, an amplifier, a filter circuit, etc. The transmitter 21 and the receiver 22 may be configured separately from each other, or may be configured integrally.

[0022] The sensor ECU 23 is a microcontroller configured using a CPU (Central Processing unit), memory, etc., and performs processes related to control of the transmitter 21 and receiver 22, generation of data based on the reflected waves received by the receiver 22, etc.

[0023] The input / output unit 24 is an interface device that establishes communication between the occupant detection device 3 and other devices in accordance with a predetermined standard such as CAN.

[0024] As described above, the sensor device 2 has a configuration in which a radio wave sensor including, for example, the transmitter 21 and the receiver 22 is integrated with the sensor ECU 23 that generates data based on transmitted and received waves obtained from the radio wave sensor.

[0025] The occupant detection device 3 included in the occupant detection system 1 is an ECU or the like including a CPU 31, a memory 32, and an input / output unit 33.

[0026] The CPU 31 executes various arithmetic processes in accordance with programs stored in, for example, the memory 32. The memory 32 may be configured using a volatile memory and a non-volatile memory. The memory 32 stores programs that cause the CPU 31 to execute various processes for realizing the functions of the occupant detection device 3, setting data, data acquired from the sensor device 2, data generated by the CPU 31, and the like.

[0027] The input / output unit 33 is an interface device that establishes communication between the sensor device 2 and other devices in accordance with a predetermined standard such as CAN.

[0028] It should be noted that the hardware configuration shown in FIG. 3 is an example, and the hardware configuration of the occupant detection system 1 is not limited to the above.

[0029] FIG. 4 is a block diagram showing an example of a functional configuration of the occupant detection system 1 according to the embodiment.

[0030] As shown in FIG. 4, the sensor device 2 included in the occupant detection system 1 includes a wave transmitting / receiving unit 201 and a generating unit 202 as functional units.

[0031] The wave transmitting / receiving unit 201 transmits a transmission wave such as an electromagnetic wave into the vehicle interior R, receives a reflected wave of the transmission wave reflected by an object present in the vehicle interior R, and generates an electrical signal indicating the intensity of the reflected wave.

[0032] The generation unit 202 generates point cloud information based on the electrical signal generated by the wave transmitting and receiving unit 201. The point cloud information is a point cloud showing one or more detection points representing the position of the occupant P present in the vehicle cabin R on a three-dimensional map corresponding to the space within the vehicle cabin R. The three-dimensional map corresponding to the space within the vehicle cabin R may be, for example, a voxel map. In this case, the detection points representing the occupant P may be voxels within the voxel map.

[0033] Here, the electrical signal generated by the wave transmitting / receiving unit 201 includes not only the occupant P in the vehicle compartment R, but also waves reflected by components of the vehicle C such as the vehicle body and seats, and luggage placed on the seats, etc. Among these, the components and luggage, etc. of the vehicle C maintain a substantially stationary state, or fluctuate in a manner highly correlated with the movement of the vehicle body as the vehicle body sways. Therefore, the components and luggage, etc. of the vehicle C also fluctuate in a manner highly correlated with the sensor device 2 installed, for example, on the ceiling 5 of the vehicle C. On the other hand, the occupant P in the vehicle compartment R is usually seated and making some kind of spontaneous movement.

[0034] Therefore, the point cloud information can be obtained by extracting, from the electrical signals generated by the wave transmitting and receiving unit 201, reflecting objects whose movements are not synchronized with the movement of the vehicle body as detection points representing the occupant P. As an example, the detection points representing the occupant P can be points where the amount of change in the intensity of the reflected waves per unit time caused by movements not synchronized with the movement of the vehicle body is greater than a predetermined threshold. However, the type of points to be set as detection points representing the occupant P can be determined appropriately depending on the function of the sensor device 2 to be used, etc.

[0035] The occupant detection device 3 included in the occupant detection system 1 includes, as functional units, an acquisition unit 301, a calculation unit 302, a determination unit 303, and a storage unit 304. These functional units may be configured, for example, by cooperation between hardware and a program as illustrated in Fig. 3. Furthermore, some or all of these functional units may be configured by dedicated hardware (circuits, etc.) such as an ASIC (Application Specific Integrated Circuit) or an FPGA (Field-Programmable Gate Array).

[0036] The acquisition unit 301 acquires point cloud information generated by the sensor device 2.

