JP2014154004A - Danger information processing method, device and system, and program - Google Patents

Abstract

PROBLEM TO BE SOLVED: To acquire even danger information that is not actualized.SOLUTION: During driving a vehicle, an image is photographed with a camera 21 of a tablet terminal 11 at a constant interval and is recorded. Also, operation information and behavior information obtained by detecting a driving operation and a vehicle behavior during driving are recorded. A danger factor which can be a factor that causes danger of an accident or the like is acquired from each image and the operation information and the behavior information. A danger type to be statistically alarmed is determined on the basis of each acquired danger factor, and danger information is registered in map data.

Description

The present invention relates to a risk information processing method, apparatus and system, and a program for driving a vehicle.

  There is known a system for preventing an accident in advance by notifying a driver of a point where an accident is likely to occur when the vehicle is traveling. In such a system, a point where the possibility of an accident or the like is high, that is, a point with a high degree of risk is identified from the detection result of the behavior of the vehicle and the driving operation together with the position of the vehicle, and the identified point is map data. Are stored in association with each other.

  For example, in the driving guidance device of Patent Document 1, an acceleration sensor and a GPS sensor that detects a position are mounted on a vehicle, a specific behavior and a position where the sensor is detected are detected by these sensors, and a specific position where the specific behavior frequently occurs is detected. Analyzing. And when approaching the specific position where a specific behavior occurs frequently during driving | running | working of a vehicle, it will alert | report.

  In addition, the danger point recording device described in Patent Document 2 records the target and surrounding situation when a dangerous situation occurs, and the alarm control device records the target and surrounding situation when the dangerous situation occurs. The alarm corresponding to the is output. In this danger point recording device, based on the output of the acceleration sensor, it is determined that the host vehicle is in a dangerous situation when it is determined that there has been a sudden deceleration or a collision with another vehicle. Yes. In addition, a vehicle front image is captured by an in-vehicle imaging device, and patterns such as preceding vehicles, crossing vehicles, oncoming vehicles, pedestrians, and bicycles are recognized and recognized from images when it is determined that the vehicle is in a dangerous situation. The object type is recorded in the database as the target when the dangerous situation occurs and the surrounding situation.

  In the latent danger point detection device described in Patent Document 3, the occurrence of a dangerous situation is determined as in Patent Document 3. The cause of the occurrence of a dangerous situation such as an image of the front of the vehicle taken by the camera immediately before the occurrence of the dangerous situation is analyzed, and obstacles such as bicycles and people jump out to the front of the own vehicle by the image analysis. The warning is made according to the cause of the occurrence.

  On the other hand, the vehicle travel support device of Patent Document 4 searches for a travel route that avoids a collision with another vehicle. In this vehicle travel support device, the degree of danger is obtained from the possibility of collision between the host vehicle and the obstacle based on the course of the host vehicle and the predicted course of the obstacle. The predicted course of the obstacle is predicted based on the position and moving speed of the detected obstacle by detecting the obstacle around the host vehicle by the obstacle sensor.

JP 2007-163472 A JP 2010-97345 A JP 2009-104531 A Patent No. 4900076

  Incidentally, all of Patent Documents 1 to 3 detect that the traveling vehicle is in a dangerous state, and register the vehicle position at that time. That is, even if an accident does not occur, it is determined that the degree of danger is high when an operation for avoiding an accident such as a sudden handle or a sudden brake is performed, and the point is registered. For this reason, there has been a problem that it is impossible to warn the driver of an unexposed danger.

  The present invention has been made to solve the above-mentioned problems, and is a risk information processing that statistically processes many images collected during driving of a vehicle, registers even latent dangers, and provides them. A method, an apparatus and a system, and a program are provided.

  In order to solve the above problem, the risk information processing method of the present invention includes a camera that captures an image around a vehicle, a position sensor that acquires position information, an image captured by the camera, and position information at the time of capturing acquired by the position sensor. A first collection step of running a plurality of vehicles equipped with a collection terminal device that collects the data in association with each other, and collecting images taken during driving of each vehicle in association with position information at the time of photography; A second collection step of collecting each image collected by the collection terminal device together with position information from each collection terminal device, and performing image recognition on each image collected in the second collection step, and associating with each image Based on the risk factor extraction step for extracting the risk factor of the road section including the point indicated in the position information and each risk factor in the same road section, a risk that should be warned in driving the vehicle in the road section And statistically determination step of classification, and information registration step of registering as a danger information of the road section corresponding risk classification alarming determined in decision step, and has a.

  The determination step determines, for each image, the likelihood of occurrence of a risk of each risk type set in advance as an individual risk based on one risk factor or a combination of a plurality of risk factors obtained from the image. It is preferable to obtain a statistical risk level from each individual risk level for the same risk type in a road section, and to determine a risk type with a statistical risk level equal to or higher than a certain value as a risk type to be warned in the road section.

  The first collection step collects the time information at the time of photographing together with the position information in association with each image, and the second collection step collects the image, the position information, and the time information collected in the first collection step. In the determination step, it is preferable to determine a risk type to be warned for each time zone based on each risk factor extracted from images taken in the same time zone within the same road section.

  The first collection step further collects the operation information and the vehicle behavior detected in the driving operation, and the second collection step further collects the operation information and the behavior information collected in the first collection step. In the extraction step, risk factors are extracted from each of the collected images and the operation information and behavior information, and in the determination step, each risk factor is extracted based on each risk factor extracted from each of the corresponding images, operation information, and behavior information. It is preferable to determine the danger type to be warned of the road section.

  Further, based on the input step for inputting the planned travel route and the planned travel time zone of the vehicle and the planned travel route and the planned travel time zone input at the input step, the risk information of each road section registered by the registration step is displayed. An acquisition step for acquiring danger information in the scheduled travel time zone of each road section on the planned travel route from within, and an output step for outputting the danger information acquired in the acquisition step or warning information corresponding thereto Is preferred.

  In the output step, it is preferable to display a map image including the planned travel route, and display the acquired risk information or the corresponding warning information in association with the corresponding road section on the displayed map image. .

  Moreover, it is preferable that an output step displays a map image and each danger information, or warning information corresponding to this on the display apparatus mounted in the vehicle.

  The danger information processing apparatus of the present invention collects, from a plurality of vehicles, images obtained by photographing the periphery of the vehicle while driving the vehicle and positional information at the time of photographing associated with each image by a camera mounted on the dangerous vehicle. The same as the risk factor extraction unit that performs image recognition on each image collected by the collection unit and extracts the risk factor of the road section including the point indicated by the position information associated with the image Based on each risk factor in the road section, a determination section that statistically determines a risk type to be warned in driving the vehicle in the road section, and a road section corresponding to the risk type to be warned determined by the determination section And an information registration unit for registering as dangerous information.

  For each image, the determination unit obtains the risk of occurrence of each preset risk type as an individual risk based on one risk factor or a combination of a plurality of risk factors obtained from the image, and the same. It is preferable to obtain a statistical risk level from each individual risk level for the same risk type in a road section, and to determine a risk type with a statistical risk level equal to or higher than a certain value as a risk type to be warned in the road section.

