Operating state analysis method and operating state analysis program

Abstract

To provide a driving state analysis method capable of detecting the presence / absence of an operation accompanied with a risk factor while suppressing a data amount subjected to a determination.SOLUTION: The present invention relates to a vehicle driving state analysis method based on information detectable by an on-vehicle unit 10. The information detected by the on-vehicle unit 10 is fetched into storage means 22, the information stored in the storage means 22 is read out. Arithmetic means 24 detects vehicle start timing based on time information that timing in which a speed of a vehicle rises over a predetermined threshold is coincident with timing in which an acceleration of the vehicle exceeds a predetermined threshold and detects, as vehicle stop timing, timing in which the speed of the vehicle is below the predetermined threshold and the speed rise occurs as long as or longer than a predetermined time.SELECTED DRAWING: Figure 2

Description

Technical Field

[0001] The present invention relates to the analysis of the driving state of a vehicle, and particularly to a method and a program for analyzing the driving state of a vehicle based on information that can be obtained by a drive recorder.

Background Art

[0002] In recent years, it has become widespread to attach a drive recorder that records images during driving to vehicles such as automobiles. Therefore, from the perspective of safe driving, it has come to be carried out to investigate the driving state of a driver based on the images recorded via the drive recorder. Originally, in such an investigation based on images, by looking at the captured images, the presence or absence of dangerous driving operations during driving was determined. However, dangerous driving operations to be determined are often extremely few in the long-duration images, and the labor required for the judgment staff is large.

[0003] For this reason, attempts have been made to detect dangerous driving operations such as sudden acceleration and sudden braking based on the values of acceleration sensors mounted on vehicle-mounted devices such as drive recorders. However, changes in acceleration often occur not due to sudden acceleration or sudden braking, and are also easily affected by noise. Therefore, there has been a situation where there are many misjudgments.

[0004] Among such circumstances, in the technique disclosed in Patent Document 1, based on the speed data and video data recorded by a vehicle-mounted device such as a drive recorder, in the frame of the video data in the time zone when a speed change (mainly deceleration) occurs, the filling rate of the circumscribed rectangle of the region detected by the detection unit and the size of the corner region at the diagonal of the circumscribed rectangle are used to identify the detection target due to dangerous driving operations. However, in recent years when the high definition of recorded videos is advancing, the amount of data to be determined becomes enormous in the determination using video data, and it is considered that it takes time to identify the detection target.

[0005] On the other hand, Patent Document 2 discloses a technique for determining whether or not sudden braking has occurred based on a comparison between the longitudinal acceleration derivative obtained by the time derivative of the vehicle's longitudinal acceleration and the vertical acceleration derivative obtained by the time derivative of the vehicle's vertical acceleration. With such a technique, since the data used for determination is only acceleration data, the amount of data can be reduced compared to video data, even when the recording time is long.

[0006] However, the technology disclosed in Patent Document 2 cannot determine actions that involve other risk factors, such as when starting or turning left or right, or when switching between driving actions, and is therefore unlikely to contribute to the detection of general dangerous driving behaviors. [Prior art documents] [Patent Documents]

[0007] [Patent Document 1] Patent No. 5983749 [Patent Document 2] Patent No. 6406041 [Overview of the project] [Problems that the invention aims to solve]

[0008] Therefore, the present invention aims to provide a driving state analysis method and a driving state analysis program that can detect the presence or absence of operations involving risk factors while keeping the amount of data to be analyzed to a minimum. [Means for solving the problem]

[0009] The driving state analysis method according to the present invention for achieving the above objective is a method for analyzing the driving state of a vehicle based on information obtainable by an in-vehicle device, wherein the information obtained by the in-vehicle device is taken into a storage means of a computer, the information stored in the storage means is read out, the computer's calculation means determines in advance a minimum travel distance calculated based on the vehicle's position information, detects five detection points with the minimum travel distance as the unit of distance, defines the detection point located in the center of the five detection points as the reference detection point, the detection points located before and after the reference detection point as adjacent detection points, the detection point located on the opposite side of the reference detection point via the adjacent detection points as the starting point detection point and the ending point detection point, respectively, determines the local radius of curvature that draws a circle passing through the reference detection point and the two adjacent detection points, and the global radius of curvature that draws a circle passing through the reference detection point, the starting point detection point and the ending point detection point, and detects the time information of the detection of the reference detection point as the right or left turning timing of the vehicle when both the local radius of curvature and the global radius of curvature are below a predetermined threshold.

