JP7784656B2 - Danger recognition and/or danger prediction evaluation device - Google Patents

Description

The present invention relates to a device for evaluating hazard perception and/or hazard prediction, which can evaluate whether a subject's hazard perception and/or hazard prediction is appropriate.

Driving simulators are used to evaluate driving characteristics at driving schools, driver's license centers, etc. Evaluation of driving characteristics includes evaluation of driving ability and driving tendencies.
In recent years, the increase in car accidents caused by elderly people has become a social problem, and there is a demand for technology that can more accurately evaluate the driving characteristics of elderly people.

Patent Document 1 discloses a cognitive behavioral assessment device for automobile driving that can determine whether a subject recognizes and/or anticipates danger based on the amount of sweating on the palm of the subject's hand in response to video of situations requiring caution while driving (for example, situations requiring caution for safe driving, such as when a traffic light changes from green to yellow, when approaching and stopping a vehicle waiting at a traffic light, or when a group of children are walking).

However, the technology in Patent Document 1 had the problem that it was difficult to objectively evaluate whether the subject's danger recognition and/or danger prediction was appropriate for the situation requiring caution.

Patent No. 5366248

The present invention has been made in consideration of the above circumstances, and aims to provide an evaluation device for danger recognition and/or danger prediction that can objectively evaluate whether a subject's danger recognition and/or danger prediction is appropriate as danger recognition and/or danger prediction for a situation requiring caution.
In the present invention, danger recognition means that a subject who encounters a situation requiring caution that is potentially dangerous, such as "a situation in which a person suddenly appears in front of them," recognizes the danger. In the present invention, danger prediction means that a subject who encounters a situation requiring caution that is potentially dangerous, such as "a situation in which a person suddenly appears in front of them," predicts the danger that may be encountered (for example, at an intersection with poor visibility, there is a possibility that a person may suddenly appear in front of them).
In the present invention, situations requiring caution are not limited to situations requiring caution related to driving a car as described above, but also include situations requiring caution related to working at heights, situations requiring caution related to disasters, etc.

The present invention provides the following risk perception and/or risk prediction evaluation devices [1] to [9].

[1] A device for evaluating danger recognition and/or danger prediction, comprising: an image presentation device that presents an image including a situation requiring caution to a subject; a sweat rate sensor attached to the palm and/or sole of the subject's foot to detect the amount of sweat; a cerebral blood flow sensor that detects the cerebral blood flow of the subject; and a correlation evaluator that determines the appropriateness of danger recognition and/or danger prediction for the situation requiring caution based on the correlation between the time change in the sweat rate and the time change in the cerebral blood flow.
[2] The device for evaluating danger perception and/or danger prediction described in [1], wherein the sweat rate sensor comprises an air supply means, a housing capsule whose opening is in close contact with the skin surface of the subject and which mixes the sweat dispersed from the skin surface with the air supplied from the air supply means in its internal space, an exhaust passageway which exhausts the mixed gas formed by mixing the sweat and the air supplied from the air supply means in the internal space of the housing capsule to the outside of the housing capsule, and a measuring means which measures the amount of sweat on the skin surface based on the difference between the humidity of the air supplied from the air supply means to the housing capsule and the humidity of the mixed gas exhausted from the exhaust passage.
[3] The risk perception and/or risk prediction evaluation device according to [1] or [2], which has a storage device that stores a standard pattern of the correlation, and performs the judgment by comparing with the standard pattern.
[4] The risk perception and/or risk prediction evaluation device according to any one of [1] to [3], comprising a skin potential sensor attached to the palm and/or sole of the foot to detect the skin potential reflex, an operation amount sensor to detect the operation amount of an operation device operated by the subject, and a recording unit to record the detection values detected by the sweat rate sensor, the cerebral blood flow sensor, the skin potential sensor, and the operation amount sensor.
[5] The risk perception and/or risk prediction evaluation device described in [4] is equipped with a reaction latency measurement unit that determines the time from the time the video presentation device displays the video of the caution situation when the skin potential reflex detected by the skin potential sensor first fluctuates beyond a predetermined value, identifies the reaction start time when the fluctuation begins, measures the reaction latency up to this reaction start time, and records the reaction latency in the recording unit, and a reaction amount detection unit that identifies the first peak value of the sweat amount after the reaction start time and records this peak value as the reaction amount in the recording unit.
[6] The device for evaluating hazard recognition and/or hazard prediction according to [5], wherein the video is a video of a roadway including a situation requiring caution when driving a vehicle.
[7] The risk perception and/or risk prediction evaluation device according to [6], further comprising a judgment unit that judges the driving characteristics of the subject based on the correlation between the time change in the sweat rate and the time change in the cerebral blood flow, and the reaction latency and/or the reaction amount.
[8] A risk perception and/or risk prediction evaluation device as described in [7], which has a gaze detection means for measuring the gaze of the subject, and the judgment unit judges the driving characteristics of the subject based on the correlation between the time change in the amount of sweating and the time change in the cerebral blood flow, the reaction latency and/or the reaction amount, and gaze measurement data by the gaze detection means.
[9] The device for evaluating hazard perception and/or hazard prediction according to any one of [6] to [8], wherein the operating device is a brake pedal, an accelerator pedal, and/or a steering wheel.