[0037] The calculation unit 302 counts each point cloud included in the point cloud information and identifies the peak positions and peak values of these point clouds. In the 3D map in which the point clouds are arranged and correspond to the space within the vehicle cabin R, seating areas where the occupant P should be seated are set corresponding to the multiple seat positions within the vehicle cabin R. Furthermore, these seating areas are further divided into multiple areas each having a predetermined width in the ±X direction. The calculation unit 302 identifies the peak positions and peak values of the point cloud within these areas.

[0038] The determination unit 303 determines whether or not an occupant P is present in each of the multiple regions dividing the multiple seating regions. That is, the determination unit 303 extracts a region having a peak from the multiple regions, and if the peak value of the peak in that region is equal to or greater than a predetermined threshold, determines that the occupant P is present in that region. The predetermined threshold is appropriately determined based on the maximum peak value of one or more peaks in the point cloud identified by the calculation unit 302.

[0039] The storage unit 304 stores, for example, programs and control parameters that realize various functions of the occupant detection device 3. The storage unit 304 also stores point cloud information acquired by the occupant detection device 3 from the sensor device 2, peak positions and peak values of the point cloud in multiple regions that divide the seating region, and determination results such as the presence or absence of an occupant P. The storage unit 304 also stores information such as a three-dimensional map corresponding to the space within the vehicle interior R, the size and arrangement of each seating region in the three-dimensional map, and the size and arrangement of regions set in each seating region.

[0040] (Determination method for occupant detection system) Next, a determination method by the occupant detection system 1 according to the embodiment will be described with reference to Fig. 5 to Fig. 10. Fig. 5 is a schematic diagram showing an example of an analysis method by the occupant detection system 1 according to the embodiment.

[0041] 5 shows a state in which multiple areas A1 to A13 are assigned to the seat row SR2 based on the point cloud information generated by the sensor device 2. The horizontal axis of the graph indicates the position of the vehicle C in the left-right direction (±X direction), and the vertical axis of the graph indicates the position of the vehicle C in the height direction (+Z direction).

[0042] As shown in Figure 5, the sensor device 2 generates point cloud information on a three-dimensional map corresponding to the space inside the vehicle compartment R, in which a point cloud PP, which is one or more detection points determined to be reflections from the occupant P, is arranged asynchronously with the movement of the vehicle body.

[0043] On a three-dimensional map corresponding to the space within the vehicle cabin R, seating areas SR2(L), SR2(C), and SR2(R) are set at positions corresponding to the three seats in seat row SR2. These seating areas SR2(L), SR2(C), and SR2(R) are also collectively referred to as seating area SR2(ST).

[0044] Of the three seats in seat row SR2, seating area SR2(L) corresponds to the seats on the left side of vehicle C, seating area SR2(R) corresponds to the seats on the right side of vehicle C, and seating area SR2(C) corresponds to the center seat. In this way, the graph in Fig. 5 shows the seat positions as viewed from sensor device 2, which is installed slightly rearward and above seat row SR2. In other words, on the graph in Fig. 5, seating areas SR2(L), SR2(C), and SR2(R) are shown in this order from the left side of the page.

[0045] When viewed from above the vehicle C, that is, from the ceiling 5 where the sensor device 2 is arranged, these seating areas SR2(L), SR2(C), and SR2(R) are set to include at least approximately the entire seating surface of each corresponding seat and a portion of the floor surface in front of these seats.

[0046] In a side view of the vehicle C, that is, in the height direction, these seating areas SR2(L), SR2(C), and SR2(R) are set to include a range within a predetermined height position, for example, 1 m downward (in the +Z direction) from the ceiling 5 (0 m). In the example of FIG. 5, a range of 0.8 m in the +Z direction from the ceiling 5 (0 m) is set as the seating areas SR2(L), SR2(C), and SR2(R). In this way, by setting the height positions of the seating areas SR2(L), SR2(C), and SR2(R) higher than the seat surfaces of each seat, it is possible to exclude reflections from the seat surfaces and luggage, etc., and to detect reflections from the occupant P, who is taller than the seat surfaces and luggage, etc.