  The collection unit collects time information at the time of shooting associated with the image together with the position information, and the determination unit extracts each risk factor extracted from the images shot in the same time zone within the same road section. Based on the above, it is preferable to determine the risk type to be warned for each time zone.

  In addition, the collection unit further collects operation information for detecting driving operations and behavior information for detecting vehicle behavior, and a risk factor extraction unit extracts risk factors from each of the collected images, operation information, and behavior information. The determination unit preferably determines the risk type to be warned for each road section based on each risk factor extracted from the corresponding image, the operation information, and the behavior information.

  A risk information processing system of the present invention includes the above-described risk information processing device and an information providing device, and the information providing device is input at an input unit that inputs a planned travel route and a planned travel time zone of the vehicle, and an input step. Based on the planned travel route and the planned travel time zone, risk information in the planned travel time zone of each road section on the planned travel route is acquired from the risk information of each road section registered by the risk information processing apparatus. An acquisition unit and an output unit that outputs danger information acquired by the acquisition unit or warning information corresponding thereto are provided.

  The output unit may be a display that displays a map image including a planned travel route and displays each acquired risk information or corresponding warning information in association with a corresponding road section on the displayed map image. preferable.

  Moreover, it is preferable that a display is a thing of the apparatus mounted in a vehicle.

  The program of the present invention is a collection step of collecting, from a plurality of vehicles, images obtained by photographing the periphery of the vehicle while the vehicle is being driven, and positional information at the time of photographing associated with each image by a camera mounted on the vehicle, A risk factor extraction step that performs image recognition on each image collected in the collection step and extracts a risk factor of a road segment including the point indicated by the position information associated with the image, and within the same road segment A determination step for statistically determining a risk type to be warned in driving the vehicle in the road section based on each risk factor, and risk information of the road section corresponding to the risk type to be warned determined in the determination step The information registration step to be registered as is executed by the computer.

  According to the present invention, since the risk type to be warned is determined based on various risk factors extracted from a plurality of images taken during driving of the vehicle, it is possible to acquire even a latent risk. it can.

It is a block diagram which shows the structure of a danger information system It is a block diagram which shows the structure of a tablet terminal. It is a block diagram which shows the structure of a danger information processing apparatus. It is explanatory drawing which shows the example of the danger information of the road area registered into the map database. It is a block diagram which shows the structure of a driving | running analysis apparatus. It is explanatory drawing which shows the example of the driving | running tendency for every driver ID. It is explanatory drawing which shows the example of the correspondence of a driving tendency and danger classification. It is explanatory drawing which shows the example which displayed danger information on the map image on a display in normal display mode. It is explanatory drawing which shows the example of the hard copy which output danger information in normal display mode. It is explanatory drawing which shows the example which displayed danger information on the map image on a display in narrowing-down display mode. It is explanatory drawing which shows the example of the hard copy which output danger information in narrowing-down display mode. It is explanatory drawing which shows the example which highlighted the danger information according to the driving tendency with respect to other danger information, and displayed it on the map image on a display. It is a block diagram which shows the structure of the tablet terminal in the example which displays danger information on the display of a car navigation system, and a car navigation system. It is explanatory drawing which shows the example of a display on the display of the car navigation system in the example of FIG.

  In FIG. 1, the risk information processing system 10 acquires risk factors that can be factors that cause danger such as accidents from a plurality of images recorded in association with shooting positions (position information) while the vehicle is running, This system provides a driver with risk information and warning information by determining a risk type to be statistically warned based on each acquired risk factor. In this embodiment, the case of an automobile will be described as an example of the vehicle. However, the vehicle is not particularly limited, and may be a motorcycle, a bicycle, or the like in addition to the automobile.

  The risk information processing system 10 includes a tablet terminal 11, a risk information processing device 12, a map DB (database) 14, and a driving analysis device 15.

  Each tablet terminal 11 is mounted on a vehicle and becomes a collection terminal device. The tablet terminal 11 collects information-added images in which position information, time information, operation information, behavior information, and a driver ID for each driver are added to an image captured by a built-in camera during driving of the vehicle, and the collected information addition The image is sent to the danger information processing apparatus 12. The tablet terminal 11 functions as an information providing device. That is, the display 27a (see FIG. 2) constituting the touch panel 11a of the tablet terminal 11 includes the touch panel 11a on which the map image and the danger information acquired from the map DB 14 are displayed in the information display mode. The tablet terminal 11 is connected to an external sensor unit 19 that detects a driver's operation on the vehicle, and operation information is input from the external sensor unit 19.

  The tablet terminal 11, the danger information processing device 12, the map DB 14, and the driving analysis device 15 are connected via a network 18, and exchange various data with each other via this network 18. The network 18 may be the Internet, a dedicated line, a mobile phone line, a public wireless LAN, or a combination thereof.

  In this embodiment, one tablet terminal is used as the collecting terminal device and the information providing device, but the collecting terminal device and the information providing device are not limited to this. For example, in addition to a tablet terminal, a portable terminal device such as a smart phone (multifunctional telephone) or a notebook PC can be employed. Further, the car navigation device may have functions of a collection terminal device and an information providing device. Furthermore, a dedicated device that functions as a collection terminal device and an information providing device may be used.

  The collection terminal device and the information providing device may be separate devices. The collection terminal device may be any device as long as it is mounted on a vehicle and can be collected in association with various information such as images and position information during traveling. Further, the camera and the sensor whose details will be described later may not be incorporated in the collection terminal device.

  The information providing device may be any device that can output the acquired map image and danger information. For example, it may be a PC installed at the driver's home. In view of the danger information being confirmed by the driver during driving of the vehicle, the information providing device is particularly preferably capable of mounting the output unit on the vehicle.

  Furthermore, the method of sending the information-added image to the risk information processing device 12 and the method of acquiring the map image and the risk information are not limited to communication as described above. For example, information addition images, map images, and danger information may be exchanged between devices using a storage medium such as a memory card.

  The risk information processing apparatus 12 receives the information added image from each tablet terminal 11. The risk information processing device 12 acquires a risk factor from each image and each operation / behavior information in each information-added image, and warns for each road section and each time zone by statistical processing based on the acquired risk factor. The risk type to be determined is determined. Examples of danger types include “contact with pedestrians”, “contact with bicycles”, “jumping out of people”, “collision with forward vehicle”, “collision with oncoming vehicle”, and “contact with parallel vehicle”. , “Encounter collision”, “slip”, etc.

  The map DB 14 stores map data for displaying a map image and road information data. The road information data includes position information of each road on which the automobile on the map image travels and connection information of each road. In the road information data, each road including the intersection is divided into one or a plurality of road sections. The road information data includes information indicating the area of each road section, and the risk information for each road section. Is stored. The danger information is a danger type determined by the danger information processing apparatus 12 to warn. Each road section is identified by a section ID.

  The distance or section position of the road section can be set as appropriate, but it is preferable to set the same road section as a range where the traffic state of the road and the surrounding environment can be regarded as the same. In addition, since the intersection and the connecting portion of the road to the intersection are different in risk generally occurring in other road portions and the degree of fear of the occurrence, it is preferable to make them different road sections. Alternatively, road sections may be set for each traveling direction, or road sections may be set for intersections and branch roads, for example, straight ahead, right turn, and left turn.

