JP7754650B2 - Vehicle abandonment detection device - Google Patents

Description

The present invention relates to a vehicle abandonment detection device, and more particularly to a vehicle abandonment detection device that detects whether a living organism has been left behind in the vehicle cabin.

In the past, there have been cases where drivers forget that a child is in the back seat and get out of the vehicle, resulting in so-called "abandonment." Particularly in the summer, when this happens, the interior of the vehicle can become very hot, putting the child at risk.

So, devices have been invented to deal with this kind of abandonment.

In the invention described in Patent Document 1, while the vehicle is stopped, it is determined based on heat source information whether a heat source that meets the target conditions is occurring, the inside of the vehicle is photographed using a camera, and if a living organism is detected in the captured image, notification information is output to the outside of the vehicle. In this way, it is possible to quickly notify those outside the vehicle that a living organism has been left inside the vehicle.

In the invention described in Patent Document 2, a microphone is installed in the buckle of the rear seat seatbelt device. When the microphone picks up biological sounds other than conversation, such as crying, it determines that there is an occupant in need of assistance, such as an infant, in the rear seat. If there is no occupant in the driver's seat and there is an occupant in need of assistance in the rear seat for a predetermined period of time or more, an alarm is activated. In this way, it is possible to detect occupants in the vehicle cabin based on the sounds picked up by the microphone.

International Publication No. 2020/065811 JP 2019-99086 A

However, the detection method described above leaves room for improvement in terms of effectively detecting abandoned items.

Specifically, the invention described in Patent Document 1 had the problem that when it detected that a child had been left behind based on heat source information, it was unable to detect the child's abandonment during the summer when the temperature inside the vehicle is higher than the child's body temperature.

Furthermore, in the invention described in Patent Document 2, a microphone is installed in the buckle of the rear seat seatbelt device, but continuous monitoring using the microphone can result in excessive power consumption.

The present invention was made in consideration of these problems, and its object is to provide a vehicle abandonment detection device that can effectively detect abandoned items in the vehicle cabin while minimizing energy consumption.

The present invention is a vehicle abandonment detection device that detects whether a living thing has been left behind in the passenger compartment of a vehicle, and is equipped with a sound collection unit and an arithmetic and control unit. The sound collection unit collects sounds generated in the passenger compartment, and the arithmetic and control unit compares the audio waveform characteristics of judgment sounds classified into multiple categories with the audio waveform characteristics of the sound collected by the sound collection unit, selects from the audio waveform characteristics of the judgment sounds classified into multiple categories that is closest to the audio waveform characteristics of the sound collected by the sound collection unit, determines a time for audio monitoring based on the selected closest category, and determines whether the living thing is present in the passenger compartment based on the sound collected by the sound collection unit within the determined audio monitoring time .

With a vehicle abandonment detection device according to one embodiment of the present invention, when determining whether a living being, such as a child, has been left behind inside the vehicle, the amount of power required for the determination can be suitably controlled by changing the time or amount of data required for the determination in accordance with sound information within the vehicle cabin. This prevents excessive power consumption by the vehicle abandonment detection device while the vehicle is stopped, and prevents excessive battery discharge.

1 is a diagram showing a vehicle equipped with a vehicle abandonment detection device according to an embodiment of the present invention; 1 is a connection diagram showing a vehicle abandonment detection device according to an embodiment of the present invention; 4 is a flowchart showing a method for detecting abandonment using the vehicle abandonment detection device according to the embodiment of the present invention, illustrating a flow for determining whether or not it is necessary to start determination. 1 is a flowchart showing a method for detecting abandonment using the vehicle abandonment detection device according to the embodiment of the present invention, illustrating a flow of determining an in-vehicle voice. 10 is a table showing a method for detecting leaving of an object using the vehicle leaving detection device according to the embodiment of the present invention, showing a table for distinguishing collected sounds. 1 is a diagram showing a vehicle parking detection device according to an embodiment of the present invention, and is a graph showing changes in the sound pressure of collected sound.

The vehicle parking detection device 10 according to an embodiment of the present invention will be described in detail below with reference to the drawings. In the following description, the front, back, top, bottom, left, and right directions are used, but left and right refer to the left and right when viewing the vehicle 11 from the rear. Furthermore, in the following description, the same components will generally be given the same reference numerals, and repeated explanations will be omitted.

Figure 1 shows a vehicle 11 equipped with a vehicle parking detection device 10.

