Abnormal state detection system and method, and storage medium

Abstract

The present invention addresses the problem of providing an abnormal state detection system, an abnormal state detection method, an abnormal state detection computer-readable storage medium, and a program therefor, whereby it is possible to detect an abnormal state of a vehicle without preparing amazing data in advance. The abnormal state detection system includes: a voice abnormality detection unit including a model generation unit that generates, as a normal model, a model of normal voice data of a person taking a vehicle on the basis of voice data including voice of the person; detecting the voice abnormity based on the current voice data of the person and the normal model; a vehicle body abnormality detection unit that detects an abnormality in a vehicle body of the vehicle on the basis of acceleration data of the vehicle; and an abnormal state determination unit configured to determine whether the vehicle is in an abnormal state based on a detection result of the voice abnormality detection unit and a detection result of the vehicle body abnormality detection unit.

Description

Technical Field

[0001] The present disclosure relates to an abnormal state detection system, an abnormal state detection method, and an abnormal state detection storage medium. Background Art

[0002] Generally, there are known technologies for detecting abnormal conditions of vehicles such as accidents. In addition, as such technologies, there are known technologies for detecting the occurrence of accidents by detecting screams of people riding in vehicles. For example, Patent Document 1 describes the following technology: a model of shouting, which is a model learned from feature quantities of multiple shouting sounds, is prepared in advance, and based on the feature quantities extracted from the input voice, the acoustic model group, and the likelihood, whether the input voice is a shouting sound is determined.

[0003] [Prior art literature]

[0004] [Patent Document]

[0005] [Patent Document 1] Japanese Patent Application Publication No. 2017-187676 Summary of the invention

[0006] [Problems to be solved by the invention]

[0007] However, when the data of screams is prepared in advance as in the technology described in Patent Document 1, data of various screams must be prepared, which increases the amount of data. In addition, screams of patterns that have not been prepared in advance cannot be identified as screams. Therefore, the prior art is not sufficient to detect abnormal conditions of vehicles.

[0008] An object of the present disclosure is to provide an abnormal state detection system, an abnormal state detection method, and an abnormal state detection program that can detect an abnormal state of a vehicle even without preparing scream data in advance.

[0009] [Technical means to solve the problem]

[0010] In order to achieve the above purpose, the abnormal state detection system disclosed in the present invention includes: a voice abnormality detection unit, which includes the following model generation unit, namely, a normal time model generation unit, which generates a model of the normal voice data of a person riding in a vehicle based on voice data including the voice of the person, as a normal time model, and detects the abnormality of the voice based on the current voice data of the person and the normal time model; a body abnormality detection unit, which detects the abnormality of the body of the vehicle based on the acceleration data of the vehicle; and an abnormal state determination unit, which determines whether the vehicle is in an abnormal state based on the detection results of the voice abnormality detection unit and the detection results of the body abnormality detection unit.

[0011] In addition, in order to achieve the above purpose, the abnormal state detection method disclosed in the present invention utilizes a processor included in an abnormal state detection system to generate a model of the normal voice data of a person riding in a vehicle based on voice data including the voice of the person as a normal time model, and detects voice abnormalities based on the current voice data of the person and the normal time model, detects abnormalities of the body of the vehicle based on the acceleration data of the vehicle, and outputs a determination result of whether the vehicle is in an abnormal state based on the detection results of the voice abnormality and the detection results of the body abnormality.

[0012] In addition, in order to achieve the above purpose, the abnormal state detection computer-readable storage medium disclosed in the present invention stores a program for causing the processor included in the abnormal state detection system to perform the following processing: generating a model of the normal speech data of a person riding in a vehicle as a normal model based on speech data of the person's speech, detecting speech abnormalities based on the person's current speech data and the normal model, detecting abnormalities of the vehicle body based on the acceleration data of the vehicle, and outputting a determination result of whether the vehicle is in an abnormal state based on the detection results of the speech abnormality and the detection results of the vehicle body abnormality.

