JP2020201795A - Automobile driving capability decision system - Google Patents

Abstract

To provide an automobile driving capability decision system that compares information, which is obtained by a monitoring device mounted in a vehicle, with a database stored in artificial intelligence, and objectively decides on a degree of danger of a traveling vehicle.SOLUTION: An automobile driving capability decision system includes: a monitoring device 3 having a feature of monitoring at least a scene ahead of a vehicle 1; a degree-of-danger decision device 4 having a feature of deciding on a degree of driving danger concerning collision of the vehicle 1 with an obstacle; a degree-of-danger ranking device having a feature of ranking resultant information; and an integration device having a feature of integrating the ranking device. The degree-of-danger decision device 4 includes a database in which information on a result of driving by a normal driver and information on a result of traveling by an aged person or a person who should return a driving license are put together.SELECTED DRAWING: Figure 1

Description

The present invention relates to, for example, a vehicle driving function determination device attached to a vehicle in order to check the driving skill and cognitive function of an elderly driver.

In recent years, automobile accidents have been increasing due to delays in forward recognition of automobile drivers, especially elderly people, mistakes in accelerator and brake depression, and mistakes in selection of forward and reverse gears. Therefore, the return of the driver's license for the elderly is being discussed, but the current situation is that the driver's license is not returned. In addition, there are individual differences in the driving skills and cognitive functions of the elderly, and the current situation is that it is not possible to judge the quality of driving skills based on age alone.

Conventionally, as a driving support device, a device that enhances the reliability of the diagnosis result of the driver's driving behavior and improves the evaluation method of the diagnosis result has been known. Because the driving support device diagnoses the driving type of the driver from various angles based on the combination of the judgment results of a plurality of driving characteristics. Compared with the case where the driving type is diagnosed only by the judgment result of a specific driving characteristic, the reliability of the diagnosis result of the driving behavior of the driver can be improved, and the evaluation method of the diagnosis result can be improved (for example). , Patent Document 1).

Further, as a distance correction device of a monitoring system, a device that corrects an error including an error caused by a horizontal deviation of a stereo camera is known. The stereo arithmetic circuit in the stereo camera calculates the parallax by stereo matching based on a pair of captured images obtained from the stereo camera. The recognition unit calculates the distance to the object based on the parallax and the vanishing point parallax. The correction calculation unit calculates a plurality of spatially parallel approximate straight lines extending in the distance direction on one of the captured image planes, calculates the first vanishing point from the revolution of the approximate straight lines, and calculates the first vanishing point of the other. In the captured image plane, a plurality of parallel approximate straight lines extending in the distance direction are calculated, a second vanishing point is calculated from the intersection of the approximate straight lines, and the first vanishing point and the second vanishing point are The vanishing point diopter is corrected based on the amount of deviation (see, for example, Patent Document 2).

Further, as a collision avoidance system, an electronic range-finding scanner, a retroreflector installed in a target automobile, and a collision avoidance system suitable for application to an automobile including a processor unit are known. In the collision avoidance system, the ranging scanner provides the processor with data on the distance and angle to the target vehicle. The processor continuously monitors the position, speed and acceleration of each target vehicle, determines the distance, angle, speed acceleration of the target vehicle and the predicted interval when the target vehicle obstructs the course, and transmits it. To do. When the predicted interval at the time of course obstruction is smaller than a predetermined value, a warning signal or a command for taking a collision avoidance action is issued (see, for example, Patent Document 3).

JP-A-2017-191133 Japanese Unexamined Patent Publication No. 2003-83742 Japanese Unexamined Patent Publication No. 6-12598

As in the above cited document, many collision avoidance systems for avoiding vehicle collision accidents caused by driving a car have been proposed. In addition, in the past, driving skills were sometimes judged using a drive recorder, but the judgment itself is ambiguous because the judgment itself is made by humans. In addition, regarding the driving skill of a car, the decline of body, cognitive ability, judgment, etc. is quantified using artificial intelligence (AI) regardless of age, and the numerical value is used to judge fairly, that is, objectively. There is a question of how to do this.