Vehicle 11 is, for example, a passenger car, such as an engine-equipped vehicle, EV (Electric Vehicle), HV (Hybrid Vehicle), PHV (Plug-in Hybrid Vehicle), or FCV (Fuel Cell Vehicle). Vehicle 11 is equipped with a vehicle parking detection device 10, which will be described later.

The vehicle body 18 of the vehicle 11 is equipped with a front-left door 161, a front-right door 162, a rear-left door 163, and a rear-right door 164. The front-left door 161 and the like may open and close using a hinged mechanism or a sliding mechanism. Each of the front-left doors 161 and the like is equipped with an opening/closing sensor 24 (Figure 2) that detects opening and closing operations. Each of the front-left doors 161 and the like is also equipped with a sliding glass window (not shown here).

The vehicle interior 12 of the vehicle body 18 is equipped with a front left seat 191, a front right seat 192, a rear left seat 193, and a rear right seat 194. Here, the driver 20 is seated in the front right seat 192, and an infant 14 is seated in the rear left seat 193. Here, the infant 14 is one form of living creature that may be left behind in the vehicle interior 12. In addition to the infant 14, other living creatures may also be used, such as elderly people, or animals such as dogs and cats.

A sound collection unit 21 and a sound collection unit 22 are provided in the vehicle interior 12. The sound collection unit 21 is located at the front of the vehicle interior 12, and the sound collection unit 22 is located at the rear of the vehicle interior 12. For example, microphones with directionality in the front-to-rear direction can be used as the sound collection unit 21 and the sound collection unit 22. This prevents the sound collection unit 21 and the sound collection unit 22 from picking up noise that enters the vehicle interior 12 via the windows, and allows for good collection of sounds such as the voice of the infant 14.

Figure 2 is a connection diagram showing the vehicle parking detection device 10.

The vehicle abandonment detection device 10 is a device that detects whether a living thing has been left behind in the passenger compartment 12 of a vehicle 11. Specifically, the vehicle abandonment detection device 10 has a sound collection unit 21, a sound collection unit 22, a timer 23, an open/close sensor 24, a calculation control unit 15, a communication unit 25, and an alarm unit 17.

Sound collection units 21 and 22 collect sounds generated within the vehicle cabin 12 and transmit sound information indicating the sound pressure and frequency of the sounds to the calculation control unit 15.

The timer 23 measures the time or duration and transmits information indicating this to the calculation control unit 15.

The opening/closing sensor 24 senses the opening/closing status of the front left door 161, front right door 162, rear left door 163, and rear right door 164 shown in Figure 1, and transmits information indicating these opening/closing statuses to the calculation control unit 15.

The calculation control unit 15 is composed of a CPU, ROM, RAM, etc., and determines whether or not a living thing is present in the vehicle cabin 12 based on sound information input from the sound collection unit 21, etc. Furthermore, when making this determination, the calculation control unit 15 changes the time it takes to make the determination or the amount of data used in the determination, depending on the sound information.

The communication unit 25 is connected to an information communication network such as a telephone communication line. For example, the communication unit 25 is used to notify the smartphone or other device of the user of the vehicle 11 that a child has been left behind in the passenger compartment 12.

The alarm unit 17 is configured to notify the outside that a child 14 has been left behind in the vehicle interior 12. Specifically, the alarm unit 17 may be a horn, hazard lights, headlights, etc.

Figure 3 is a flowchart showing a method for detecting whether an object is left behind using the vehicle-mounted object detection device 10, and shows the flow for determining whether or not a determination needs to be started.

In step S10, the vehicle 11 stops traveling and the engine and ACC (Adaptive Cruise Control) are turned off. If the vehicle 11 uses a motor as a drive source, such as an EV, the motor is turned off.

In step S11, based on step S10, the calculation control unit 15 determines whether the front right door 162 (vehicle front door), which is the door on the driver's 20 side, has been opened or closed using the opening/closing sensor 24.

If the answer to step S11 is YES, i.e., if the front right door 162 has been opened or closed, the calculation control unit 15 proceeds to step S12.

If step S11 is NO, i.e., if the front right door 162 has not been opened or closed, proceed to step S15.

In step S12, the calculation control unit 15 determines whether the rear left door 163 (rear door of the vehicle), which is the door on the child 14's side, has been opened or closed using the opening/closing sensor 24.

If the answer to step S12 is YES, i.e., if the rear left door 163 has been opened or closed, the calculation control unit 15 proceeds to step S13.

If the answer to step S12 is NO, i.e., if the rear left door 163 has not been opened or closed, the calculation control unit 15 proceeds to step S14.