[0013] [Effects of the Invention]

[0014] According to the present disclosure, even without preparing scream data in advance, it is possible to detect an abnormal state of a vehicle. BRIEF DESCRIPTION OF THE DRAWINGS

[0015] Figure 1 This is a diagram for explaining the outline of an abnormal state detection system.

[0016] Figure 2 This is a circuit diagram showing an example of the hardware configuration of the abnormal state detection system.

[0017] Figure 3 This is a structural diagram showing an example of the structure of an abnormal state detection system.

[0018] Figure 4 It is a diagram for explaining the function of the abnormal state detection system.

[0019] Figure 5 This is a diagram for explaining a data frequency / data accumulation degree measuring unit.

[0020] Figure 6 This is a flowchart showing an example of the flow of abnormal state detection processing.

[0021] Figure 7 This is a flowchart showing an example of the flow of speech anomaly detection processing.

[0022] Figure 8This is a flowchart showing an example of the flow of vehicle body abnormality detection processing.

[0023] [Explanation of Symbols]

[0024] 1: Abnormal state detection system

[0025] 51: CPU

[0026] 10: Voice Anomaly Detection Unit

[0027] 12: Vehicle body abnormality detection unit

[0028] 14: Abnormal status determination unit

[0029] 28: Current model generation department

[0030] 32: Normal model generation unit DETAILED DESCRIPTION

[0031] Hereinafter, embodiments of the present disclosure will be described in detail with reference to the accompanying drawings. In addition, the following embodiments do not limit the technology of the present disclosure.

[0032] like Figure 1 As shown, the abnormal state detection system 1 of this embodiment is equipped inside the body 3 of the vehicle 2. The abnormal state detection system 1 is a system that determines whether the vehicle 2 is in an abnormal state based on the abnormality of the voice of the person riding in the vehicle 2 and the abnormality of the body 3. In addition, in this embodiment, the so-called "person riding in the vehicle 2" specifically refers to all people existing in the body 3, including people other than the driver.

[0033] Figure 2 FIG. 2 shows an example of the hardware configuration of the abnormal state detection system 1 according to the present embodiment. Figure 2 As shown, the abnormal state detection system 1 of this embodiment includes an acceleration sensor 40, a control unit 50, a storage unit 52, a sound collector 54, a recording device 56, and a communication interface (Interface, I / F) 60. The acceleration sensor 40, the control unit 50, the storage unit 52, the sound collector 54, the recording device 56, and the communication I / F 60 are connected via a bus 69 such as a system bus or a control bus so that various information can be transmitted and received to each other.

[0034] The control unit 50 includes a central processing unit (CPU) 51, and controls the entire abnormal state detection system 1. The CPU 51 of the present embodiment is an example of a processor of the present disclosure.

[0035] The storage unit 52 stores the voices of the persons riding in the vehicle 2, or the details thereof. The current model generating unit 28 and the display area 62 (both refer to Figure 3) etc. The storage unit 52 is implemented by a storage medium such as a hard disk drive (HDD), a solid state drive (SSD) or a flash memory.

[0036] The sound collector 54 has a function of collecting the voice of the person riding in the vehicle 2, and is, for example, a microphone. The recording device 56 has a function of recording the voice collected by the sound collector 54, and the recorded voice is stored in the storage unit 52. In addition, the acceleration sensor 40 has a function of detecting the acceleration of the body 3 of the vehicle 2.

[0037] The communication I / F 60 communicates various information with an external device of the abnormal state detection system 1 or a device outside the vehicle 2 through wireless communication or wired communication.

[0038] Furthermore, refer to Figure 3 The structure of the abnormal state detection system 1 of this embodiment is described. Figure 3 As shown, the abnormal state detection system 1 of the present embodiment includes a voice abnormality detection unit 10 , a vehicle body abnormality detection unit 12 , and an abnormal state determination unit 14 .