[0047] These seating areas SR2(L), SR2(C), and SR2(R) are subdivided according to positions in the ±X directions, and a plurality of areas A1 to A13, each having a predetermined width, such as 10 cm, in the ±X directions, are set. In the example of Fig. 5, of the plurality of areas A1 to A13, areas A1 to A5 are assigned to the seating area SR2(L), areas A6 to A8 are assigned to the seating area SR2(C), and areas A9 to A13 are assigned to the seating area SR2(R).

[0048] Such regions A1 to A13 are also called regions of interest (ROI).

[0049] The calculation unit 302 counts the points PP for each of the regions A1 to A13 and identifies the peak positions and peak values of the points PP within the regions A1 to A13. The results are shown in FIGS.

[0050] 6 to 8 are schematic diagrams showing examples of analysis results by the occupant detection system 1 according to the embodiment. The graphs shown in Fig. 6 to 8 show peak positions and peak values of the point cloud PP identified by the occupant detection system 1 within multiple regions A1 to A13.

[0051] The horizontal axes of the graphs shown in Figures 6 to 8 indicate coordinates corresponding to the regions A1 to A13. That is, for example, "2" on the horizontal axis of the graph corresponds to region A2, and "8" on the horizontal axis of the graph corresponds to region A8. The vertical axes of the graphs shown in Figures 6 to 8 indicate the density of the point clouds PP in each of the regions A1 to A13, that is, the number of point clouds PP. The point clouds PP will have a peak in a region among the multiple regions A1 to A13 where the density of the point clouds PP is higher than in other adjacent regions.

[0052] 6 shows the analysis result of the occupant detection system 1 at time t0. Time t0 is the timing of the first analysis by the occupant detection system 1. After starting detection of the occupant P at time t0, the occupant detection system 1 of the embodiment repeatedly generates and analyzes point cloud information at a predetermined cycle, such as every 200 ms.

[0053] At the first analysis timing, the occupant detection system 1 determines whether or not an occupant P is present in the vehicle compartment R using a threshold value TH1 obtained by multiplying the maximum peak value of one or more peaks identified at that timing by a predetermined ratio R1. In the example of FIG. 6, one peak with a peak value of 24 is detected in area A10. Also, in the example of FIG. 6, the threshold value TH1 is 7.3, which is calculated by multiplying the peak value of the peak in area A10 by a predetermined ratio of 30%.

[0054] Here, if the point cloud PP indicates an occupant P, the point cloud PP tends to appear at a high density in a specific area of the areas A1 to A13. On the other hand, if Tengu PP appear discretely in several areas of the areas A1 to A13, those point clouds PP are likely to be components and luggage of the vehicle C, or simple noise, etc.

[0055] Therefore, the calculation unit 302 described above identifies the portion of the regions A1 to A13 where the count of the point cloud PP is maximum among the mutually adjacent regions as the peak of the point cloud PP. A region where the point cloud PP appears as a peak means that the point cloud PP exists at high density, and as described above, there is a high possibility that the point cloud PP represents the occupant P. In the example of Fig. 6, the point cloud PP appears in the mutually adjacent regions A8 to A11, and among these regions A8 to A11, the region A10 where the count of the point cloud PP is maximum is identified as the region where the peak of the point cloud PP exists.

[0056] The determination unit 303 determines whether an occupant P is present in the area A10 based on whether the peak of the area A10 is equal to or greater than the threshold value TH1. In the example of Fig. 6, the peak value of the peak of the area A10 is 24, which is equal to or greater than the threshold value TH1 of 7.3. Therefore, the determination unit 303 determines that one occupant P is present in the seating area SR2(L) corresponding to the area A10.

[0057] 7 shows the analysis result of the occupant detection system 1 at time t1. Time t1 is a timing after the above-mentioned time t0, which is the timing of the first analysis by the occupant detection system 1.

[0058] At times after time t0, the occupant detection system 1 determines whether or not an occupant P is present in the vehicle compartment R using a threshold value TH2 obtained by multiplying the maximum peak value of one or more peaks identified at that time by a predetermined ratio R2, or a threshold value TH3 obtained by multiplying the maximum peak value of one or more peaks identified at that time by a predetermined ratio R3. Which of the threshold values TH2 and TH3 to use at time t1 is determined based on the analysis result at the analysis time immediately before time t1.