In step S13, the calculation control unit 15 determines that there is a high possibility that occupants, including the infant 14, have exited the vehicle due to the opening and closing of the front right door 162 and the rear left door 163. However, the risk that the infant 14 has been left behind cannot be ruled out. Therefore, the calculation control unit 15 determines that there is a medium risk of the infant 14 being left behind.

In step S14, the calculation control unit 15 determines that there is a high risk of being left behind because only the front right door 162 has been opened and closed, and the rear left door 163 has not been opened or closed, which means that only the driver 20 seated in the front right door 162 has exited the vehicle.

In step S15, the front right door 162 has not been opened or closed, and the driver 20 has not exited the vehicle, so there is no risk of the infant 14 being left behind. Therefore, the calculation control unit 15 ends the flow for detecting whether the infant has been left behind.

In step S16, because there is a medium risk, the calculation control unit 15 sets the monitoring threshold used to determine whether or not an object has been abandoned (described later) to a relatively large value, threshold value S1. By setting threshold value S1 to a relatively large value, it is possible to prevent a situation from being erroneously determined to be abandoned when it is not actually the case.

In step S17, because there is a high risk of abandonment, the calculation control unit 15 sets the monitoring threshold used to determine whether or not an object has been abandoned, described below, to threshold value S2, which is a relatively small value.

Here, the threshold value S2 set in step S17 is a smaller value than the threshold value S1 set in step S16. By setting the threshold value S2 relatively small, it is possible to more reliably detect the occurrence of abandonment.

In step S18, the calculation control unit 15 proceeds to a flow for monitoring audio within the vehicle compartment 12. Details of this monitoring flow will be described later with reference to Figure 4.

Figure 4 shows the in-vehicle voice detection flow illustrating a method for detecting whether an object is left behind using the vehicle-mounted object detection device 10.

In step S20, the calculation control unit 15 determines whether the door glass windows of the vehicle 11 are open. Specifically, based on the output of a sensor (not shown), the calculation control unit 15 determines whether any of the windows of the front left door 161, front right door 162, rear left door 163, and rear right door 164 is open.

If the answer to step S20 is YES, i.e., if the window is open, the calculation control unit 15 proceeds to step S21.

If step S20 is NO, i.e., if the window is not open, the calculation control unit 15 proceeds to step S22. For example, in the subsequent steps, the calculation control unit 15 uses a noise removal filter that assumes that external noise will not enter the vehicle interior 12.

In step S21, the calculation control unit 15 switches the noise removal filter. Specifically, when the windows of the vehicle 11 are open, external noise enters the passenger compartment 12, so the calculation control unit 15 switches to a noise removal filter that can remove this noise.

In step S22, the calculation control unit 15 begins monitoring the level of sound information input from the sound collection unit 21 and the sound collection unit 22.

In step S23, the calculation control unit 15 determines whether the level of the sound information input from the sound collection unit 21 and the sound collection unit 22 is equal to or greater than a threshold value. Here, the threshold value is the aforementioned threshold value S1 or threshold value S2. If the threshold value S1 set in step S16 is used as the threshold value, threshold value S1 is a relatively large value, and therefore step S23 will not return a YES result unless a relatively loud sound is generated in the vehicle interior 12. This prevents an erroneous YES result due to noise in step S23. On the other hand, if the threshold value S2 set in step S17 is used as the threshold value, threshold value S2 is a relatively small value, and therefore step S23 will return a YES result even if the sound generated in the vehicle interior 12 is relatively quiet. This allows for accurate detection of abandonment even when the sound made by the abandoned infant 14 is quiet.

If the answer to step S23 is YES, i.e., if the level of the input sound information is equal to or greater than the threshold, the calculation control unit 15 proceeds to step S24 to determine whether the object has been left behind.

If step S23 returns NO, i.e., if the level of the input sound information is below the threshold, the calculation control unit 15 returns to step S22 and continues monitoring the level of the sound information.

In step S24, the calculation control unit 15 determines whether the sound matches the determination sound (determination trigger sound). Here, the calculation control unit 15 determines whether the sound matches by comparing the determination sound, which is a sound emitted in various situations, with the sound collected by the sound collection unit 21 and the sound collection unit 22.

The determination sounds used in step S24 are explained with reference to the table in Figure 5. Here, the determination sounds are classified into multiple categories, and the audio waveform characteristics, detection timer, and notes for each classification are shown.

First, the judged sounds are broadly classified into voice, hitting sounds, rubbing sounds, scratching sounds, breathing sounds, and environmental noise.