[0039] The speech anomaly detection unit 10 includes a speech acquisition unit 20, a speech extraction unit 22, a spectrum creation unit 24, a data frequency / data accumulation measurement unit 26, a current model generation unit 28, a multiplexer 30, a normal model generation unit 32, a difference detection unit 34, and a multiplexer 36.

[0040] The voice acquisition unit 20 includes the sound collector 54 and the recording device 56. Figure 4 As shown, the voice acquisition unit 20 has a function of acquiring voice data including the voice of the person 6 riding in the vehicle 2. In fact, the sound collected by the sound collector 54 of the voice acquisition unit 20 also includes sounds other than the voice of the person 6 riding in the vehicle 2. As sounds other than the voice of the person 6, sounds emitted by the vehicle 2 itself or environmental sounds can be listed. In this embodiment, the sounds other than the voice of the person 6 in the sounds collected by the sound collector 54 of the voice acquisition unit 20 are called "noise".

[0041] The voice extraction unit 22 has a function of extracting the voice data of the person 6 riding in the vehicle 2 from the voice data acquired by the voice acquisition unit 20. In other words, the voice extraction unit 22 has a function of removing noise from the voice data acquired by the voice acquisition unit 20. The method by which the voice extraction unit 22 extracts the voice of the person 6, that is, the human voice, from the acquired voice data is not particularly limited, and a known method can be used. For example, a filter that only passes the frequency of the human voice can be applied.

[0042] The spectrum creation unit 24 has a function of creating a waveform histogram for the speech data output from the speech extraction unit 22. In addition, the method by which the spectrum creation unit 24 generates a waveform histogram is not particularly limited, and a known method can be used. For example, a waveform histogram can be generated by performing a Fast Fourier Transform (FFT) or a Discrete Fourier Transform (DFT) on the speech data. In this way, the frequency characteristics of the speech data can be known, and the characteristics of the speech data become clear. Figure 4 As shown in the figure, the waveform histogram of normal speech data has different characteristics from the waveform histogram of emergency situations such as accidents. The frequency of speech data corresponds to the height of the sound, and the amplitude corresponds to the size of the sound. Figure 4 As shown, in the waveform histogram during normal times, there are many waveforms with small amplitudes and low frequencies. On the other hand, compared with normal voices, screams tend to be louder and higher in volume. Therefore, in an emergency, waveforms with large amplitudes and high frequencies are included. In this way, when the waveform histogram includes features that are significantly deviated from the features included in the waveform histogram during normal times, it can be determined that the voice data includes screams.

[0043] The data frequency / data accumulation measurement unit 26 has the function of measuring the frequency and accumulation of voice data. The data frequency / data accumulation measurement unit 26 determines whether the voice data is correct and sufficient by measuring the frequency and accumulation of voice data. The data frequency / data accumulation measurement unit 26 of this embodiment divides the interval into intervals according to each specified time, measures the frequency of data in each interval, and uses the interval where the frequency of data is higher than a predetermined threshold as the effective period of voice data. Figure 5 In the example shown, periods 70_1, 70_2, and 70_3 are valid periods for voice data due to their high data frequencies. On the other hand, period 72 has a low data frequency and is therefore outside the valid period for voice data. That is, the voice data for period 72 is excluded from the creation of the normal time model. In addition, the so-called low data frequency is when the person 6 speaks less. Therefore, in the case of low data frequency, the abnormal state detection system 1 preferably encourages the person 6 to speak.

[0044] The current model generation unit 28 has a function of generating a current voice data model (referred to as a "current model" in this embodiment) using the current voice data. In addition, the method by which the current model generation unit 28 generates the current model is not particularly limited. For example, the current model as a classifier of voice features can be generated by performing opportunistic learning using a waveform histogram of voice data as input and without using teacher data of the model using a known method.