[0059] More specifically, if multiple occupants P are detected at the analysis timing immediately before time t1, the determination unit 303 determines the presence or absence of an occupant P using a threshold value TH2 obtained by multiplying the maximum peak value by the ratio R2. Also, if a single occupant P is detected at the analysis timing immediately before time t1, the determination unit 303 determines the presence or absence of an occupant P using a threshold value TH3 obtained by multiplying the maximum peak value by a ratio R3 that is larger than the ratio R2. In other words, if only one occupant P is detected at a predetermined time, the occupant detection system 1 makes the detection conditions for the occupant P stricter at the next analysis timing.

[0060] In the example of Fig. 7, peaks of the point cloud PP are recognized in regions R4, R7, and R10, and among these, the peak with the highest peak value of 55 is found in region A7. In the example of Fig. 7, the ratio R2 is set to 10% and multiplied by the maximum peak value to obtain 5.5, which is the threshold value TH2. Furthermore, the ratio R3 is set to 30% and multiplied by the maximum peak value to obtain 16.5, which is the threshold value TH3.

[0061] Therefore, if multiple occupants P are detected at the analysis timing immediately before time t1, the determination unit 303 determines that peaks having peak values equal to or greater than the threshold value TH2 indicate the presence of an occupant P. In the example of Fig. 7, since the peak values of all peaks in regions R4, R7, and R10 are equal to or greater than the threshold value TH2, the determination unit 303 determines that one occupant P is present in each of the seating regions SR(L), SR(C), and SR(R) corresponding to regions R4, R7, and R10, respectively, and that three occupants P are present in the entire seat row SR2.

[0062] Furthermore, if only one occupant P is detected at the analysis timing immediately before time t1, the determination unit 303 determines that a peak having a peak value equal to or greater than the threshold value TH3 indicates the presence of an occupant P. In the example of Fig. 7, among the regions R4, R7, and R10, only the peak value of the peak in region R7 is equal to or greater than the threshold value TH3, so the determination unit 303 determines that one occupant P is present in the seating region SR(C) corresponding to region R7.

[0063] In addition, if the number of detected occupants P is zero at the analysis timing one before time t1, the number of occupants at time t1 can be determined using threshold value TH1 multiplied by ratio R1, as in the first analysis timing shown in Figure 6 above.

[0064] As described above, if only one occupant P is detected, the threshold is set higher at the next analysis timing to make the detection conditions for the occupant P stricter. This makes it possible to suppress erroneous detections such as overestimating the number of occupants P at the next analysis timing.

[0065] On the other hand, at the analysis timing immediately after multiple occupants P are detected, the threshold value is set lower to relax the detection conditions for the occupants P. This increases the detection sensitivity for the occupants P, making it easier to detect all occupants P in the vehicle compartment R.

[0066] Furthermore, as shown in Figure 6 above, by setting a higher threshold value and tightening the detection conditions for occupant P even at the initial analysis timing, it is possible to suppress erroneous detections such as overestimating the number of occupants P at the initial analysis timing, which serves as the basis for subsequent analyses.

[0067] Here, the ratio R2 applied when multiple occupants P were detected at the immediately previous analysis timing is an example of a first ratio, and the threshold value TH2 obtained by multiplying the ratio R2 by the maximum peak value is an example of the first threshold. Also, the ratio R3 applied when a single occupant P was detected at the immediately previous analysis timing is an example of a second ratio, and the threshold value TH3 obtained by multiplying the ratio R3 by the maximum peak value is an example of the second threshold.

[0068] Furthermore, time t1, which is after the above-mentioned time t0, which is the first analysis timing, is an example of the second time, and the analysis timing immediately before time t1 is an example of the first time.

[0069] 6 and 7, the ratio R1 applied at time t0 and the ratio R3 applied at time t1 after time t0 are both set to 30%, but the ratios R1 and R3 may be different values. However, it is preferable that both ratios R1 and R3 be greater than the ratio R2.

[0070] 8 shows the analysis result of the occupant detection system 1 at time tX. Time tX is an arbitrary timing of analysis by the occupant detection system 1. In other words, time tX may be time t0, which is the timing of the first analysis, or time t1, which is a timing after time t0, and the analysis method of the occupant detection system 1 described below can be applied to the entire time period during which detection of the occupant P continues.

[0071] At a time tX that is an arbitrary timing, the occupant detection system 1 distinguishes between peaks that are counted as occupants P and peaks that are not counted, depending on the peak positions of the point cloud PP.