Therefore, the applicant has quantified the driving skill of the elderly using artificial intelligence (AI), etc., and the quantified data is used in many conventional databases. We compared them and considered whether we could make an objective judgment on driving skills. In addition, when renewing a driver's license, a cognitive function test and a training course for the elderly are conducted, but the content is not sufficient, and the elderly who have had an accident also undergo these tests before renewing their driver's license. At present, it does not have sufficient functions and systems. Also, even if my family recommends that the elderly return their driver's license, the elderly often do not agree with me because I am okay. In addition, when elderly people develop early-stage dementia, it may be difficult to accurately determine their own driving skills.

An object of the present invention is to solve the above-mentioned problems, to measure the driving condition of a vehicle by using a monitoring device, and to compare the measured value with the artificial intelligence (AI) data having a large number of situations in a database. By doing so, the risk level of the traveling vehicle is objectively judged by quantifying the normal driver, the elderly person who has a solid driving function and cognitive function, and the person whose driving skill is about to be dangerous. It is to provide a vehicle driving function determination device to be attached to a vehicle, which is composed of enabling various evaluations.

The present invention is a monitoring device having a function of monitoring one of the front, left-right and rear directions of a vehicle, or all directions of the vehicle, and a function of determining a collision driving risk between the vehicle and an obstacle. The risk determination device has a risk determination device, a risk ranking device having a function of ranking information from the risk determination device, and an integration device having a function of integrating the ranking device. Is equipped with a database that aggregates driving result information of normal drivers, driving result information of elderly people, and driving result information of people who should return their driver's license, and monitors the driving skills of test drivers with the above-mentioned monitoring device. The present invention relates to a vehicle driving function determination device to be attached to a vehicle, which comprises quantifying and ranking the driving skills and cognitive functions of the test driver by comparing the information obtained in the above with the database.

Further, in this vehicle driving function determination device, the driving information from the risk level determination device is transmitted and stored in a predetermined server determined in advance through a predetermined communication base station in the traveling area of the vehicle, and the server transmits and stores the driving information. The result of comparing the received driving information by the vehicle with the database stored in the server is transmitted to the risk level determination device of the vehicle through the communication base station.

Further, the risk level determination device in this vehicle driving function determination device recognizes the traveling position of the vehicle by GPS, and obtains the driving information of the test driver and the proper driving information on the road map on which the vehicle is traveling. It has a function of comparing and determining whether or not the driving information of the test driver is appropriate.

Further, the monitoring device is composed of any one of a lidar, a millimeter wave radar, a stereo camera, and a monocular camera, or a combination thereof.

Further, the risk level determination device is implemented by logical reasoning AI and machine learning AI.

In addition, this automobile driving function determination device records a high-risk driving point with a camera by the risk-level ranking device.

As described above, the automobile driving function determination device attached to the vehicle according to the present invention uses LIDAR, millimeter-wave radar, stereo camera, and monocular camera as monitoring devices, so that it is possible to reliably grasp the momentary situation of the vehicle, that is, information. It is possible to judge the risk level of an objective traveling vehicle by comparing it with the AI database accumulated using such information, and quantify the results to obtain an objective evaluation. Become. The database is data on the past risk levels of ordinary young drivers, elderly people who have solid driving and cognitive functions, and people whose driving skills are about to be at risk. Based on the results obtained by this car driving function judgment device, as an example, it can be used as a material for judging whether to renew or return a driver's license at a family meeting including the elderly. By discussing with people who think "I am good at driving" based on objective data, people who are elderly but who are good at driving skills and who can drive well with cognitive function and those who are not It is possible to make a judgment.

It is the schematic explanatory drawing which shows the Example of the automobile driving function determination device attached to the vehicle by this invention.