  The driving analysis device 15 acquires operation information / behavior information through the danger information processing device 12, and analyzes the driving tendency of the driver based on the acquired operation information / behavior information. The driving analysis device 15 accumulates the analyzed driving tendency for each driver in a database.

  In FIG. 2, each part of the tablet terminal 11 is controlled by the control unit 20. For example, a dedicated application is installed in the tablet terminal 11, and the functions of each unit are realized by executing the dedicated application. The tablet terminal 11 includes a camera 21, a clock circuit 22, a built-in sensor unit 23, an interface circuit (I / F) 24, a built-in memory 25, a communication unit 26, a screen display unit 27, a touch sensor 28, and a route search unit 29. ing. An external sensor unit 19 is connected to the interface circuit 24.

  The camera 21 has its photographing lens exposed on the back surface opposite to the touch panel 11a. The tablet terminal 11 is disposed in the vehicle in a posture with the touch panel 11a facing the driver and the back facing forward. Thereby, the camera 21 images the front through the windshield of the vehicle. The camera 21 shoots images (still images and moving images) at a constant shooting interval under the control of the control unit 20 during driving of the vehicle. The shooting interval can be determined as appropriate. The clock circuit 22 measures the current time and outputs time information.

  In addition to the front of the vehicle, the rear and sides may be photographed, and a plurality of directions may be photographed. In addition, while the vehicle in which the image is being taken is being driven, the vehicle may be running or stopped, but the image may be taken only while the vehicle is running. Moreover, based on the detection result of the external sensor part 19 and the built-in sensor part 23, you may image | photograph, when the impact with respect to a vehicle is detected, when a sudden handle | steering-wheel or a sudden brake operation is detected. You may further image | photograph at the timing which detected the impact with respect to a vehicle, a sudden handle, a sudden brake operation, etc., performing imaging | photography at a fixed space | interval.

  The built-in sensor unit 23 includes a position sensor 23a, an acceleration sensor 23b, and an angular acceleration sensor 23c. The position sensor 23a is configured by a GPS (global positioning system) receiver or the like, detects the position of the vehicle on which the tablet terminal 11 is mounted, and outputs position information. It is preferable to perform position detection reliably and with high accuracy by using a positioning system that performs position detection based on the acceleration sensor 23b and travel distance information. The acceleration sensor 23b and the angular acceleration sensor 23c detect the acceleration and angular velocity of the vehicle, respectively. The control unit 20 detects the traveling speed of the vehicle based on the acceleration obtained from the acceleration sensor 23b. The behavior information indicating the behavior of the vehicle includes the detected traveling speed, acceleration, and angular velocity.

  The external sensor unit 19 detects operation timings and operation amounts of various operations of the driver and outputs them as operation information. The external sensor unit 19 includes a handle sensor 19a, an accelerator sensor 19b, and a brake sensor 19c, and detects a handle operation, a brake operation, and an accelerator operation.

  The sensors are not limited to the types of sensors described above, and various sensors that acquire risk factors to be extracted and information necessary for the driving tendency to be analyzed can be used. For example, a sensor for detecting the operation of the direction indicator, the operation of turning on / off the light, etc. may be provided, and the operation timing of the direction indicator, the timing of turning on / off the light, etc. may be acquired as part of the operation information.

  Each time the camera 21 captures an image, the control unit 20 generates an information-added image in which position information and time information at the time of shooting, operation information / behavior information, and a driver ID are added to the image obtained by shooting. And recorded in the built-in memory 25. Thereby, an image, position information, time information, operation information, and behavior information are associated with each other. The built-in memory 25 stores a plurality of information-added images. The control unit 20 collectively sends the accumulated information images to the risk information processing apparatus 12 via the communication unit 26. In this example, the information-added image is transmitted when the tablet terminal 11 is instructed to transmit. However, the information-added image is transmitted when a certain number of information images are accumulated, or communication with the risk information processing apparatus 12 is possible. Each information image stored for each time may be transmitted.

  The driver ID is input to the tablet terminal 11 by the driver prior to the start of operation, for example. For the operation information / behavior information, for example, information for the same time as the shooting interval is added to the image centering on the time of shooting. Thereby, it is possible to know the various operations of the driver and the behavior of the vehicle in the period before and after the image is taken. Note that all operation information / behavior information during traveling may be recorded as continuous information in association with time information, and may be associated with each image by the risk information processing device 12.

  The communication unit 26 communicates with the danger information processing device 12, the map DB 14, and the driving analysis device 15. The screen display unit 27 as an output unit includes a display 27a and a drive circuit 27b that drives the display 27a. The screen display unit 27 displays the map and the danger information acquired from the map DB 14 acquired by the control unit 20 on the display 27a. Various operation screens are displayed on the display 27a.

  A transparent touch sensor 28 is disposed on the surface of the display 27a. Thereby, the touch panel 11a which inputs by touching the surface according to the operation screen etc. which are displayed on the display 27a is comprised. For operation of the touch panel 11a, input of driver ID, instruction to start collecting information images, instruction to transmit stored information images, switching between normal display mode and narrowed display mode under information display mode, and route search Various inputs can be made.

  The route search unit 29 searches for a planned travel route for going from the current location to the destination and inputs the route to the control unit 20. Further, a route search unit 29, for example, a scheduled travel time zone in each road section on the planned travel route based on the departure time is obtained and input to the control unit 20. The current location, destination, and departure time are input by operating the touch panel 11a.

  The normal display mode of the information display mode is a mode for displaying the danger information of the scheduled travel time period for each road section of the planned travel route. The narrowed display mode is a mode for narrowing down and displaying the danger information output in the normal display mode to the danger information corresponding to the driving tendency for each driver. In any display mode, a map format and a list format can be selected as a display format.

  The road map information for route search may be stored in the tablet terminal 11 and the planned travel route may be searched based on this information. However, the route search site published on the Internet may be accessed for driving. The scheduled route and the scheduled traveling time zone of each road section may be acquired. Further, instead of searching for the planned travel route in this way, the planned travel route may be input by tracing a road on a map displayed on the touch panel 11a, for example.

  The control unit 20 is an acquisition unit that acquires danger information in the information display mode. In the normal display mode, the control unit 20 acquires from the map DB 14 map data in a range including the planned travel route and danger information registered in the same time zone as the planned travel time zone for each road section of the planned travel route. To do. Further, in the narrowed-down display mode, the control unit 20 acquires map data and danger information as in the normal display mode, but acquires only danger information according to the driving tendency of the driver. As the driving tendency of the driver, the driving tendency corresponding to the driver ID is acquired from the driving tendency DB 42. The memory 20a stores in advance correspondence data in which danger information indicating the type of danger that is likely to occur for each driving tendency is associated with each driving tendency, and the danger information corresponding to the driving tendency is specified by referring to this data. Is done.

  When the map format is selected, the control unit 20 drives the screen display unit 27 based on the acquired map data and the danger information, and displays a map image including the planned travel route and the danger information on the display 27a. . The danger information is displayed as a symbol or the like on or near the corresponding road section on the displayed map image. Note that warning information such as a warning message corresponding to the danger information may be displayed. When the list format is selected, the control unit 20 generates a list obtained by dividing each acquired danger information for each time zone, and displays the list on the display 27a. Similar to each display on the display 27a, a map image with danger information and danger information for each time zone may be printed by a printer.