[0072] In the example of FIG. 8, peaks of the point cloud PP appear in regions A5, A7, and A11. Of these, the peak in region A7 shows the largest peak value of 60, and this is multiplied by, for example, ratio R3 to obtain 18, which is used to determine the presence or absence of an occupant P. Of regions A5, A7, and A11, the peak value of the peak in region A11 is less than threshold TH3, so it is not counted. On the other hand, of regions A5, A7, and A11, the peak values of the peaks in regions A5 and A7 are both equal to or greater than threshold TH3.

[0073] However, here, the occupant detection system 1 is configured to count only the peak having the largest peak value among one or more peaks appearing in a region within a predetermined range adjacent to each other. The region within the predetermined range may be, for example, three consecutive regions.

[0074] 8, areas A5 and A7 are adjacent to each other with area A6 in between, and are included in the range of three consecutive areas. Therefore, the determination unit 303 does not count each of the peaks in areas A5 and A7 as occupant P, but only counts the peak in area A7, which has the higher peak value, among these peaks.

[0075] If the multiple peaks identified by the calculation unit 302 are located in adjacent regions within the predetermined range, there is a high possibility that a single occupant P who is seated in an unsteady posture, for example, has been detected as multiple peaks. In this case, it is highly likely that the peak position of the peak with the largest peak value among the multiple peaks detected in the region within the predetermined range is the actual seating position of the occupant P. Therefore, by counting only the peak in region A7, which has the higher peak value, as the occupant P, of the peaks in regions A5 and A7, it is possible to prevent multiple peaks corresponding to a single occupant P from being detected as multiple occupants P.

[0076] As described above, in FIG. 8, the determination unit 303 counts only the peak in area A7 among areas A5, A7, and A11 as an occupant P, and determines that one occupant P is present only in the seating area SR(C) corresponding to area A7.

[0077] 6 to 8, 13 areas A1 to A13 each having a predetermined width, such as 10 cm, are set as the seating area SR2 (ST). However, the width and number of the areas A1 to A13 are not limited to the above example and can be determined arbitrarily.

[0078] The above analysis method can also be applied to the seat row SR3 behind the seat row SR2.

[0079] That is, a seating area corresponding to seat row SR3 is set in a 3D map corresponding to the space in the vehicle interior R, and the seating area is further divided into multiple regions. Then, the points PP in the multiple regions obtained by dividing the seating area are counted, the peak positions and peak values of the points PP are identified, and a predetermined threshold value corresponding to the maximum peak value is used to determine whether or not to count one or more identified peaks as an occupant P. Furthermore, the conditions shown in FIG. 8 can also be applied to the analysis of seat row SR3.

[0080] This allows the occupant P in the seat row SR3 to be detected in the same manner as in the seat row SR2.

[0081] Incidentally, when detecting an occupant P seated in seat row SR2, other conditions are set for the occupant detection system 1. The other conditions imposed on the occupant detection system 1 are shown in FIGS. 9 and 10.

[0082] 9 and 10 are schematic diagrams showing examples of different settings of the seating areas SR2(ST) and SR2(NR) applied to the occupant detection system 1 according to the embodiment.

[0083] As shown in Figures 9 and 10, in the occupant detection system 1, the setting ranges of the seating areas SR2(ST) and SR2(NR) corresponding to seat row SR2 differ depending on whether or not an occupant P is present in seat row SR3 behind seat row SR2.

[0084] FIG. 9 shows an example of setting the seating area SR2(ST) when no occupant P is present in the seat row SR3.

[0085] 9, seating areas SR2(ST) and ST3(ST) corresponding to the seat rows SR2 and SR3, respectively, are set on the three-dimensional map of the vehicle interior R. When viewed from the ceiling 5 where the sensor device 2 is arranged, these seating areas SR2(ST) and ST3(ST) are set to include substantially the entire backrests and seat surfaces of the corresponding seats and a portion of the floor surface in front of these seats.

[0086] In this way, the seating area SR2 (ST) shown in Figure 9 is set to a wide area that includes almost the entire backrest and seat surface of each seat in the seat row SR2, and is applied when there is no occupant P in the rear seating area SR3 (ST).

[0087] As shown in Fig. 10, seating areas SR2(NR) and ST3(ST) corresponding to seat rows SR2 and SR3, respectively, are set in the three-dimensional map of the vehicle interior R. Of these, the setting range of seating area SR3(ST) is the same as the example in Fig. 9 described above.