The present invention provides a device for evaluating hazard recognition and/or hazard prediction that can objectively evaluate whether a subject's hazard recognition and/or hazard prediction is appropriate for a situation requiring caution.

1 is a diagram illustrating the configuration of an evaluation device for risk perception and/or risk prediction according to an embodiment. 1 is a diagram illustrating the configuration of a sweat rate sensor that constitutes an evaluation device for risk perception and/or risk prediction according to an embodiment. FIG. FIG. 10 is a diagram illustrating the configuration of a risk perception and/or risk prediction evaluation device according to another embodiment. FIG. 10 is a diagram showing an example of display of a determination result.

The following describes an embodiment in which the present invention is applied to an automobile driving simulator, but the "hazard recognition and/or hazard prediction evaluation device" of the present invention is not limited to the embodiment described below, and can also be applied to operation simulators for vehicles other than automobiles, such as bullet trains, trains, and airplanes, work simulators for dangerous work such as working at heights, and evacuation simulators in the event of a disaster such as an earthquake.

[Embodiment 1]
This embodiment is an evaluation device for hazard recognition and/or hazard prediction when driving a car, and as shown in FIG. 1, comprises an image presentation device 1 that presents an image including a situation requiring caution to a subject, a sweat rate sensor 2 that is attached to the palms and/or soles of the subject's feet and detects the amount of sweat, a cerebral blood flow sensor 3 that detects the cerebral blood flow of the subject, and a correlation evaluator 4 that determines the appropriateness of hazard recognition and/or hazard prediction for the situation requiring caution based on the correlation between the time change in the sweat rate and the time change in the cerebral blood flow.
In addition to the above configuration, a storage device 5 for storing a standard pattern of the correlation may be provided, and the judgment may be made by comparing with the standard pattern.

(Image presentation device 1)
In this embodiment, the image display device 1 displays on the display an image of a roadway including a situation requiring caution when driving a vehicle. Specifically, the image displays a situation requiring caution in the middle of an image that does not require much attention when driving a vehicle (for example, an image of a vehicle traveling on a straight road without traffic lights). In addition to the above-mentioned situations, various situations requiring caution include a situation where a stop sign appears, a situation where a parked vehicle is being overtaken, or a situation where a person, bicycle, or the like is about to jump out or has already jumped out from a side street. If necessary, ambient sounds can also be output from the speaker while driving.

(Sweat rate sensor 2)
The perspiration rate sensor 2 is not particularly limited as long as it has the function of measuring the perspiration rate on the palms and/or soles of the subject.
An example of the sweat rate sensor 2 is a capsule-type sweat rate sensor including an air supply means, a housing capsule whose opening is in close contact with the skin surface of the subject and which mixes the sweat dispersed from the skin surface with the air supplied from the air supply means in its internal space, an exhaust passageway which exhausts the mixed gas formed by mixing the sweat and the air supplied from the air supply means in the internal space of the housing capsule to the outside of the housing capsule, and a measuring means which measures the amount of sweat on the skin surface based on the difference between the humidity of the air supplied from the air supply means to the housing capsule and the humidity of the mixed gas exhausted from the exhaust passageway.
As shown in Fig. 2, the capsule-type perspiration rate sensor 2 includes an air blower 113 as the air supply means. The capsule-type perspiration rate sensor 2 further includes a housing capsule 103 having an opening 102 that is attached to the skin surface SK, an intake hole 104 for drawing natural air into the housing capsule 103, a mixing chamber 105 that communicates with the opening 102 to dissipate sweat on the skin surface SK and mixes the dispersed sweat with the natural air to form mixed air, an exhaust hole 106 for discharging the mixed air from the mixing chamber 105, a first humidity sensor 107 for measuring the humidity of the natural air, and a second humidity sensor 108 for measuring the temperature and humidity of the mixed air. The capsule-type perspiration rate sensor 2 further includes an air blower 113 that discharges natural air as the air supply means, upstream of the first humidity sensor 107, and an exhaust passage (flexible pipe) 117 downstream of the second humidity sensor 108. Furthermore, a calculation device 19 is used as a measurement means.