  In FIG. 3, the risk information processing apparatus 12 includes a collection unit 31, a risk factor extraction unit 32, a specification unit 33, a determination unit 34, a risk DB 35, and an information registration unit 36. The risk information processing apparatus 12 realizes the functions of each unit by executing a risk information processing program on a computer, for example.

  The collection unit 31 collects information-added images from the tablet terminal 11 and includes an HDD 31a and a communication circuit 31b. The collection unit 31 communicates with each tablet terminal 11 via the communication circuit 31 b and receives an information-added image from the tablet terminal 11. Each information added image received is stored in the HDD 31a. As the HDD 31a, a large-capacity HDD capable of storing each information added image from the plurality of tablet terminals 11 is used.

  The collection unit 31 reads information-added images one by one from the HDD 31a and separates them into operation information, behavior information, time information, position information, driver ID, and images. Of these, the image, the operation information, and the behavior information are sent to the risk factor extraction unit 32. Further, the operation information and behavior information are sent to the driving analysis device 15 together with the driver ID. Further, the time information and position information are sent to the specifying unit 33.

  The risk factor extraction unit 32 includes an image analyzer 32a and an information analyzer 32b. The image analyzer 32a performs image recognition processing on the image, analyzes a subject in the image, and extracts risk factors. The information analyzer 32b analyzes the operation information and the behavior information, and extracts risk factors for the driver's operation and the vehicle behavior. Each analyzer 32a, 32b sends a risk factor score of a score corresponding to the presence / absence of each risk factor and the content (extent etc.) of the extracted risk factor to the determination unit 34.

  Risk factors extracted from the image include, for example, those related to the road itself such as “lane classification”, “road width”, “puddle”, “number of pedestrians”, “number of bicycles”, “inter-vehicle distance”, “road `` Pedestrians crossing '', `` Oncoming vehicles protruding from the vehicle lane '', `` Number of oncoming vehicles '', `` Vehicles crossing the front of the vehicle '', `` Distance to pedestrians '', `` Almost front Asahi / West ”, Etc. related to the surrounding situation. Risk factors extracted from operation information and behavior information include “rapid steering operation”, “rapid acceleration”, “rapid braking”, “traveling speed”, and the like. The risk factors to be extracted are set in advance.

  Each analyzer 32a, 32b gives a score between 0 and 1.0 to the risk factor score. “Lane classification”, “Pedestrians crossing the road”, “Oncoming vehicles that protrude from the vehicle lane”, “Rearly front of Asahi / West”, “Vehicles that cross the front of the vehicle”, “Sudden steering wheel operation” The risk factor score is “1.0” when the risk factor is extracted, such as “”, “rapid acceleration”, “rapid brake”, etc., and “0” when the risk factor is not extracted. Is done. For example, when there is a pedestrian crossing a road in the image, the risk factor score of “pedestrian crossing the road” is “1.0”. When such a pedestrian is not in the image, “ 0 ”.

  In addition, if the degree of risk factors such as “road width”, “puddle”, “number of pedestrians”, “number of bicycles”, “distance between vehicles”, “distance to pedestrians”, “traveling speed” are examined, The value is closer to “1.0” as the risk increases. For example, “road width” is such that the risk factor score is closer to “1.0” as the road width is narrower, and the risk factor score is “1” as the number of “pedestrians”, “bicycles”, etc. increases. .0 ". The “traveling speed” is a score closer to “1.0” as the speed is higher, but depending on the excess speed based on the speed limit that can be obtained from the signs in the image and the display on the road. You may score.

  The road width can be determined from the road width in the image and the known photographing magnification. The number of pedestrians and bicycles can be detected by counting the number of detected pedestrians by image recognition. The inter-vehicle distance is based on the fact that the license plate size of the previous vehicle is constant, and the distance to the pedestrian is considered to be the same as the size in the image, assuming that the pedestrian's face size is constant. Detection can be performed based on the magnification.

  The specifying unit 33 checks a point on the road information data indicated by the position information, specifies a road section including the point, and outputs the section ID. Further, a time zone including the time indicated in the time information is specified. For example, the time zone is divided into 24 time zones every hour from 0:00. The identified section ID and time zone of the road section are sent to the determination unit 34 and the information registration unit 36, and are used as information indicating the road section and time zone that are to be processed.

  The determination unit 34 determines the risk type to be warned by statistical processing based on the risk factor score. The determination unit 34 includes an individual risk calculator 34a, a statistical risk calculator 34b, and a comparator 34c, and a risk DB 35 is connected to the statistical risk calculator 34b.

  The individual risk calculator 34a obtains an individual risk level for each risk type based on one image and one or more risk factor scores obtained from the corresponding operation information and behavior information, This is sent to the statistical risk calculator 34b. The individual risk level indicates the likelihood of occurrence of the risk indicated by the corresponding risk type.

  In the individual risk calculator 34a, a calculation formula for calculating the individual risk for each risk type using the risk factor score is set. The calculation formula is created in accordance with the condition for generating the danger indicated by the danger type. For example, if there is a risk of an accident of “contact with pedestrians” when “there are many pedestrians, the traveling speed is high, and the distance to the pedestrians is short”, the “number of pedestrians” ”,“ Traveling speed ”, and“ distance to pedestrian ”are f1 to f3, respectively, and the individual risk Ga of“ contact with pedestrian ”is“ Ga = k1f1 · f3 ”.・ F3 ”(k1 is an appropriate coefficient). The same calculation is made for “contact with the bicycle”.

  In addition, for example, there is a risk of an accident of “collision with an oncoming vehicle” occurring when “the road is narrow and there are many oncoming vehicles, or the oncoming vehicle is running outside the host vehicle lane”. Assuming that there is a risk factor score of “road width”, “number of oncoming vehicles”, and “oncoming vehicles that protrude from the host vehicle lane” as f4 to f6, respectively, "Gb = (k4.f4.k5f5 + k6f6" (k4 to k6 are appropriate coefficients).

  In addition, for example, a “collision with a vehicle ahead” has a high traveling speed and a short inter-vehicle distance, and a “meeting collision” has a high traveling speed and there is a vehicle crossing the front. When there is a puddle and the traveling speed is high, it can be calculated that these are likely to occur. Further, assuming that a road section where the speed limit is unintentionally exceeded, a speed limit violation may be set as the risk type. In this case, on the basis of the speed limit extracted from the signs and signs in image analysis, the excess of the running speed obtained from the behavior information is output as a risk factor score, and the individual risk level can be obtained from the risk factor score. That's fine.

  The statistical risk calculator 34b is, for each risk type, based on the individual risk levels of the same road section, the same time zone, and the same risk type, the likelihood of occurrence of the risk statistically indicated by the risk type. The statistical risk level indicating In this example, an average value of each individual risk, for example, an arithmetic average is used as the statistical risk. Each individual risk used for calculating the statistical risk is the current individual risk input from the individual risk calculator 34a and each individual risk in the past. Each past individual risk is acquired from the risk DB 35 using the section ID of the road section specified by the specifying unit 33, the time zone, and the risk type corresponding to the statistical risk to be calculated. That is, the statistical risk level of one risk type is obtained statistically based on each risk factor obtained from each image of the same road section, the same time zone, and each behavior operation information. The statistical risk calculator 34b sends the calculated statistical risk to the comparator 34c. The statistical risk calculator 34b registers the current individual risk level used for calculating the statistical risk level in the risk level DB 35.