[0088] On the other hand, the seating area SR2 (NR) is set to include, when viewed from the ceiling 5 where the sensor device 2 is placed, almost the entire seat surface excluding the backrest of each corresponding seat, and a portion of the floor surface in front of these seats.

[0089] In this way, the seating area SR2 (NR) shown in Figure 10 is set to a narrow area with its rear end reduced away from the seating area SR3 (ST), and is applicable when there are one or more occupants P in the rear seating area SR3 (ST).

[0090] The seating area SR2 (NR) is also set up to include seating areas SR2 (L), SR2 (C), and SR2 (R) which are reduced in size so that their rear ends are away from the seating area SR3 (ST), and a plurality of areas A1 to A13 which further divide these reduced seating areas SR2 (L), SR2 (C), and SR2 (R).

[0091] As described above, when there is no occupant P in the seating area SR3(ST), by applying the wide seating area SR2(ST), it is possible to detect a point cloud PP indicating the presence of the occupant P over a wider area, thereby improving the detection accuracy of the occupant P. On the other hand, when there is an occupant P in the seating area SR3(ST), by applying the narrow seating area SR2(NR), it is possible to prevent the occupant P in the rear seat row SR3 from being mistakenly detected as an occupant P seated in the seat row SR2.

[0092] (Processing example of occupant detection device) Next, an example of the occupant detection process of the occupant detection device 3 according to the embodiment will be described with reference to Fig. 11. Fig. 11 is a flow chart showing an example of the procedure of the occupant detection process by the occupant detection device 3 according to the embodiment. Fig. 11 shows a determination process per one time for the seating area SR2(ST) or the seating area SR2(NR).

[0093] As shown in FIG. 11, the acquisition unit 301 acquires point cloud information that shows a point cloud of the occupant P on a three-dimensional map of the vehicle interior R, which is detected by the transmitting and receiving unit 201 of the sensor device 2 and generated by the generation unit 202 (step S101).

[0094] The calculation unit 302 checks whether an occupant P has been detected in the seat row SR3 behind the seat row SR2 at the most recent timing (step S102). If an occupant P has not been detected in the seat row SR3 (step S102: No), the calculation unit 302 selects to use the seating area SR2 (ST) including the wide areas A1 to A13 to detect the occupant P in the seat row SR2 (step S103). If an occupant P has been detected in the seat row SR3 (step S102: Yes), the calculation unit 302 selects to use the seating area SR2 (NR) including the narrow areas A1 to A13 to detect the occupant P in the seat row SR2 (step S104).

[0095] The calculation unit 302 also counts the points PP in each of the regions A1 to A13 (step S105), and identifies the peak position and peak value of the points PP from the density of the counted points PP (step S106).

[0096] The determination unit 303 determines which of the threshold values TH1 to TH3 to apply when determining whether or not an occupant P is present.

[0097] That is, the judgment unit 303 checks whether this is the first analysis timing (step S107), and if it is the first analysis timing (step S107: Yes), selects to apply a threshold value TH1 having a ratio of, for example, 30% of the maximum peak value of the point cloud PP (step S110).

[0098] If this is not the first analysis timing (step S107: No), the determination unit 303 checks whether multiple occupants P were detected last time (step S108), and if multiple occupants P were detected (step S108: Yes), it selects to apply threshold value TH2 having a ratio of 10% of the maximum peak value of the point cloud PP (step S109).If there is a single occupant P or zero occupants P (step S108: No), the determination unit 303 selects to apply threshold value TH3 having a ratio of 30% of the maximum peak value of the point cloud PP, or threshold value TH1 (step S110).

[0099] Furthermore, the determination unit 303 determines whether the peak value of a peak present in any of the regions A1 to A13 is equal to or greater than any of the selected thresholds TH1 to TH3 (step S111). If the peak value is equal to or greater than any of the thresholds TH1 to TH3 (step S111: Yes), the determination unit 303 determines whether a peak having a higher peak value than the peak in the region currently being determined exists in an adjacent region within a predetermined range (step S112). If no higher peak exists in the adjacent region (step S112: No), the peak in the region currently being determined is counted as the occupant P (step S113).