Hereinafter, examples of the automobile driving function determination device attached to the vehicle according to the present invention will be described with reference to the drawings. The automobile driving function determination device attached to this vehicle is attached to the vehicle 1 and a test driver such as an elderly person drives the actual road for a certain period of time or longer to judge the skill and cognitive function. This vehicle driving function determination device includes a monitoring device 3 having any one or a combination of a front monitoring function of the vehicle 1, a rear monitoring function of the vehicle, and a left / right monitoring function of monitoring the left / right direction of the vehicle. There is. The monitoring device 3 includes a rider (LIDER), a millimeter-wave radar, a stereo camera, a monocular camera, and in some cases, a vehicle speed sensor. LIDER, millimeter-wave radar, and stereo cameras can measure obstacles in the direction of travel of the vehicle, their size, and the distance to the obstacles. In addition, the monocular camera can recognize objects and detect shooting records, and can shoot dangerous scenes.

The vehicle driving function determination device according to the present invention is a monitoring device 3 and a vehicle having a function of monitoring a vehicle 1 equipped with tires 2, that is, one of the front, left-right and rear directions of the vehicle, or all directions of the vehicle 1. Collision between 1 and an obstacle A risk judgment device having a function of judging the risk of operation 4, a data logger that saves information obtained by monitoring with the risk judgment device 4, and information from the risk judgment device 4. It has a risk leveling device having a function of ranking the above and an integrating device having a function of integrating the ranking device. The risk determination device 4 provides driving result information of a normal driver, that is, a driver having normal driving skill and cognitive function, driving result information of an elderly person, and driving result information of a person who should return a driver's license. It is equipped with an aggregated database and an operation information transmission device. Further, the risk level determination device 4 includes a communication module, an accelerometer, and the like. In particular, this automobile driving function judgment device stores the information obtained by monitoring the driving skill of the test driver with the monitoring device 1 in the data logger 5, and compares the data with the database of the risk judgment device 4. The feature is that the driving skill and cognitive function of the test driver can be quantified and ranked. Further, the data logger 5 is provided with a function of mechanically determining an event that is difficult for the human eye to discriminate and not easily digitizing, and intermittently storing information, and records the result. That is, the printer data can be transferred and printed out.

In this automobile driving function determination device, the monitoring device 3 is composed of any one of a lidar, a millimeter wave radar, a stereo camera, a monocular camera, and a vehicle speed sensor, or a combination thereof. The English name of the rider is LIDAR (Laser Imaging).
Detection and Ranging), which means light detection and ranging or laser image detection and ranging, remote sensing using light, which measures scattered light for pulsed laser irradiation and is at a long distance. It analyzes the distance to an object or the nature of the object. LIDAR uses electromagnetic waves with a wavelength much shorter than that of radar, for example, ultraviolet rays, visible rays, and rays close to infrared rays. Millimeter-wave radar is a radar technology that detects vehicle 1 and pedestrians using radio waves, and is a sensor that measures the reflected wave of the emitted radio waves and detects the position and speed of obstacles and objects. A stereo camera is a camera that can record information in the depth direction of an obstacle or an object by simultaneously photographing the object from a plurality of different directions. A monocular camera is a camera that detects automobiles, pedestrians, white lines on roads, etc. from captured images, and estimates the distance to the detected object.

Regarding this automobile driving function determination device, the driving information from the risk level determination device 4 is transmitted and stored in a predetermined server 7 determined in advance through a predetermined communication base station 6 in the traveling area of the vehicle 1, and the server 7 transmits and stores the driving information. , The result of comparing the received driving information by the vehicle 1 with the database stored in the server 7 is transmitted to the risk level determination device 4 of the vehicle 1 through the communication base station 6. That is, the risk level determination device 4 mounted on the vehicle 1 transmits the vehicle driving information of the test driver to a predetermined communication base station 6 in the traveling area of the vehicle 1. The communication base station 6 transmits the AI judgment result of the comparison result with the database to the risk level judgment device 4, and records the result in the data logger 5. That is, the communication base station 6 transmits the operation information of the risk level determination device 4 to the server 7, and the received server 7 uses the logical reasoning AI (artificial intelligence) and the machine learning AI (artificial intelligence). It will be judged whether the driving information of the test driver is appropriate.