  The risk DB 35 stores individual risk levels obtained so far for each road section on the road information data in the map DB 14. Each individual risk level in the risk level DB 35 is given a corresponding risk type, time zone, and section ID, and by specifying them, each individual risk level with the same road section, time zone, and risk type can be obtained. Can be read.

  The comparator 34c checks whether or not it is a risk type to be warned in driving the vehicle by comparing the statistical risk level of the input risk type with a predetermined threshold value. In this comparison, when the statistical risk is equal to or higher than the threshold value, the comparator 34c determines that the risk type should be warned, and sends a registration instruction for this risk type to the information registration unit 36. On the other hand, if the statistical risk is less than the threshold value, it is determined that the risk type is not a warning risk type, and an instruction to delete this risk type is sent to the information registration unit 36.

  Note that the method for determining whether or not the risk type should be warned is not limited to the above.

  When the information registration unit 36 receives a risk type registration instruction from the comparator 34c, the information registration unit 36 uses the risk type as road information and road information specified by the specification unit 33 as road information in the map DB 14. Register to data. In addition, when a risk type deletion instruction is received, if the risk type is already registered in the road information data of the map DB 14 as the risk information in the road section and time zone specified by the specifying unit 33, Delete the danger information.

  As shown in FIG. 4, each road section of the road information data is registered with danger information that the road section has. FIG. 4 shows that the risk type corresponding to the column with “◯” is registered as the risk information for that time zone. In the road section of the section ID “56ff381de” shown in the figure, for example, “contact with a bicycle” and “popping out of a person” are registered as danger information in the time zone from 9:00 to 10:00, and the time from 11:00 to 12:00 “Collision with an oncoming vehicle” is registered as danger information in the time zone and from 17:00 to 18:00.

  In FIG. 5, the driving | running analysis apparatus 15 has the analysis part 41 and driving | operation tendency DB42. The analysis unit 41 analyzes the driving tendency of the driver, for example, “the frequency of sudden braking is high”. The driving tendency is analyzed for a plurality of analysis items. The analysis unit 41 analyzes the latest driving tendency based on the input operation information / behavior information and the results of the analysis items analyzed so far. The driving tendency obtained by the analysis is stored in the driving tendency DB 42 in association with the driver ID.

  In FIG. 6, the driving tendency DB 42 stores trend data in which driving tendencies of a plurality of drivers are associated with each driver. The trend data is composed of records for each driver ID, and each analysis item of driving tendency is written in each record. In FIG. 6, the recorded driving tendency items are indicated by “◯”. The driving tendency includes, for example, “slow brake operation”, “high frequency of sudden braking”, “high frequency of sudden steering”, “high frequency of sudden start”, and the like.

  The analysis items of the driving tendency are not limited to those exemplified above. Various analysis items can be set such as those that are dangerous in themselves, those that lead to danger, and driving habits that do not directly lead to danger. The driving tendency may be analyzed by combining each information included in the operation information and the behavior information. For example, based on the vehicle speed or acceleration information and the steering operation information, “the speed on the curve is high”, the steering wheel Based on the information on the operation and the brake operation, it is possible to analyze the driving tendency such as “operating the steering wheel while operating the brake”. Furthermore, for example, when a sensor for detecting the operation of the direction indicator is provided, the information obtained from the sensor, the steering wheel operation, and the vehicle speed information indicate that “the turn signal operation is slow when turning right or left”, “the turn signal is not emitted. You can also analyze the driving tendency such as “turn left and right”.

  In addition to the operation information / behavior information, for example, an image captured by the camera 21 may be used for analysis of driving tendency. The driving tendency can be analyzed from only the photographed image, and the driving tendency can be analyzed by combining the photographed image with the operation information and the behavior information. For example, it is possible to analyze “running to the right of the lane” from the position of the center line in the image. Also, by combining information such as the inter-vehicle distance, road width, road surface wetness, and paused intersections obtained from the images with information on vehicle speed and brake operation, “the inter-vehicle distance during high-speed driving is short”, “the road width Driving tendencies such as “high speed on narrow roads”, “high speed on wet roads”, “incomplete pause”.

  As shown in an example in FIG. 7, danger information to be acquired for each driving tendency is written in the memory 20 a as corresponding data. For example, danger information such as “a collision with a preceding vehicle” and “a person jumping out” is associated with “the frequency of sudden braking is high”. Further, danger information (target danger information) such as “contact with a parallel running vehicle” and “slip” is associated with “the frequency of sudden steering is high”.

  Next, the operation of the above configuration will be described. When driving the vehicle, the driver touches the touch panel 11a of the tablet terminal 11 to start the dedicated application and then inputs the driver ID assigned to the driver. Thereafter, the vehicle is driven. While driving the vehicle, the camera 21 captures the front of the vehicle at regular intervals. Each time this shooting is performed, the input driver ID, the position information from the position sensor 23a and the clock circuit 22, and the time information are acquired and added to the shot image.

  Further, various operations of the driver and detection of the behavior of the vehicle are continuously performed by the acceleration sensor 23b, the angular velocity sensor 23b, and the external sensor unit 19, and operation information and behavior information before and after photographing are added to the image. . As described above, each time an information-added image in which position information, time information, operation information, and behavior information are added to the image is generated, the information-added image is sequentially stored in the built-in memory 25.

  Accordingly, for example, if the vehicle travels on a road section with many pedestrians, an image including many pedestrians is captured, and if the vehicle travels close to the preceding vehicle, a state in which the inter-vehicle distance is short is captured. Also, behaviors such as the speed of the vehicle at the time of shooting, and operations performed by the driver such as brake operation and steering wheel operation are recorded.

  After the driving is finished, for example, the driver removes the tablet terminal 11 from the vehicle, moves to a place where communication with the danger information processing apparatus 12 is possible, and then touches the touch panel 11a to instruct transmission. As a result, the information addition images stored in the built-in memory 25 are sequentially read out by the control unit 20 and transmitted to the collection unit 31 via the communication unit 26. Each information added image transmitted from the tablet terminal 11 is received by the collection unit 31 and written to the HDD 31a.

  When the writing of the information addition image is completed, the first information addition image is read from the HDD 31a, and this information addition image is separated into an image, operation information / behavior information, time information, position information, and a driver ID. Then, the separated first image and operation information / behavior information are sent to the risk factor extracting unit 32, and time information and position information are sent to the specifying unit 33. Further, the operation information / behavior information and the driver ID are sent to the driving analysis device 15.

  When the time information and the position information are input, the specifying unit 33 specifies from which time zone the image output from the collection unit 31 and the operation information / behavior information belong. Further, by referring to the road information data in the map DB 14, a road section including the point indicated by the position information is specified, and the section ID is acquired. Then, the section ID and the time zone are sent to the statistical risk calculator 34 b and the information registration unit 36. Since the location information may be shifted due to positioning errors, if the location indicated by the location information is not on the road, the location of the road section must be specified after correcting the location to be on the road. Good to do.