[0100] On the other hand, if the peak value of the peak in the area currently being judged is less than any of the selected thresholds TH1 to TH3 (step S111: No), or if a larger peak exists in the adjacent area (step S112: Yes), the judgment unit 303 does not count the peak in the area currently being judged as an occupant P (step S114).

[0101] Thereafter, the occupant detection device 3 determines whether or not the determination for all the areas A1 to A13 has been completed (step S115). If the determination for all the areas A1 to A13 has been completed (step S115: Yes), the occupant detection device 3 ends the process. If the determination for all the areas A1 to A13 has not been completed (step S115: No), the occupant detection device 3 repeats the processes from step S111 onwards.

[0102] With the above, one cycle of the occupant detection process of the occupant detection device 3 of the embodiment is completed.

[0103] The occupant detection device 3 starts the process shown in Fig. 11 at a predetermined timing and repeats the process until the predetermined timing is reached. For example, when the determination result of the occupant detection device 3 is used for the seat reminder function, the above process is repeated from when the engine of the vehicle C is turned on until it is turned off.

[0104] (Overview) There is a technology for detecting occupants in a vehicle cabin using a radio wave sensor or the like. However, if an occupant is sitting in a position that is out of alignment with the seat, multiple peaks of the point cloud may be detected across multiple seating areas, and each of these may be detected as an occupant. In this way, for example, if there is only one occupant, there is a tendency for the occupant to be mistakenly detected as multiple occupants.

[0105] According to the embodiment of the occupant detection device 3, when one occupant P is detected at a predetermined time, the judgment unit 303 calculates a threshold value TH3 at a time after the predetermined time by multiplying the maximum peak value of one or more peaks identified for the point cloud PP at the later time by a ratio R3 higher than the ratio R2, and counts the peaks among the one or more peaks having a peak value greater than or equal to the threshold value TH3 as the number of occupants P.

[0106] As described above, the point cloud PP representing the occupant P appears at high density in a predetermined area, while the point cloud PP representing objects other than the occupant P appears at low density scattered across multiple areas. As described above, by using one of the thresholds TH2 and TH3 based on the maximum peak value of one or more peaks, the peak of the point cloud PP representing the occupant P can be extracted with high accuracy, and the occupant P can be detected.

[0107] In addition, by referring to the number of occupants P detected at the previous analysis timing, and applying a high threshold value TH3 to tighten the detection conditions after one occupant P is detected, it is possible to prevent one occupant P from being mistakenly detected as multiple occupants P.

[0108] In this way, the number of occupants P can be identified with high accuracy.

[0109] According to the occupant detection device 3 of the embodiment, even if multiple peaks are identified in the point cloud PP, if two or more peaks are identified in areas among the multiple areas A1 to A13 that are within a predetermined range of each other, only the peak with the highest peak value among the two or more identified peaks is counted as the number of occupants P.

[0110] This prevents, for example, a single occupant P who is seated in a slumped posture from being mistakenly detected as multiple occupants P. Therefore, the number of occupants P can be identified with even higher accuracy.

[0111] According to the embodiment of the occupant detection device 3, when an occupant P is detected in another seating area SR3(ST) behind the seating areas SR2(L), SR2(C), and SR2(R), the calculation unit 302 counts the point clouds PP that appear in the multiple areas A1 to A13 that have been reduced so that the rear ends of the multiple areas A1 to A13 set in the seating areas SR2(L), SR2(C), and SR2(R) are moved away from the other seating area SR3(ST), and identifies the peak position and peak value of the point cloud PP.

[0112] This prevents the occupant P seated in the seat row SR3 behind the seat row SR2 from being mistakenly detected as the occupant P seated in the seat row SR2. Therefore, the number of occupants P can be identified with even higher accuracy.

[0113] In the above embodiment, when multiple peaks appear in adjacent regions within a predetermined range, only the peak with the largest peak value is counted as the occupant P. However, the method for preventing a single occupant P who is seated with poor posture from being mistakenly detected as multiple occupants P is not limited to this.

[0114] As shown in Fig. 8 above, among the seating areas SR(L), SR(C), and SR(R) corresponding to a seat row SR2 having three seats, the occupant P seated in the central seating area SR(C) is closest to the sensor device 2 installed on the ceiling 5 in the center of the vehicle compartment R, for example, and is therefore likely to have the strongest peak. Therefore, instead of defining the area within the above-mentioned predetermined range as any three consecutive areas, if a peak having the maximum peak value appears in any of the areas A6 to A8 corresponding to the central seating area SR(C), the area from that area to the area separated by one area may be defined as the predetermined range.