<Blower section 113>
When the blower 113 is located outside the casing capsule 103, for example, a compressor can be used as the blower 113. The compressor draws in natural air and then blows out the natural air. The natural air blown out from the compressor is supplied to the mixing chamber 105 of the casing capsule 103 via a flexible pipe 114. When the blower 113 is located inside the casing capsule 103, it is preferable that the blower 113 be arranged upstream of the mixing chamber 105. The blower 113 arranged upstream of the mixing chamber 105 draws in natural air through the intake hole 104 and blows it into the mixing chamber 105. For example, an electric air fan, air pump, or air blower can be used as the blower 113.

<Housing capsule 103>
The shape of the housing capsule 103 is not particularly limited, but for example, it can be formed in a substantially tubular shape (for example, a substantially cylindrical shape) with a closed upper end.
The material of the housing capsule 103 is not particularly limited, but it can be made of, for example, synthetic resin.

The air intake hole 104 for drawing natural air into the interior of the housing capsule has a smaller diameter (smaller size) than the inner diameter of the cylindrical housing capsule 103.

The housing capsule 103 has an opening 102 at its lower end (on the skin surface SK side) that is attached to (attached to) the skin surface SK.
The opening area of the opening 102 is not limited, but can be, for example, 1 cm ^2. When the unit of the amount of sweating on the skin surface SK is expressed in mg/cm ² min, if the opening area of the opening 102 is 1 cm ² , the measured value of the amount of sweating can be used as is without converting it by dividing it by the area.
The casing capsule 103 can be attached to the skin surface SK by applying, for example, double-sided tape, adhesive, or pressure-sensitive adhesive to the periphery of the opening 102 .

The housing capsule 103 has a mixing chamber 105 therein that communicates with the opening 102 .
The mixing chamber 105 functions as a space for dissipating sweat from the skin surface SK while in contact with the skin surface SK and for mixing the dispersed sweat with natural air.

The exhaust hole 106 for exhausting the mixed air from the mixing chamber 105 has a smaller diameter (smaller size) than the inner diameter of the cylindrical housing capsule 103.

<First Humidity Sensor 107>
The first humidity sensor 107 has a function of measuring the humidity of the natural air blown by the blower 113 .
2, a first humidity sensor 107 is disposed inside a box 115 located outside the housing capsule 103. In this embodiment, the first humidity sensor 107 measures the absolute humidity of natural air sent out from a compressor, which is the air blower 113, and supplied into the box 115.
Instead of the first humidity sensor 107 that measures absolute humidity, a relative humidity sensor and a temperature sensor may be disposed inside the box 115 to determine the absolute humidity of the natural air that has flowed into the box 115. A temperature sensor and a humidity sensor formed integrally may also be used.

<Second Humidity Sensor 108>
The second humidity sensor 108 has a function of measuring the humidity of the mixed air.
2, a second humidity sensor 108 is disposed inside a box 116 that is outside the housing capsule 103. In this embodiment, the second humidity sensor 108 measures the absolute humidity of the mixed air exhausted from the housing capsule 103 and supplied to the inside of the box 116.
Instead of the second humidity sensor 108 that measures absolute humidity, a relative humidity sensor and a temperature sensor may be disposed inside the box 116 to determine the absolute humidity of the natural air that has flowed into the box 116. A temperature sensor and a humidity sensor formed integrally may also be used.