[0045] The multiplexer 30 has the following functions: taking the current model generated by the current model generating unit 28 as input, based on the measurement result of the data frequency / data accumulation measuring unit 26, outputting data obtained by excluding speech data with low data frequency as deviation values ​​from the speech data in normal times.

[0046] The normal time model generation unit 32 has a function of generating a model (normal time model) of the normal time speech of the character 6. The method by which the normal time model generation unit 32 generates the normal time model is not particularly limited, and can be set to be the same as the method by which the current model generation unit 28 generates the current model.

[0047] The difference detection unit 34 has a function of sequentially comparing the current model generation unit 28 with the normal model generation unit 32 and outputting the comparison result. As described above, the difference detection unit 34 compares the waveform histogram of the current model generation unit 28 with the waveform histogram of the normal model generation unit 32, and outputs information indicating the above when the difference exceeds a predetermined threshold. That is, the difference detection unit 34 outputs information indicating the above when detecting a scream.

[0048] The multiplexer 36 has a function of outputting a detection result of "abnormality exists" to the abnormal state determination unit 14 when information indicating that a scream has been detected is input from the difference detection unit 34. On the other hand, the multiplexer 36 has a function of outputting a detection result of "no abnormality" to the abnormal state determination unit 14 when information indicating that a scream has been detected is not input from the difference detection unit 34. Specifically, the multiplexer 36 of the present embodiment outputs "1" to the abnormal state determination unit 14 when "abnormality exists", and outputs "0" to the abnormal state determination unit 14 when "no abnormality" is detected.

[0049] In this way, according to the voice anomaly detection unit 10, when the person 6 makes a scream, a detection result indicating the presence of an abnormality can be output.

[0050] On the other hand, the vehicle body abnormality detection unit 12 includes the acceleration sensor 40, the difference detection unit 42, and the multiplexer 44. As described above, the acceleration sensor 40 has a function of outputting acceleration data of the vehicle 2.

[0051] The difference detection unit 42 has a function of successively outputting the change of the acceleration data as a difference. The difference becomes larger when the vehicle 2 collides or brakes suddenly. Therefore, when the change (difference) of the acceleration data exceeds a predetermined threshold, the difference detection unit 42 outputs information indicating that an abnormality of the vehicle body 3 is detected.

[0052] The multiplexer 44 has the following function: when information indicating that an abnormality of the vehicle body 3 is detected is input from the difference detection unit 42, the detection result of "abnormality exists" is output to the abnormal state determination unit 14. On the other hand, it has the following function: when information indicating that an abnormality of the vehicle body 3 is detected is not input from the difference detection unit 42, the detection result of "no abnormality" is output to the abnormal state determination unit 14. Specifically, the multiplexer 44 of this embodiment outputs "1" to the abnormal state determination unit 14 in the case of "abnormality exists", and outputs "0" to the abnormal state determination unit 14 in the case of "no abnormality".

[0053] In this way, when an abnormality occurs in the vehicle body 3 due to a collision of the vehicle 2 or the like, the vehicle body abnormality detection unit 12 can output a detection result indicating the presence of the abnormality.

[0054] The abnormal state determination unit 14 has the following functions: based on the detection result of the voice abnormality detection unit 10 and the detection result of the vehicle body abnormality detection unit 12, it determines whether the vehicle 2 is in an abnormal state, and outputs the determination result. As an example, the abnormal state determination unit 14 of this embodiment is an OR circuit, and in at least one of the cases where the detection result of the voice abnormality detection unit 10 is an abnormal situation and the detection result of the vehicle body abnormality detection unit 12 is an abnormal situation, the determination result that the vehicle 2 is in an abnormal state is output to the outside of the voice abnormality detection unit 10.

[0055] Next, the operation of the abnormal state detection system 1 according to the present embodiment will be described with reference to the drawings. Figure 6 1 is a flowchart showing an example of the process of abnormal state detection processing executed by the CPU 51 of the control unit 50. In the present embodiment, the CPU 51 executes an abnormal state detection processing program (omitted from the diagram) stored in the control unit 50, thereby executing Figure 6 As an example, in this embodiment, when the engine of the vehicle 2 is installed, the abnormal state detection process is executed. Figure 6 The abnormal state detection process is shown.