In addition, each major category is further divided into smaller categories. Specifically, sounds can be classified into infant speech, dog barking, cat meowing, etc. Knocking sounds can be classified into knocking on the door glass, knocking on the door trim, kicking on the seat, stomping on the floor, etc. Rubbing sounds can be classified into rubbing of clothes against a seat, rubbing of clothes against clothes, rubbing of a head against a seat, etc. Scratching sounds can be classified into scratching on a seat, scratching on the door glass, scratching on clothes, etc. Breathing sounds can be classified into human breathing, dog breathing, etc. Environmental noise can be classified into wind noise, street noise, shaking of the car body, conversation outside the car, etc.

Furthermore, frequency peaks, sound volume (gain), and continuity can be used as audio waveform characteristics for each determined sound classified in this way.

In this embodiment, the most likely detection targets are voice, hammering sounds, rubbing sounds, scratching sounds, breathing sounds, and environmental noise, in that order. In other words, voice is most likely to be a detection target, and environmental noise is least likely to be a detection target.

For this reason, the timer 23, which will be described later, is set according to each determined sound. For example, if the determined sound is a voice, the timer 23 is set to 300 seconds; if it is a slapping sound, it is set to 120 seconds; if it is a rubbing sound, it is set to 120 seconds; if it is a scratching sound, it is set to 30 seconds; if it is a breathing sound, it is set to 30 seconds; and if it is environmental noise, it is set to 10 seconds.

In step S24, the audio waveform characteristics of the trigger determination sound are compared with the audio waveform characteristics of the sound collected by the sound collection units 21 and 22. This comparison is performed, for example, with respect to one or more of frequency peak, magnitude (gain), and continuity. The determination sound is then compared with the sound collected by the sound collection units 21 and 22 with respect to one or more of frequency peak, magnitude (gain), and continuity, the closest classification is selected, and the detection timer for that classification is determined. For example, if the audio waveform characteristics of the sound collected by the sound collection units 21 and 22 are closest to the audio waveform characteristics of sheet scratching sounds, the time for setting the timer 23 (described below) is set to 120 seconds.

In this way, by setting the monitoring timer 23 according to each determination sound, it is possible to monitor for a longer period of time if the sound indicates a high possibility of abandonment, and for a shorter period of time if the sound indicates a low possibility of abandonment. This optimizes the time and amount of data required for monitoring, and reduces unnecessary power consumption by the vehicle parking detection device 10. Furthermore, when comparing the collected sound with the determination sound, rather than performing frequency component analysis, the comparison focuses on frequency peaks, magnitude, and continuity, thereby reducing the amount of calculations that the calculation control unit 15 must perform and reducing energy consumption.

If the answer to step S24 is YES, i.e., if the comparison between any of the detected sounds and the collected sound matches, the calculation control unit 15 proceeds to step S25.

If the answer to step S24 is NO, i.e., if the comparison between any of the determination sounds and the collected sound does not match, the calculation control unit 15 returns to step S22.

In step S25, the calculation and control unit 15 starts sound processing to determine whether or not a living being, such as an infant 14, is present in the vehicle compartment 12. Specifically, the calculation and control unit 15 collects sounds generated inside the vehicle compartment 12 using the sound collection unit 21 and the sound collection unit 22, and converts these sounds into waveform data. In this embodiment, the determination of whether or not a living being, such as an infant 14, is present in the vehicle compartment 12 is made from step S25 onwards.

In step S26, the calculation control unit 15 sets the timer 23 to monitor sounds generated inside the vehicle compartment 12. Specifically, in step S26, as shown in FIG. 5, if the sound collected by the sound collection unit 21 or the like is from a living thing such as an infant 14, the sound monitoring time is extended to, for example, approximately 300 seconds, and the amount of data used for monitoring is increased. On the other hand, if the sound collected by the sound collection unit 21 or the like is not from a living thing, the sound monitoring time is shortened to, for example, approximately 10 seconds, and the amount of data used for monitoring is reduced. In this way, the presence or absence of a living thing inside the vehicle compartment 12 can be accurately detected, and further, the energy consumed by the calculation control unit 15 can be reduced, preventing the battery from running out.