  In the risk factor extraction unit 32, the first image is input to the image analyzer 32a. Image recognition processing is performed on the image by the image analyzer 32a, and each risk factor in the image is extracted. Then, the risk factor score corresponding to the presence or absence of the risk factor or the degree is output for each risk factor. In addition, operation behavior information is input to the information analyzer 32b, risk factors are extracted from the operation information / behavior information, and risk factor scores for each risk factor are output. Then, each risk factor score from each analyzer 32a, 32b is sent to the individual risk calculator 34a.

  When the risk factor score is input, the individual risk calculator 34a first extracts each risk factor score necessary for calculating the first risk type, for example, “contact with pedestrian”, Applies to the formula for “contact with pedestrians”. Thereby, the individual risk of the first risk type determined from the first image is obtained. Then, the individual risk level of the first risk type is sent to the statistical risk level calculator 34b.

  When the individual risk of the first risk type is input to the statistical risk calculator 34b, the statistical risk calculator 34b is registered in the road section and time zone specified by the specifying unit 33. Each individual risk level of the th risk type is read from the risk level DB 35. Therefore, each individual risk read at this time is calculated based on each image taken in the same time zone within the same road section as the first image, and operation information / behavior information added thereto. Is.

  The statistical risk calculator 34b obtains the average of the read individual risk and the individual risk from the individual risk calculator 34a, and sets this as the statistical risk of the first risk type. Then, this statistical risk is sent to the comparator 34c. Further, after obtaining the statistical risk in this way, the individual risk from the individual risk calculator 34 a is stored in the risk DB 14. In this storage, the individual risk is stored as the time zone of the road section specified by the specifying unit 33.

  The comparator 34c compares the statistical risk level of the first risk type with a threshold value. If the statistical risk is equal to or greater than the threshold, the comparator 34c sends the registration instruction to the information registration unit 36, assuming that the first risk type is the risk type to be warned. In this case, in response to this registration instruction, the information registration unit 36 registers the first risk type as risk information in the time zone of the road section specified by the road information data specifying unit 33 of the map DB 14. At this time, if the same danger information is already registered, it is overwritten, but nothing need be registered.

  On the other hand, when the statistical risk level is smaller than the threshold value, the comparator 34c sends a registration deletion instruction to the information registration unit 36, assuming that it is not a risk type to be warned. In this case, the information registration unit 36 checks whether or not the first risk type is registered as the risk information in the time zone of the road section specified by the road information data specifying unit 33. If so, delete the danger information. If it is not registered, the process is completed without doing anything.

  When the processing of the first risk type is completed as described above, the individual risk level is calculated for the second risk type, for example, “contact with a bicycle” by the individual risk level calculator 34a and the statistical risk level calculator 34b. The statistical risk is calculated sequentially. The calculation of each risk level is the same as that of the first risk type, but the calculation formula for the second risk type is used in calculating the individual risk level. In the calculation of the statistical risk, each individual risk of the second risk type registered in the road section and time zone specified by the specifying unit 33 is read from the risk DB 35 and calculated as these. The statistical risk is calculated as an average with each individual risk of the second risk type.

  Subsequently, the comparator 34c compares the statistical risk level of the second risk type with a threshold value. If the statistical risk level is equal to or higher than the threshold value, a registration instruction is sent to the information registration unit 36, and the threshold value is higher than the threshold value. When it is small, a delete instruction is sent. Then, based on the registration instruction or the deletion instruction, the risk information is registered or deleted in the time zone of the road section specified by the road information data specifying unit 33 of the map DB 14 in the same manner as described above.

  Thereafter, in the same manner, for the third and subsequent risk types, the calculation of the individual risk, the calculation of the statistical risk, and the comparison between the statistical risk and the threshold are sequentially performed. Then, the risk type whose statistical risk level is equal to or greater than the threshold is registered in the map DB 14 as risk information, and when it is less than the threshold, the risk information is deleted from the map DB 14.

  When the process for the last risk type is completed, the second information-added image is read out, and risk factors are extracted and risk factor scores are output in the same procedure as the first information-added image. Degree, statistical risk level, statistical risk level and threshold value, and risk information registration / deletion based on the comparison result. In this case, calculation of statistical risk and registration / deletion of risk information are performed for a time zone including a shooting time in a road section including a point where an image separated from the second information-added image is shot. Thereafter, processing is performed in the same procedure up to the last information added image.

  On the other hand, the operation information / behavior information separated from each image is sequentially input to the driving analysis device 15 together with the driver ID. Each input operation information / behavior information is analyzed by the driving analysis device 15 together with the past driving tendency, so that a new driving tendency is generated. Then, the generated driving tendency is stored in the driving tendency DB 42 in association with the driver ID.

  As described above, analysis of danger information using each image and each operation information / behavior information obtained by one driving of one driver and analysis of driving tendency are performed.

  In the same manner as described above, the driver ID is input to the tablet terminal 11 mounted on another vehicle, the vehicle is driven, information additional images are collected during driving, and each information additional image accumulated after the end of traveling is displayed. Transmit to the collection unit 31. As a result, each information-added image is collected for each travel of each vehicle, collected by the collection unit 31, and written in the HDD 31a. Thereafter, in the same procedure, based on the image, operation information / behavior information, time information, position information, and driver ID separated from the information addition image newly written in the HDD 31a, the risk factor extraction unit 32 and the determination unit 34 The risk type to be warned is determined and the driving tendency is analyzed by the driving analyzer 15.

  As each vehicle travels on various road sections in various time zones, the individual risk levels of the respective risk types in each road section are sequentially accumulated. Then, risk information statistically determined based on the accumulated individual risk is registered in the map DB 14. Therefore, the accuracy of the danger information registered in the map DB 14 is increased.

  When referring to danger information before driving, the surface of the touch panel 11a is touched to enter the information display mode, and either the normal display mode or the narrowed display mode is selected. For example, when the normal display mode is selected, a route search is performed by inputting a departure place, a destination, and a departure time.

  The route search unit 29 searches for a planned travel route based on the input departure point and destination. Further, the route search unit 29 obtains the scheduled traveling time zone of each road section of the scheduled traveling route based on the input departure time. After this, the control unit 20 accesses the map DB 14 via the communication unit 26 and is registered in the same time zone as the map data in the range including the planned travel route obtained by the route search unit 29 and the planned travel time zone. Each risk information about each road section on the planned travel route is acquired from the map DB 14.

  For example, a map format is set as an initial value of the display format. After each information is acquired, the control unit 20 drives the screen display unit 27 based on the acquired map data and the danger information, and displays a map image on the display 27a. Danger information is displayed. At this time, the driver displaying each danger information on or near the corresponding road section can view this as a screen of the touch panel 11a.

  For example, in the case where the scheduled traveling time is from 9:30 to 11:30, the scheduled traveling time zone includes 9am to 10am, 10am to 11am, and 11am to 12am. Therefore, each road section on the planned travel route has one of these three time zones as the planned travel time zone. “Contact with pedestrians” and “People jumps out” are registered at 10:00 in one road section with a scheduled travel time of 9:00 to 10:00 and another time section in another road section. It is assumed that “collision with the vehicle ahead” is registered in the same time zone of the road section at ˜11 o'clock, and that danger information is not registered in each road section between 11:00 and 12:00.