[0056] exist Figure 6 In step S100, the CPU 51 starts the process of detecting the abnormality of the voice through the voice abnormality detection unit 10. Specifically, the CPU 51 starts the abnormal state detection process (see Figure 7 ).

[0057] In the next step S102, the CPU 51 starts a process of detecting an abnormality of the vehicle body 3 through the vehicle body abnormality detection unit 12. Specifically, the CPU 51 starts a vehicle body state detection process (see Figure 8 ).

[0058] In the next step S104, it is determined whether the vehicle body 3 is in an abnormal state. Based on the detection results of the above steps S100 and S102, the detection results are output from the abnormal state determination unit 14 as described above. In the case where the detection result output from the abnormal state determination unit 14 indicates that the vehicle body 3 is not in an abnormal state, the determination of step S104 is a negative determination, and the process is transferred to step S108. On the other hand, in the case where it indicates that the vehicle body 3 is in an abnormal state, the determination of step S104 is a positive determination, and the process is transferred to step S106.

[0059] In step S106 , the CPU 51 outputs information indicating that the vehicle body 3 is in an abnormal state to the outside via the communication I / F 60 .

[0060] In the next step S108, the CPU 51 determines whether to terminate the abnormal state detection process. As an example, in the present embodiment, when the engine of the vehicle 2 is turned off, the abnormal state detection process is terminated. Therefore, until the engine of the vehicle 2 is turned off, the determination of step S108 is negative, and the process returns to step S104, and the process of step S104 and step S106 is repeated. On the other hand, when the engine of the vehicle 2 is turned off, the determination of step S108 is positive, and the process is terminated. Figure 6 The abnormal state detection process is shown.

[0061] Next, refer to Figure 7 The abnormal speech state detection process is described. Figure 7 A flowchart showing an example of the flow of speech abnormal state detection processing is shown in FIG.

[0062] exist Figure 7 In step S120, the CPU 51 causes the voice acquisition unit 20 to start acquiring voice data.

[0063] In the next step S122, the CPU 51 determines whether a normal time model is generated. In the present embodiment, a normal time model is generated at a specified time point. Therefore, it is determined whether it is a specified time point. In the case where a normal time model is not generated, the determination of step S122 is a negative determination, and the process moves to step S126. On the other hand, in the case where a normal time model is generated, the determination of step S122 is a positive determination, and the process moves to step S124.

[0064] In step S124 , the CPU 51 generates a normal-time model by the normal-time model generating unit 32 as described above.

[0065] In the next step S126 , the CPU 51 causes the current model generating section 28 to generate a current model as described above.

[0066] In the next step S128 , as described above, the CPU 51 causes the difference detection unit 34 to detect the difference between the normal-time model and the current model.

[0067] In the next step S130 , the CPU 51 outputs the detection result of the abnormal voice detection section 10 to the abnormal state determination section 14 .

[0068] In the next step S134, the CPU 51 determines whether to end the voice anomaly detection process. As described above, until the engine of the vehicle 2 is turned off, the determination of step S132 is negative, and the process returns to step S122 to repeat the process of step S122 to step S130. On the other hand, when the engine of the vehicle 2 is turned off, the determination of step S132 is positive, and the process is transferred to step S134.

[0069] In step S134, the CPU 51 ends the acquisition of the voice data by the voice acquisition unit 20. When the processing of step S134 ends, Figure 7 The illustrated speech anomaly detection process ends.

[0070] Furthermore, refer to Figure 8 The vehicle body abnormal state detection process will be described. Figure 8 A flowchart showing an example of the flow of vehicle body abnormal state detection processing is shown in FIG.