<Circuit configuration>
2, the first humidity sensor 107 and the second humidity sensor 108 are electrically connected to filter circuits F1 and F2, respectively. When the detection signals output from the sensors are input to the filter circuits F1 and F2, the filter circuits F1 and F2 remove noise components contained in the detection signals and then amplify the detection signals by a predetermined amplification factor.
A differential amplifier DA1 is connected to the output sides of the filter circuits F1 and F2.
The output side of the filter circuit F1 is connected to the inverting input terminal (-) of the differential amplifier DA1.
The output side of the filter circuit F2 is connected to the non-inverting input terminal (+) of the differential amplifier DA1.
With this configuration, the differential amplifier DA1 outputs a signal that is the difference between a signal corresponding to the absolute humidity of natural air detected by the first humidity sensor 107 and a signal corresponding to the absolute humidity of the mixed air detected by the second humidity sensor 108. The signal output from the differential amplifier DA1 corresponds to the amount of sweat dissipated from the skin surface SK into the mixing chamber 105.
When the signal output from the differential amplifier DA1 is input to the calculation device 19, the calculation device 19 calculates the actual amount of sweat dissipated from the skin surface SK into the mixing chamber 105 based on each input signal. The amount of sweat obtained by the calculation can be displayed on the output device 20.

(Cerebral blood flow sensor 3)
The cerebral blood flow sensor 3 is not particularly limited as long as it has the function of detecting changes in blood flow in the brain of the subject. The cerebral blood flow sensor may be a non-contact sensor equipped with a light source that irradiates a predetermined light onto the surface of the head and a light-receiving sensor that receives the light.

(Correlation Evaluator 4)
The inventors have found that the state of danger recognition and/or danger prediction of a subject during the period can be evaluated from the correlation between sweat rate data, which is a plot of changes in sweat rate from a baseline time to the most recent 100 seconds, and cerebral blood flow data, which is a plot of changes in cerebral blood flow from the baseline time to the most recent 100 seconds. Specifically, the inventors have found that the evaluation can be performed as follows.
If the correlation is a "positive correlation" and both the amount of sweating and cerebral blood flow decrease for approximately the last 100 seconds from the baseline (hereinafter also referred to as a "Type A correlation"), it can be evaluated that "the subject predicted danger and became increasingly tense" during that time.
If the correlation is a "negative correlation" and the amount of sweating decreases and cerebral blood flow increases for approximately 100 seconds immediately preceding the baseline (hereinafter also referred to as a "Type B correlation"), it can be evaluated that "the subject paid sufficient attention during that time, but did not recognize any imminent danger."
If the correlation is a "negative correlation" and the amount of sweating increases and cerebral blood flow decreases for approximately 100 seconds immediately preceding the baseline time (hereinafter also referred to as a "Type C correlation"), it can be evaluated that "the subject recognized a sudden danger" during that time.
The present invention is based on this finding, and the correlation evaluator 4 determines the appropriateness of the subject's danger recognition and/or danger prediction for the situation requiring caution based on the correlation between the time change in the amount of sweat detected by the sweat rate sensor 2 and the time change in cerebral blood flow detected by the cerebral blood flow sensor 3.

(Storage device 5)
The above-mentioned judgment can also be made by comparing with a standard pattern stored in the storage device 5.
For example, if the standard pattern for a specific video is a "Type A correlation" and the subject's data also shows a "Type A correlation," it can be determined that the subject's danger perception and/or danger prediction was appropriate. By adding the element of time to the determination and taking into account the timing of changes in sweat rate, a more accurate determination can be made.
Furthermore, it is also possible to judge multiple situations requiring caution by converting the rate of match with a standard pattern into a score.

The standard pattern can be derived from a data set of multiple subjects accumulated as past test results.
Video scenes in which the standard pattern shows a "Type A correlation" are video scenes that require sustained attention while predicting danger, such as video scenes of a bicycle overtaking or a straight road with good visibility.
Video scenes in which the standard pattern shows a "Type B correlation" are scenes in which danger is predicted, but the danger is not imminent, and the sense of tension is reduced. For example, scenes in which the driver drives straight through an urban area or stops to wait for an oncoming vehicle can be exemplified.
Video scenes in which the standard pattern shows a "Type C correlation" are difficult to predict and are video scenes that evoke tension or surprise, such as turning right in an urban area or seeing a runner approaching from the front in a residential area.
Standard patterns are prepared for different age groups and levels of proficiency, and by using standard patterns with the same attributes as the test subject, more accurate judgments can be made.
Using a standard pattern derived from the subject's own past data, the present invention can also be applied to assessing improvement in cognitive function, for example, in rehabilitation for stroke or the like.