Further, in this vehicle driving function determination device, the risk level determination device 4 recognizes the traveling position of the vehicle 1 by GPS (global positioning system), and displays the driving information of the test driver and the road map on which the vehicle 1 is traveling. It has a function to judge whether or not the driving information of the test driver is appropriate by comparing with the proper driving information. For example, if there is road map information in the road map, that is, determined driving information, for example, the road map information should have a one-way display or an entry prohibition display for the vehicle 1 currently traveling. Regardless, if the test driver overlooks and goes straight, the risk determination device 4 determines that the test driver has a cognitive function of NG.

This automobile driving function determination device can be configured in the following two types. In the first type, this car driving function judgment device can be configured to be always mounted on the vehicle 1, but in that case, the server cost is not required, but the in-vehicle device becomes large and expensive. .. In the second type, when this vehicle driving function determination device is attached to and detachably mounted on the vehicle 1, the server 7 communicates with the vehicle 1 through the external communication base station 6 to determine the risk level of the driving information of the vehicle 1. It will be completed. The monitoring device 3 is configured in the optimum system by selecting either or both types of the first type and the second type. For example, as the monitoring device 3, it is possible to select (1) millimeter-wave radar, GPS and map data, (2) LIDAR and monocular camera, or (3) monocular camera only. That is, the purpose of this automobile driving function determination device is achieved by the function of the monitoring device 3 (LIDAR, millimeter wave radar, monocular camera, stereo camera), GPS, and map data in cooperation with each other.

In this automobile driving function determination device, the risk level determination device 4 is particularly characterized in that it is executed by logical reasoning AI (artificial intelligence) and machine learning AI (artificial intelligence). In this embodiment, the logical reasoning AI refers to various events that occur during driving, and the AI accumulates these events and uses them as judgment information. For example, event A indicates that there is an obstacle in front of the vehicle 1. Event B means that the height of the obstacle in front of the vehicle 1 is 25 cm or more. When the monitoring device 3 detects the event A and the event B, the judgment, that is, the result is recorded as extremely dangerous. In this way, as an event, the event to occur is narrowed down by combining the condition of the distance to the obstacle in front of the vehicle 1 in the traveling direction, the condition of the width, the condition of time, and the like. Further, in this embodiment, the machine learning AI (artificial intelligence) is defined as A. Recording an image of a human standing. B. Recording images of humans walking. C. Recording images of human children. D. Recording an image of a human falling and crawling on his stomach. Here, if all of the above A, B, C, and D are human beings, the machine is made to learn, that is, to memorize them. It is judged as "human" by comparing the actual event with the machine-learned one. That is, this automobile driving function determination device determines the degree of risk while compensating for the shortcomings of both by combining the logical reasoning AI and the machine learning AI. For example, in logical reasoning AI, the risk is judged to be low because the height of the obstacle is 25 cm or less, but in machine learning AI, if it is detected as a hungry person from the image by the monitoring device, the risk is high because it is a human. It will be judged to be expensive.

In addition, the information from the monitoring device 1 is input to AI (artificial intelligence) in this car driving function judgment device. A device for making a judgment by comparing with the data possessed by AI is provided.
Further, this automobile driving function determining device has a risk level determining device 4 as a function of determining the driving risk level. The risk judgment device 4 shall divide the risk judgment of the vehicle into five stages.
1. The first stage is the case where the vehicle is staggering.
2. 2. The second stage is the case where the acceleration / deceleration of the vehicle is large.
3. 3. Make the third stage a little dangerous.
4. The fourth stage is a dangerous situation.
5. The fifth stage is extremely dangerous.