[0071] exist Figure 8 In step S140 , the CPU 51 causes the acceleration sensor 40 to start acquiring acceleration data.

[0072] In the next step S142 , as described above, the CPU 51 causes the difference detection section 42 to detect the difference in acceleration.

[0073] In the next step S144 , the CPU 51 outputs the detection result of the vehicle body abnormality detection unit 12 to the abnormal state determination unit 14 .

[0074] In the next step S146, the CPU 51 determines whether to end the vehicle body abnormality detection process. As described above, until the engine of the vehicle 2 is turned off, the determination of step S146 is negative, and the process returns to step S142 to repeat the processes of step S142 and step S144. On the other hand, when the engine of the vehicle 2 is turned off, the determination of step S146 is positive, and the process is transferred to step S148.

[0075] In step S148, the CPU 51 ends the acquisition of acceleration data by the acceleration sensor 40. When the processing of step S148 ends, the CPU 51 ends. Figure 8 The vehicle body abnormality detection process is shown.

[0076] As described above, the abnormal state detection system 1 of the present embodiment includes a voice abnormality detection unit 10, and the voice abnormality detection unit 10 includes a normal-time model generation unit 32, that is, the normal-time model generation unit 32 generates a model of the normal-time voice data of the person 6 as a normal-time model based on the voice data including the voice of the person 6 riding in the vehicle 2, and detects the abnormality of the voice based on the current voice data of the person 6 and the normal-time model. In addition, the abnormal state detection system 1 includes a vehicle body abnormality detection unit 12, and the vehicle body abnormality detection unit 12 detects the abnormality of the vehicle body 3 of the vehicle 2 based on the acceleration data of the vehicle 2. Furthermore, the abnormal state detection system 1 includes an abnormal state determination unit 14, and the abnormal state determination unit 14 determines whether the vehicle 2 is in an abnormal state based on the detection results of the voice abnormality detection unit 10 and the detection results of the vehicle body abnormality detection unit 12.

[0077] According to the above structure, normal voice of the person 6 can be acquired while riding in the vehicle 2, and screams can be detected based on the normal voice. Therefore, according to the abnormal state detection system 1 of this embodiment, even if scream data is not prepared in advance, the abnormal state of the vehicle can be detected.

[0078] The following additional notes are further disclosed with respect to the above-mentioned embodiments.

[0079] (Note 1)

[0080] An abnormal state detection system, comprising:

[0081] A voice anomaly detection unit, comprising a model generation unit, that is, a normal time model generation unit, which generates a model of the normal voice data of the person as a normal time model based on the voice data including the voice of the person riding in the vehicle, and detects the abnormality of the voice based on the current voice data of the person and the normal time model;

[0082] a vehicle body abnormality detection unit that detects an abnormality of a vehicle body of the vehicle based on acceleration data of the vehicle; and

[0083] The abnormal state determination unit determines whether the vehicle is in an abnormal state based on the detection result of the voice abnormality detection unit and the detection result of the vehicle body abnormality detection unit.

[0084] (Note 2)

[0085] According to the abnormal state detection system described in Supplement 1,

[0086] The speech anomaly detection unit also includes a current model generating unit, that is, a current model generating unit that generates a model of the current speech data of the character as a current model based on the current speech data of the character, and sets the speech anomaly as a current model when it is determined that the current speech data of the character contains a scream based on a comparison result between the current model and the normal model.

[0087] (Note 3)

[0088] The abnormal state detection system according to Supplement 1 or Supplement 2, wherein

[0089] The voice anomaly detection unit determines that the voice anomaly is detected when it is determined that the current voice data of the person includes a scream based on a comparison result between a waveform histogram obtained from normal voice data and a waveform histogram obtained from current voice data.

[0090] (Note 4)

[0091] An abnormal state detection system according to any one of Notes 1 to 3, wherein

[0092] The abnormal state determination unit determines that the vehicle is in an abnormal state when at least one of the detection result of the voice abnormality detection unit is a detection result of voice abnormality and the detection result of the vehicle body abnormality detection unit is a detection result of vehicle body abnormality.