[Embodiment 2]
As shown in FIG. 3, in addition to the configuration of embodiment 1, a skin potential sensor 6 attached to the palm and/or sole of the foot to detect the skin potential reflex, an operation amount sensor 7 to detect the operation amount of an operation device operated by the subject, and a recording unit 8 to record the respective detection values detected by the sweat rate sensor, the cerebral blood flow sensor, the skin potential sensor, and the operation amount sensor may also be provided.
In addition to the above configuration, the device may also be provided with a reaction latency measuring unit 9 that determines the time from the time the video presentation device displays the video of the caution-requiring situation when the skin potential reflex detected by the skin potential sensor first fluctuates beyond a predetermined value, identifies the reaction start time when the fluctuation begins, measures the reaction latency up to this reaction start time, and records the reaction latency in the recording unit, and a reaction amount detecting unit 10 that identifies the first peak value that the sweat amount reaches after the reaction start time, and records this peak value as the reaction amount in the recording unit.
In addition to the above configuration, a judgment unit 11 may be provided that judges the driving characteristics of the subject based on at least one of the correlation between the time change in the amount of sweating and the time change in the cerebral blood flow, the reaction latency, and the reaction amount.
In addition to the above configuration, a gaze detection means 12 for measuring the gaze of the subject may also be provided.

(Skin potential sensor 6)
The skin potential sensor 6 is not particularly limited as long as it has the function of detecting skin potential reflexes of the palms and/or soles of the subject. For example, an electrodermal activity sensor having the function of measuring skin activity can be used.

(Operation amount sensor 7)
The operation amount sensor 7 is not particularly limited as long as it has a function of detecting the operation amount of the operation device 71 operated by the subject. For example, an operation amount sensor that detects the rotation amount of the steering wheel on the shaft of the steering wheel can be exemplified.
The operation device 71 to be detected by the operation amount sensor 7 may be a brake pedal, an accelerator pedal, or the like in addition to the steering wheel.

(Recording unit 8)
The detected values of the sweat rate sensor, the cerebral blood flow sensor, the skin potential sensor, and the operation amount sensor are recorded.

(Response latency measurement unit 9)
The reaction latency measurement unit 9 monitors whether the absolute value of the skin potential reflex detected by the skin potential sensor 6 fluctuates beyond a predetermined voltage value, such as 0.1 mV, from the point when the caution-requiring situation start time has elapsed. When the reaction latency measurement unit determines that this predetermined voltage value has been exceeded, it identifies the elapsed time in the video at which the skin potential reflex value already recorded in the recording unit 8 begins to fluctuate, that is, at which the skin potential reflex value begins to increase slightly (for example, 0.02 mV) from 0, and sets this as the reaction start time. The reaction latency measurement unit 9 records the time interval from the display of the caution-requiring situation to the reaction start time as the reaction latency of the skin potential reflex in the recording unit 8 in such a way that it is possible to identify which caution-requiring situation the reaction latency belongs to.
When the reaction latency exceeds a predetermined maximum allowable time, it can be determined that the driving characteristics for recognizing and predicting danger are low. The maximum allowable time is determined in advance and recorded in a recording unit. The maximum allowable time may be determined by having a large number of healthy drivers drive in a simulated manner and recording the detected values, and the maximum detection time for a situation requiring caution may be used. The maximum allowable time may also be changed according to gender or age.

(Reaction amount detection unit 10)
The reaction amount detection unit 10 records the first peak value as the reaction amount from the detected values of the amount of sweat recorded after the reaction started in the recording unit 8 so that it can be identified as the reaction amount for which caution is required.
When the reaction amount is smaller than a predetermined minimum reaction amount, it can be determined that the driving characteristics for recognizing and predicting danger are low. The minimum reaction amount is determined in advance and recorded in a recording unit. As the minimum reaction amount, a large number of healthy drivers may be made to perform simulated driving, the detected values are recorded, and the minimum reaction amount among those to situations requiring caution may be used. Furthermore, the minimum reaction amount may be changed according to gender or age.

(Determination unit 11)
The determination unit 11 comprehensively determines the driving characteristics of the subject based on the determination result by the correlation evaluator 4, the reaction latency and/or the reaction amount. In addition to the determination result by the correlation evaluator 4 and the reaction latency and/or the reaction amount, the determination unit 11 may combine the detection value of the operation amount sensor 7 and/or the gaze measurement data by the gaze detection means 12 to determine the driving characteristics from three viewpoints, namely, danger prediction, danger recognition, and avoidance behavior, as shown in FIG.
The determination unit 11 may make a determination based on one characteristic situation requiring attention in the video, or may make a determination based on multiple situations requiring attention in the video. When the determination unit 11 makes a determination based on multiple situations requiring attention, the determination unit 11 may further determine whether the driving characteristics are good or bad depending on the number of times the driving characteristics are determined to be poor.