  In this case, “contact with a pedestrian”, “popping out a person”, and “collision with a preceding vehicle” are acquired, and a display screen 51 is displayed on the touch panel 11a as shown in FIG. Is done. On the displayed display screen 51, a map image 52 including a travel route, marks 53a and 53c, and a dotted line 53b are displayed as danger information. That is, a hatched mark 53a is superimposed and displayed on a road section with danger information of “contact with pedestrians”, and a road section with danger information of “popping out of people” is displayed along the road section. A dotted line 53b is displayed. Furthermore, a star-shaped mark 53c is displayed in the vicinity of the road section having the danger information of “collision with the preceding vehicle”. Here, it is the scheduled travel time zone of each road section where the mark 53a and the dotted line 53b are displayed corresponding to the danger information of “contact with pedestrian” and “popping out of the person”, and “ The scheduled travel time zone of the road section where the mark 53c is displayed corresponding to the danger information of “collision with the vehicle ahead” is from 10:00 to 11:00. This provides the driver with a danger that may occur on the planned travel route and prompts attention. Note that the symbol R1 indicates the planned travel route.

  When the display in the list format is instructed, the control unit 20 generates data in which the acquired risk information is classified into the list format for each time zone, and is sent to the screen display unit 27. As a result, as shown in FIG. 9, a display screen 54 is displayed on the touch panel 11a showing the danger information of each road section where the driver plans to travel for each time zone. On the display screen 54, “contact with a pedestrian” and “popping out a person” are displayed in the time zone from 9:00 to 10:00, and “collision with the vehicle ahead” is displayed in the time zone from 10:00 to 11:00. Yes. No danger information is displayed in the time zone from 11:00 to 12:00.

  On the other hand, when the narrowed-down display mode is selected, the route search is performed by inputting the driver ID and then inputting the departure place, the destination, and the departure time. As in the normal display mode, the planned travel route is searched based on the departure point, the destination, and the departure time, and the planned travel time zone of each road section of the planned travel route is obtained. Next, the control unit 20 reads the driving tendency corresponding to the input driver ID from the driving tendency DB 42, and refers to the memory 20a to identify the danger information corresponding to the driving tendency. Thereafter, the map DB 14 is accessed, and the specified risk among the map data in the range including the planned travel route and each risk information registered in the same time zone as the planned travel time zone for each road section on the planned travel route. Get information.

  For example, in the case of the same planned travel route and the planned travel time zone of each road section as in the normal display mode, as shown in FIG. 6, the driver ID is “AA001”, and the driving tendency of the driver Is assumed to be only “the brake operation is slow”. Further, as shown in FIG. 7, the danger information associated with “slow brake operation” includes “collision with the vehicle ahead” and “jumping out of person”, but “contact with pedestrian” Is not included. In this case, the control unit 20 acquires “collision with the vehicle ahead” and “popping out of a person” as danger information, but does not acquire “contact with a pedestrian”.

  Therefore, in the map format, as shown in FIG. 10, the map screen 55 displayed on the touch panel 11a does not display the mark corresponding to the danger information of “contact with pedestrians”, and the danger of “popping out of people”. A dotted line 53b is displayed on the road section with the information, and a star mark 53c is displayed on the road section with the danger information “collision with the preceding vehicle”. Further, in the case of the list format as shown in FIG. 11, “popping out a person” is displayed in the time zone from 9:00 to 10:00, and “collision with the preceding vehicle” is displayed in the time zone from 10:00 to 11:00. The display screen 56 is displayed on the touch panel 11a.

  Thus, in the narrowed-down display mode, only the danger information required by the driver is narrowed down and provided according to the driving tendency. In this example, the risk information corresponding to the driving tendency is narrowed down, but the risk information may be narrowed down under other conditions. Also, the planned travel route can be considered as one narrowing condition. Therefore, for example, it is possible to remove the narrowing condition of the planned travel route and display only the danger information corresponding to the driving tendency of the driver for each road section of the area displayed as the map image. Alternatively, an arbitrary time zone may be designated and only the danger information for the designated time zone may be displayed. One or a plurality of danger information may be arbitrarily designated, and only the designated danger information may be displayed. Of course, danger information to be displayed by combining a plurality of conditions may be narrowed down.

  In the above-described narrowed display mode, the danger information corresponding to the driving tendency of the driver is displayed, and other danger information is not displayed, so that their output modes are different. It is not limited. In the example of FIG. 12, the danger information corresponding to the driving tendency is displayed more emphasized than the other danger information. In FIG. 12, the danger information corresponding to the driving tendency is displayed as marks 53a and 53b, and the danger information mark 53c not corresponding to the driving tendency is displayed at a lighter density than the marks 53a and 53b. . Thereby, the marks 53a and 53b are emphasized with respect to the mark 53c.

  13 and 14 show an example in which danger information is displayed on the guidance display screen of the car navigation device. In addition, except being demonstrated below, it is the same as the said embodiment, The same code | symbol is attached | subjected to the same structural member, and the detailed description is abbreviate | omitted.

  The tablet terminal 11 includes an interface circuit 62 for writing / reading data to / from the memory card 61, and the memory card 61 is detachable. The control unit 20 acquires risk information from the map DB 14 along with the position information of the corresponding road section based on the planned travel route and the planned travel time zone of each road section, and the acquired risk information and their position information, The planned travel route is written to the memory card 61 attached to the tablet terminal 11 via the interface circuit 62. The memory card 61 is removed from the tablet terminal 11 after completion of the writing of the danger information and the position information, and is attached to the car navigation device 64 mounted on the vehicle.

  The car navigation device 64 displays the current position of the vehicle on a map image displayed as is well known and provides route guidance to the destination. The car navigation device 64 includes a storage unit 65 that stores map display data, a GPS receiver 66 that measures the current position of the host vehicle, a navigation control unit 67 that controls each unit, and a display 68. The navigation control unit 67 reads out map data around the current location from the storage unit 65 and displays a guidance display screen 69 on the display 68 as shown in FIG. A map image 70 around the current location is displayed on the guidance display screen 69 based on the read map data. Further, the host vehicle mark P is superimposed on the map image 70 at the position of the host vehicle based on the planned route R2 indicating the planned driving route read from the memory card 61 and the measurement result of the GPS receiver 66.

  When the planned route R2 is displayed, the position information of each road section written in the memory card 61 is referred to, and the danger information of each road section in the displayed map range is read from the memory card 61. It is read out and displayed on the corresponding road section on the guidance display screen 69 or in the vicinity thereof. In FIG. 14, danger information marks 71 of “contact with pedestrians” are displayed on the currently traveling road section and the road section ahead.

  Thus, by displaying the danger information on the displayed guidance display screen of the car navigation device 64, the danger information can be provided to the driver during traveling. Also in the display of this car navigation device, the display of the danger information specified according to the driving tendency and the other danger tables can be made different.