[0093] (Note 5)

[0094] A method for detecting abnormal state,

[0095] Using a processor included in the abnormal state detection system,

[0096] A model of the normal speech data of a person riding in a vehicle is generated based on speech data including the speech of the person riding in a vehicle as a normal model, and abnormalities in speech are detected based on the current speech data of the person and the normal model.

[0097] detecting an abnormality of a body of the vehicle based on acceleration data of the vehicle,

[0098] Based on the detection result of the voice abnormality and the detection result of the vehicle body abnormality, a determination result of whether the vehicle is in an abnormal state is output.

[0099] (Note 6)

[0100] A computer-readable storage medium for detecting an abnormal state stores a program for causing a processor included in an abnormal state detection system to perform the following processing:

[0101] A model of the normal speech data of a person riding in a vehicle is generated based on speech data including the speech of the person riding in a vehicle as a normal model, and abnormalities in speech are detected based on the current speech data of the person and the normal model.

[0102] detecting an abnormality of a body of the vehicle based on acceleration data of the vehicle,

[0103] Based on the detection result of the voice abnormality and the detection result of the vehicle body abnormality, a determination result of whether the vehicle is in an abnormal state is output.

Claims

1. An abnormal state detection system, comprising: A voice anomaly detection unit, comprising a model generation unit, that is, a normal time model generation unit, which generates a model of the normal voice data of the person as a normal time model based on the voice data including the voice of the person riding in the vehicle, and detects the abnormality of the voice based on the current voice data of the person and the normal time model; a body abnormality detection unit, which detects an abnormality of a body of the vehicle based on acceleration data of the vehicle; as well as The abnormal state determination unit determines whether the vehicle is in an abnormal state based on the detection result of the voice abnormality detection unit and the detection result of the vehicle body abnormality detection unit.

2. The abnormal state detection system according to claim 1, wherein The speech anomaly detection unit also includes a current model generating unit, that is, a current model generating unit that generates a model of the current speech data of the character as a current model based on the current speech data of the character, and sets the speech anomaly as a current model when it is determined that the current speech data of the character contains a scream based on a comparison result between the current model and the normal model.

3. The abnormal state detection system according to claim 1, wherein The voice anomaly detection unit determines that the voice anomaly is detected when it is determined that the current voice data of the person includes a scream based on a comparison result between a waveform histogram obtained from normal voice data and a waveform histogram obtained from current voice data.

4. The abnormal state detection system according to claim 1, wherein The abnormal state determination unit determines that the vehicle is in an abnormal state when at least one of the detection result of the voice abnormality detection unit is a detection result of voice abnormality and the detection result of the vehicle body abnormality detection unit is a detection result of vehicle body abnormality.

5. A method for detecting abnormal conditions, Using a processor included in the abnormal state detection system, A model of the normal speech data of a person riding in a vehicle is generated based on speech data including the speech of the person riding in a vehicle as a normal model, and abnormalities in speech are detected based on the current speech data of the person and the normal model. detecting an abnormality of a body of the vehicle based on acceleration data of the vehicle, Based on the detection result of the voice abnormality and the detection result of the vehicle body abnormality, a determination result of whether the vehicle is in an abnormal state is output.

6. A computer-readable storage medium for abnormal state detection, storing a program for causing a processor included in an abnormal state detection system to perform the following processing: A model of the normal speech data of a person riding in a vehicle is generated based on speech data including the speech of the person riding in a vehicle as a normal model, and abnormalities in speech are detected based on the current speech data of the person and the normal model. detecting an abnormality of a body of the vehicle based on acceleration data of the vehicle, Based on the detection result of the voice abnormality and the detection result of the vehicle body abnormality, a determination result of whether the vehicle is in an abnormal state is output.

Images