This automobile driving function determination device is characterized in that a high-risk driving point is recorded by the camera of the monitoring device 3 by a risk ranking device. In this embodiment, the monocular camera or the stereo camera constantly captures the image in front of the vehicle 1. This automobile driving function determination device receives the video information taken by the monitoring device 3 and constantly determines the degree of danger, and when the memory is full, the oldest information is automatically deleted. However, for example, if the risk level of the driving situation is classified into the first stage, the second stage, the third stage, the fourth stage, and the fifth stage, the image of "the risk level is the first stage or higher" is not deleted. Then, record in memory one by one. This allows the parties to later determine what is dangerous on the video. For example, in this car driving function judgment device, there is information that there are 5 images with a risk level of 1st stage, 2 images with a risk level of 2nd stage, and 1 image with a risk level of 5th stage. If this happens, do not discard this information, but store it in memory. In particular, this is to make the test driver recognize "images" that he / she does not notice. Because he is "confident in his driving", he says "I don't remember the risk I missed", so I need to save it in memory to cover those situations. Is.

In this automobile driving function determination device, the risk level determination device 4 has a function of objectively quantifying how much dangerous driving the elderly driver is causing and how often. .. The degree of risk is determined, for example, by dividing the frequency of the degree of risk into five stages as described above and determining how many times those stages of the degree of risk have occurred. For example, the driving skills of the test driver were 0 times in the 5th stage, 2 times in the 4th stage, 4 times in the 3rd stage, 10 times in the 2nd stage, and 8 times in the 1st stage. When the result is obtained, an objective judgment will be made based on this information. Furthermore, this automobile driving function judgment device can take a picture of a dangerous scene while driving and is used as information for judging the degree of danger.

This automobile driving function determination device has a function of outputting the driving result of the driver. This car driving function judgment device shall compare with a database of many measured values (including the results of normal drivers) and present the evaluation score and rank of the risk of the driver's results.

By presenting the evaluation score and rank of the risk of the result of the person, the elderly person and his / her family, for example, share this result and discuss whether to continue the driver's license or return it to help the judgment.

The vehicle driving skill determination device attached to the vehicle according to the present invention is preferably used for checking the driving skill and cognitive function of an elderly driver.
It is a thing.

1 Vehicle 2 Tire 3 Monitoring device 4 Danger level judgment device 5 Data logger 6 Communication base station 7 Server

Claims (6)

A monitoring device having a function of monitoring the front, left-right and rear directions of the vehicle, or all directions of the vehicle, a function of determining the risk of collision driving between the vehicle and an obstacle, and at least a primary stop. A risk judgment device having a function of judging a collision driving risk due to a decrease in cognitive function that ignores signals, a risk ranking device having a function of ranking driving information from the risk judgment device, and the above. It has an integrating device that has a function to integrate the ranking device,
The risk level judgment device is equipped with a database that aggregates driving result information of a normal driver, driving result information of an elderly person, and driving result information of a person who should return a driver's license.
An automobile attached to a vehicle consisting of quantifying and ranking the driving skill and cognitive function of the test driver by comparing the information obtained by monitoring the driving skill of the test driver with the monitoring device with the database. Driving function judgment device. The driving information from the risk determination device is transmitted and stored in a predetermined server determined in advance through a predetermined communication base station in the traveling area of the vehicle, and the server receives the driving information by the vehicle and the driving information. The vehicle driving function determination device according to claim 1, wherein the result of comparison with a database stored in the server is transmitted to the risk level determination device of the vehicle through the communication base station. The risk determination device recognizes the traveling position of the vehicle by GPS, compares the driving information of the test driver with the proper driving information on the road map on which the vehicle is traveling, and compares the driving information of the test driver with the proper driving information of the test driver. The automobile driving function determination device to be attached to the vehicle according to claim 1 or 2, which has a function of determining whether or not the driving information is appropriate. The automobile to be attached to the vehicle according to any one of claims 1 to 3, wherein the monitoring device is composed of any one of a rider, a millimeter wave radar, a stereo camera, and a monocular camera, or a combination thereof. Driving function judgment device. The vehicle driving function determination device according to any one of claims 1 to 4, wherein the risk determination device is implemented by logical reasoning artificial intelligence and machine learning artificial intelligence. The vehicle driving function determination device to be attached to a vehicle according to any one of claims 1 to 5, wherein a driving point having a high degree of risk is recorded by a camera by the risk ranking device.

Images