  In this embodiment, the information providing device is configured by the tablet terminal and the car navigation device, but the car navigation device may directly access the map DB and the driving tendency analysis device to acquire various information. In addition, as a car navigation device, for example, an application for car navigation may be installed in a tablet terminal serving as a collection terminal device or an information providing device.

  In each of the above embodiments, risk factors are extracted by analyzing collected images one by one. However, risk factors may be extracted by analyzing a plurality of collected images collectively. . In this way, for example, it is possible to extract and use risk factors such as the movement of other vehicles, that other vehicles are stopped, and that many vehicles are overtaking. A moving image may be taken instead of a still image. Further, although risk factors are extracted from each of the image and the operation behavior information, the risk factors may be extracted from only the image.

  Furthermore, when the danger information is output, the danger information of the same section may be output and displayed regardless of the time zone. In addition to the time zone, the month, season, day of the week, national holidays, specific days (for example, 5, 10, 15, 20, 25, etc.), AM / PM, An attribute such as weather may be assigned to the danger information, and the danger information may be narrowed down by the attribute and output / displayed.

  When giving a driver warning based on the registered danger information, in addition to displaying symbols and messages, light warning, warning sound and voice warning, smell warning, contact / vibration / heat warning, etc. Also good. Also in this case, in addition to not outputting them, changing the brightness and color of the light according to the danger information according to the driving tendency and others, outputting different warning sounds, outputting different odors, different contact / It is better to change the output mode such as generating vibration and heat.

DESCRIPTION OF SYMBOLS 10 Risk information processing system 11 Tablet terminal 12 Risk information processing apparatus 14 Map database 31 Collection part 32 Risk factor extraction part 34 Determination part

Claims (15)

A plurality of cameras equipped with a camera that captures images around the vehicle, a position sensor that acquires position information, and a collection terminal device that collects images captured by the camera and position information acquired by the position sensor in association with each other A first collection step of running a vehicle and collecting images taken during driving of each vehicle in association with positional information at the time of shooting;
A second collection step of collecting each image collected by the collection terminal device together with position information from each collection terminal device;
A risk factor extracting step of performing image recognition on each image collected in the second collecting step and extracting a risk factor of a road section including a point indicated by position information associated with the image;
A determination step for statistically determining a risk type to be warned in driving the vehicle in the road section based on each risk factor in the same road section;
An information registration step of registering the danger type to be warned determined in the determination step as the risk information of the corresponding road section;
A risk information processing method characterized by comprising:   In the determination step, for each image, based on one risk factor or a combination of a plurality of risk factors obtained from the image, the risk of occurrence of each risk type set in advance is determined as an individual risk level, A statistical risk level is obtained from each individual risk level for the same risk type on the same road section, and a risk type with a statistical risk level equal to or greater than a certain value is determined as a risk type to be warned for the road section. Item 1. The risk information processing method according to Item 1. The first collecting step collects time information at the time of photographing together with position information for each of the images,
The second collecting step collects the image, position information and time information collected in the first collecting step,
The determination step is to determine a risk type to be warned for each time zone based on each risk factor extracted from images taken in the same time zone within the same road section. The risk information processing method according to 1 or 2. The first collection step further collects the operation information and the behavior of the vehicle that detected the driving operation,
The second collection step further collects the operation information and behavior information collected in the first collection step,
The risk factor extraction step extracts a risk factor from each of the collected image and operation information and behavior information,
4. The determination step according to claim 1, wherein the risk type to be warned of each road section is determined based on each risk factor extracted from the corresponding image, operation information, and behavior information. The dangerous information processing method according to any one of the above items. An input step for inputting a planned travel route and a planned travel time zone of the vehicle;
Based on the planned travel route and the planned travel time zone input in the input step, the travel schedule of each road section on the planned travel route from the danger information of each road section registered by the registration step. An acquisition step for acquiring danger information in a time zone;
The risk information processing method according to any one of claims 1 to 4, further comprising: an output step of outputting the danger information acquired in the acquisition step or warning information corresponding thereto.   The output step displays a map image including the planned travel route, and displays each acquired risk information or warning information corresponding thereto in association with a corresponding road section on the displayed map image. 6. The risk information processing method according to claim 5, wherein   7. The danger information processing method according to claim 6, wherein the output step displays a map image and each danger information or warning information corresponding thereto on a display device mounted on the vehicle. An image obtained by photographing the periphery of the vehicle while driving the vehicle by a camera mounted on the vehicle, and a collection unit that collects position information at the time of photographing associated with each image from a plurality of vehicles,
Recognizing each image collected by the collection unit, a risk factor extraction unit that extracts a risk factor of a road section including a point indicated by position information associated with the image;
Based on each risk factor in the same road section, a determination unit that statistically determines a risk type to be warned in driving the vehicle in the road section,
An information registration unit that registers the risk type to be warned determined by the determination unit as the risk information of the corresponding road section;
A risk information processing apparatus comprising:   The determination unit, for each image, based on one risk factor or a combination of a plurality of risk factors obtained from the image, to determine the risk of occurrence of each risk type set in advance as an individual risk, A statistical risk level is obtained from each individual risk level for the same risk type on the same road section, and a risk type with a statistical risk level equal to or greater than a certain value is determined as a risk type to be warned for the road section. Item 9. The danger information processing apparatus according to Item 8. The collection unit collects time information at the time of photographing associated with an image together with position information,
The determination unit determines a risk type to be warned for each time zone based on each risk factor extracted from images taken in the same time zone within the same road section. The danger information processing apparatus according to 8 or 9. The collection unit further collects operation information for detecting driving operations and behavior information for detecting vehicle behavior,
The risk factor extraction unit extracts risk factors from each of the collected images and operation information and behavior information,
The said determination part determines the risk classification which should be warned of each road area based on each risk factor extracted from each of a corresponding image, operation information, and behavior information. The danger information processing apparatus according to any one of the preceding claims. A risk information processing device according to any one of claims 8 to 11 and an information providing device,
The information providing apparatus includes:
An input unit for inputting a planned travel route and a planned travel time zone of the vehicle;
Based on the planned travel route and the planned travel time zone input in the input step, the risk information of each road section on the planned travel route from among the risk information of each road section registered by the risk information processing device. An acquisition unit for acquiring danger information in the scheduled driving time zone;
An output unit that outputs the danger information acquired by the acquisition unit or warning information corresponding thereto;
A risk information processing system comprising:   The output unit is a display that displays a map image including the planned travel route, and displays each acquired risk information or warning information corresponding thereto in association with a corresponding road section on the displayed map image. 13. The danger information providing system according to claim 12, wherein the danger information providing system is provided.   The danger information processing system according to claim 13, wherein the display is of a device mounted on a vehicle. An image obtained by photographing the periphery of the vehicle during driving of the vehicle by a camera mounted on the vehicle, and a collection step of collecting position information at the time of photographing associated with each image from a plurality of vehicles;
A risk factor extraction step of performing image recognition on each image collected in the collecting step and extracting a risk factor of a road section including a point indicated by position information associated with the image;
A determination step for statistically determining a risk type to be warned in driving the vehicle in the road section based on each risk factor in the same road section;
An information registration step of registering the danger type to be warned determined in the determination step as the risk information of the corresponding road section;
A program that causes a computer to execute.

Images