The reaction latency measurement unit, reaction amount detection unit, and judgment unit may be activated while the simulated driving is being performed, or after the simulated driving has ended.

REFERENCE SIGNS LIST 1 Image display device 2 Sweat rate sensor 3 Cerebral blood flow sensor 4 Correlation evaluator 5 Storage device 6 Skin potential sensor 7 Operation amount sensor 8s Recording unit 9 Response latency measuring unit 10 Response amount detecting unit 11 Determination unit 12 Gaze detection means 19 Arithmetic unit 20 Output device

102 Opening 103 Housing capsule 104 Intake hole 105 Mixing chamber 106 Exhaust hole 107 First humidity sensor 108 Second humidity sensor 113 Blower 114 Flexible pipe 115 Box 116 Box 117 Flexible pipe

Claims (7)

an image display device that displays an image including a situation requiring attention to a subject;
a sweat rate sensor attached to the palms and/or soles of the subject to detect the amount of sweat;
a cerebral blood flow sensor for detecting cerebral blood flow of the subject;
A risk recognition and/or risk prediction evaluation device including a correlation evaluator that determines the appropriateness of risk recognition and/or risk prediction for the caution-requiring situation based on a correlation between the time change of the sweat rate and the time change of the cerebral blood flow,
the perspiration rate sensor comprises an air supply means, a capsule housing having an opening in close contact with the skin surface of the subject, which mixes the perspiration dispersed from the skin surface with air supplied from the air supply means in its internal space, an exhaust passageway for exhausting the mixed gas obtained by mixing the perspiration with the air supplied from the air supply means in the internal space of the capsule to the outside of the capsule, and a measuring means for measuring the perspiration rate on the skin surface based on the difference in humidity between the air supplied from the air supply means to the capsule housing and the mixed gas exhausted from the exhaust passageway;
The correlation is a correlation between sweat rate data obtained by plotting changes in sweat rate from a baseline time to a predetermined time immediately preceding the baseline time and cerebral blood flow data obtained by plotting changes in cerebral blood flow from a baseline time to a predetermined time immediately preceding the baseline time,
The correlation evaluator evaluates that, if both the amount of sweating and cerebral blood flow decrease during a predetermined period from the reference time, "the subject predicted danger and felt a rising sense of tension" during that period; if the amount of sweating decreases and the cerebral blood flow increases during a predetermined period immediately preceding the reference time, "the subject paid sufficient attention but did not recognize the imminent danger" during that period; and if the amount of sweating increases and the cerebral blood flow decreases during a predetermined period from the reference time, "the subject recognized a sudden danger" during that period; this is an evaluation device for danger recognition and/or danger prediction. a storage device that stores the standard pattern of the correlation;
The device for evaluating danger perception and/or danger prediction according to claim 1 , wherein the judgment is made by comparing with the standard pattern. a skin potential sensor attached to the palm and/or sole of the foot to detect the skin potential reflex;
an operation amount sensor that detects an operation amount of an operation device operated by the subject;
3. The risk perception and/or risk prediction evaluation device according to claim 1, further comprising a recording unit that records each of the detection values detected by the sweat rate sensor, the cerebral blood flow sensor, the skin potential sensor, and the operation amount sensor. a reaction latency measuring unit that determines the time when the skin potential reflex detected by the skin potential sensor first fluctuates beyond a predetermined value from the time when the video presentation device displays the video of the caution-requiring situation, identifies the reaction start time when the fluctuation begins, measures the reaction latency up to this reaction start time, and records the reaction latency in the recording unit;
The danger perception and/or danger prediction evaluation device of claim 3 further comprises a reaction amount detection unit that identifies the first peak value of the sweat amount after the reaction start point and records this peak value in the recording unit as the reaction amount. The risk perception and/or risk prediction evaluation device of claim 4, wherein the video is video of a roadway including a situation requiring caution when driving a vehicle. The risk perception and/or risk prediction evaluation device of claim 5, further comprising a judgment unit that judges the subject's driving characteristics based on the correlation between the time change in sweat rate and the time change in cerebral blood flow, and the reaction latency and/or the reaction amount. The device for evaluating hazard perception and/or hazard prediction described in claim 5 or 6, wherein the operating device is a brake pedal, an accelerator pedal, and/or a steering wheel.