[0115] That is, if, among one or more peaks appearing in regions A1 to A13, the peak with the largest peak value appears in region A7, other peaks appearing in regions A5 to A9 are not counted as occupant P. Also, if, among one or more peaks appearing in regions A1 to A13, the peak with the largest peak value appears in region A6, other peaks appearing in regions A4 to A8 are not counted as occupant P. Also, if, among one or more peaks appearing in regions A1 to A13, the peak with the largest peak value appears in region A8, other peaks appearing in regions A6 to A9 are not counted as occupant P.

[0116] In this way, as another method for suppressing erroneous detection of the number of occupants, even if multiple peaks are identified in the point cloud PP, if other peaks are identified in an area within a predetermined range from the area where the peak with the largest peak value appears, the other peaks can be not counted as the number of occupants PP. This method also prevents the number of occupants P from being overestimated, and the number of occupants P can be identified with high accuracy.

[0117] Although the embodiments of the present invention have been described above, the above embodiments are presented as examples and are not intended to limit the scope of the invention. This novel embodiment can be embodied in various other forms, and various omissions, substitutions, and modifications can be made without departing from the spirit of the invention. This embodiment and its modifications are included within the scope and spirit of the invention, and are also included in the inventions and their equivalents as set forth in the claims. [Explanation of symbols]

[0118] 1...occupant detection system, 2...sensor device, 3...occupant detection device, 5...ceiling, 21...transmitter, 22...receiver, 23...sensor ECU, 24...input / output unit, 31...CPU, 32...memory, 33...input / output unit, 201...transmitter / receiver unit, 202...generation unit, 301...acquisition unit, 302...calculation unit, 303...determination unit, 304...storage unit, C...vehicle, P...occupant, PP...point cloud, R...vehicle compartment, SR2(L), SR2(C), SR2(R), SR2(NR), SR2(ST), SR3(ST)...seating area, A1 to A13...area, SR1, SR2, SR3...seat row.

Claims

1. an acquisition unit that acquires point cloud information that indicates, as a point cloud, one or more detection points that represent positions of occupants present in the vehicle interior on a three-dimensional map corresponding to a space within the vehicle interior, based on a transmission wave that is transmitted toward the vehicle interior and reflected by the occupants in the vehicle interior; a calculation unit that counts point clouds that appear in a plurality of regions of interest that divide each of a plurality of adjacent seating areas in the vehicle cabin by a predetermined width in a direction in which the plurality of seating areas extend on the three-dimensional map, and identifies peak positions and peak values of the point clouds in the plurality of regions of interest; a determination unit that determines whether or not an occupant is present based on the peak position and the peak value, The determination unit If a plurality of occupants are detected at a first time, at a second time that is one time after the first time, a first threshold is calculated by multiplying a maximum peak value of one or more peaks identified in the point cloud at the second time by a first ratio, and counting a peak having a peak value equal to or greater than the first threshold as the number of occupants among the one or more peaks; If one occupant is detected at the first time, at the second time, a second threshold is calculated by multiplying a maximum peak value of one or more peaks identified in the point cloud at the second time by a second ratio that is higher than the first ratio, and a peak having a peak value equal to or greater than the second threshold among the one or more peaks identified at the second time is counted as the number of occupants. Occupant detection device.

2. The determination unit Even if a plurality of peaks are identified for the point cloud, if two or more peaks are identified in regions of interest that are within a predetermined range of each other among the plurality of regions of interest, only the peak with the highest peak value among the two or more identified peaks is counted as the number of occupants. The occupant detection device according to claim 1 .

3. The determination unit Even if a plurality of peaks are identified in the point cloud, if a second peak is identified in a region of interest within a predetermined range from a region of interest in which a first peak having the largest peak value appears, the second peak is not counted as the number of occupants. The occupant detection device according to claim 1 .

4. Another seating area is set behind the plurality of seating areas, The calculation unit If an occupant is detected in the other seating area, counting the point clouds that appear in the plurality of regions of interest that have been reduced so that the rear ends of the plurality of regions of interest are moved away from the other seating area, and identifying the peak positions and peak values of the point clouds. The occupant detection device according to claim 1 .

Images