In step S27, the calculation control unit 15 determines whether the sounds collected by the sound collection unit 21 and the sound collection unit 22 are noise. Various methods can be used to determine whether a sound is noise. For example, it is possible to determine whether a sound is noise by comparing the sound collected by the sound collection unit 21 with the sound collected by the sound collection unit 22. In the graph shown in Figure 6, the horizontal axis represents time and the vertical axis represents sound pressure. Also in Figure 6, the dotted line represents the change over time in the sound pressure of the sound collected by the sound collection unit 21, and the solid line represents the change over time in the sound pressure of the sound collected by the sound collection unit 22. Here, if the difference in sound pressure and time difference between the two is large, it is highly likely that the sound source is inside the vehicle interior 12. On the other hand, if the difference in sound pressure and time difference between the two is small, it is highly likely that the sound source is outside the vehicle and is noise.

The sound collected by sound collection unit 21 is compared with the sound collected by sound collection unit 22, and if either or both of the sound pressure difference and time difference are equal to or greater than a predetermined value, it is determined that the sound is occurring inside the vehicle interior 12 and is not noise. On the other hand, if either or both of the sound pressure difference and time difference are less than a predetermined value, it is determined that the sound is occurring outside the vehicle interior 12 and is noise.

If the answer to step S27 is NO, i.e., if the sound is not noise, the calculation control unit 15 proceeds to step S28 to determine whether the sound originates from a living organism.

If the answer to step S27 is YES, i.e., if the sound is noise, the calculation control unit 15 proceeds to step S22.

In step S28, the calculation control unit 15 performs frequency analysis to determine whether the sound collected by the sound collection unit 21 and the sound collection unit 22 is a sound. In step S28, when performing the sound determination, well-known formant analysis can also be performed on the sound collected by the sound collection unit 21 and the sound collection unit 22. In formant analysis, the calculation control unit 15 analyzes the temporal peaks of the frequency characteristics contained in the collected sound, prepares a formant feature table to be detected, evaluates the consistency between the collected sound and the formant feature table, and thereby estimates the presence of a living organism inside the vehicle cabin 12.

If the collected sound originates from a human, such as an infant 14, the calculation control unit 15 proceeds to step S29. Furthermore, if the collected sound originates from an animal other than a human, the calculation control unit 15 proceeds to step S30. Furthermore, if the collected sound does not originate from a human or an animal, the calculation control unit 15 proceeds to step S31.

In step S29, the calculation control unit 15 determines that the infant 14 is left behind in the vehicle compartment 12, and proceeds to step S32. Furthermore, in step S29, the presence of the infant 14 may be determined with even greater accuracy using a camera or millimeter-wave radar installed in the vehicle compartment.

In step S30, the calculation control unit 15 determines that a non-human animal, such as a dog or cat, has been left behind in the vehicle interior 12, and proceeds to step S32.

In step S31, the calculation control unit 15 determines that no one is present in the vehicle compartment 12.

In step S32, the calculation control unit 15 ends the audio determination process that began in step S25.

In step S33, the calculation control unit 15 ends the audio determination process that began in step S25, returns to step S22, and continues monitoring the input audio level.

In step S34, the calculation and control unit 15 notifies that an abandoned child has occurred in the vehicle 11. Specifically, the calculation and control unit 15 uses the communication unit 25 to notify a smartphone or other device owned by the owner of the vehicle 11 that an abandoned child has occurred in the vehicle 11. Here, the person to whom the calculation and control unit 15 notifies is not limited to the owner, and the calculation and control unit 15 may notify multiple people registered as emergency contacts, i.e., users who regularly use the vehicle 11. Furthermore, the calculation and control unit 15 may use the alarm unit 17 to sound the horn, flash the hazard lights or headlights, or the like, to notify those around the vehicle 11 that an abandoned child has occurred. Furthermore, the calculation and control unit 15 may operate the heating or cooling unit of the vehicle 11 to prevent the infant 14 from suffering from heatstroke or hypothermia.

In step S35, a decision is made as to whether or not to report the incident. This decision can be made by the person receiving the notification operating their smartphone.

If the answer to step S35 is YES, i.e., if a notification is to be made, the calculation control unit 15 proceeds to step S36.

If the answer to step S35 is NO, i.e., if a report is not to be made, the calculation control unit 15 proceeds to step S37.

In step S36, the calculation and control unit 15 notifies the telematics service call center, police, fire department, etc. In this way, it is possible to quickly rescue any infants 14 or others left behind in the vehicle interior 12.

In step S37, the calculation control unit 15 ends the in-vehicle voice determination flow. For example, the driver 20 rescues the infant 14 left behind in the vehicle 11.

The above is an explanation of the method for detecting abandonment in the vehicle compartment 12 using the vehicle abandonment detection device 10 according to this embodiment.

The above-described embodiment can achieve the following main effects:

In this embodiment, the calculation control unit 15 determines whether or not a living thing is present in the vehicle compartment 12 based on audio information input from the sound collection unit 21, etc., and changes the time required to make the determination or the amount of data used for the determination depending on the audio information. In this way, when determining whether or not a living thing such as a child has been left behind inside the vehicle 11, the time or amount of data required to make the determination can be changed depending on the audio information, thereby suitably adjusting the power required for the determination. This prevents excessive power consumption by the vehicle abandonment detection device 10 while the vehicle is stopped. Furthermore, because the vehicle abandonment detection device 10 uses an inexpensive microphone to determine whether or not an object has been left behind, rather than using expensive sensors such as conventional load sensors, cameras, or millimeter waves, the vehicle abandonment detection device 10 can be constructed inexpensively.

Furthermore, if the calculation control unit 15 determines that there is a high possibility that the audio information originates from a living organism, it increases the time or data volume. By doing this, if there is a high possibility that the audio information originates from a living organism, the presence of a living organism inside the vehicle can be detected more accurately by increasing the time or data volume.

Furthermore, when the calculation and control unit 15 detects the opening and closing of the door 16 of the vehicle 11, it determines whether or not a living creature is present in the passenger compartment 12. In this way, the calculation and control unit 15 does not operate when no living creature has been left behind, thereby reducing battery consumption.

Furthermore, when the calculation control unit 15 detects the opening and closing of only the front right door 162, it sets a smaller threshold value than when it detects the opening and closing of both the front right door 162 and the rear right door 164. By doing this, when the calculation control unit 15 detects the opening and closing of only the front right door 162, there is a high possibility that only the driver 20 has exited the vehicle, leaving the child 14 seated in the back seat behind in the vehicle 11. In such cases, more reliable audio detection can be performed to accurately detect abandonment.

Furthermore, the calculation control unit 15 changes the filter that processes sound depending on whether the vehicle 11's windows are open or closed. By doing this, when the windows are open, noise entering from outside the vehicle can be removed, making it possible to more accurately detect whether someone has been left behind.

Although the embodiments of the present invention have been described above, the present invention is not limited to these and can be modified within the scope of the present invention. Furthermore, the above-described embodiments can be combined with each other.

10 Vehicle abandonment detection device 11 Vehicle 12 Vehicle compartment 14 Infant 15 Calculation control unit 161 Front left door 162 Front right door 163 Rear left door 164 Rear right door 17 Notification unit 18 Vehicle body 191 Front left seat 192 Front right seat 193 Rear left seat 194 Rear right seat 20 Driver 21 Sound collection unit 22 Sound collection unit 23 Timer 24 Opening/closing sensor 25 Communication unit

Claims (5)

A vehicle abandonment detection device that detects whether a living thing has been left inside a vehicle compartment,
The device includes a sound collection unit and a calculation control unit,
The sound collection unit collects sounds generated in the vehicle interior,
The arithmetic and control unit
comparing the sound waveform characteristics of the determined sounds classified into a plurality of categories with the sound waveform characteristics of the sound collected by the sound collecting unit;
Selecting a classification that is closest to the audio waveform characteristics of the sound collected by the sound collection unit from among the audio waveform characteristics of the determination sounds classified into the plurality of classifications;
A vehicle abandonment detection device characterized by determining the time for audio monitoring based on the selected closest classification, and determining whether the living thing is present in the vehicle cabin based on the sound collected by the sound collection unit during the determined time for audio monitoring . The vehicle abandonment detection device described in claim 1, characterized in that the judgment sound is classified according to the degree of possibility that it is derived from the living organism, and the time for which the sound monitoring is performed is set to be longer the greater the possibility that it is derived from the living organism . The vehicle abandonment detection device according to claim 1 or 2, characterized in that the calculation control unit, when detecting the opening or closing of the vehicle door, determines whether the living thing is present in the vehicle compartment. The arithmetic and control unit
When the sound collected by the sound collecting unit is louder than a threshold value, it is determined whether or not the living thing is present in the vehicle interior;
When detecting the opening and closing operation of only the front door of the vehicle, the number of times of detection is larger than when detecting the opening and closing operation of the front door and the rear door of the vehicle.
4. The vehicle abandonment detection device according to claim 3, wherein the threshold value is reduced. A vehicle abandonment detection device as described in any one of claims 1 to 4, characterized in that the calculation control unit changes the filter that processes the sound collected by the sound collection unit depending on whether the vehicle window is open or closed.