[0010] Furthermore, in the driving state analysis method having the above-described features, the calculation means may detect the vehicle's starting timing based on time information where the timing at which the vehicle's speed rises above a predetermined threshold coincides with the timing at which the vehicle's acceleration exceeds a predetermined threshold. Having such features makes it possible to detect the vehicle's starting operation.

[0011] Furthermore, in the driving state analysis method having the above-described features, the calculation means can also detect the timing at which the vehicle stops when the vehicle's speed falls below a predetermined threshold and no further speed increase occurs for a predetermined period of time or longer. Having such features makes it possible to detect the stopping action of the vehicle.

[0012] Furthermore, in the driving state analysis method having the above-described features, the calculation means can also detect, during the stopping timing, when acceleration in the opposite direction to the direction of travel occurs in the vehicle, and detect an increase in the vehicle's speed at the timing when this opposite acceleration rises, as the timing of the vehicle's reversal. With these features, it becomes possible to detect whether or not the vehicle is performing a reversal movement.

[0013] Furthermore, in a driving state analysis method having the above-described characteristics, the in-vehicle device can acquire the vehicle's reverse gear signal, and if it detects an increase in the vehicle's speed at the timing when the reverse gear signal rises, it can also detect this as the vehicle's reversing timing. Even with such characteristics, it is possible to detect whether or not the vehicle is performing a reversing maneuver.

[0014] Furthermore, in a driving state analysis method having the above-described features, the calculation means may also detect the time period during which the vehicle is traveling straight based on the vehicle's starting timing and the vehicle's turning timing, or the vehicle's stopping timing and the vehicle's turning timing, or the vehicle's starting timing and the vehicle's stopping timing. Having such features makes it possible to grasp the length and distance of the time the vehicle is traveling straight.

[0015] Furthermore, in a driving state analysis method having the above-described features, the in-vehicle device may acquire the vehicle speed pulse signal and incorporate it into the storage means as vehicle speed information. Having such features makes it possible to acquire vehicle speed information with high accuracy and improve the accuracy of detecting various timings.

[0016] Furthermore, the driving state analysis program according to the present invention for achieving the above objective is a program that is executed using vehicle position information acquired by an in-vehicle device and stored in a storage means in association with time information, and is characterized by executing the following: reading the position information stored in the storage means via a calculation means; predetermining a minimum travel distance calculated based on the position information, detecting five detection points with the minimum travel distance as the unit of distance, defining the detection point located in the center of the five detection points as the reference detection point, the detection points located before and after the reference detection point as adjacent detection points, the detection point located on the opposite side of the reference detection point via the adjacent detection points as the starting point detection point and the ending point detection point, respectively; determining the local radius of curvature of a circle passing through the reference detection point and the two adjacent detection points, and the global radius of curvature of a circle passing through the reference detection point, the starting point detection point and the ending point detection point; and, if both the local radius of curvature and the global radius of curvature are below a predetermined threshold, determining the time information at which the reference detection point was detected as the right or left turning timing of the vehicle.

[0017] Furthermore, in a driving state analysis program having the above-described features, the information acquired by the in-vehicle device includes speed information and acceleration information associated with the time information. When this information is acquired by the storage means along with the position information, the program may also perform the following processes: reading the speed information and acceleration information stored in the storage means via the calculation means; detecting the timing at which the vehicle's speed rises above a predetermined threshold based on the speed information, and detecting the timing at which the vehicle's acceleration exceeds a predetermined threshold based on the acceleration information, and recognizing the time information at which both coincide as the vehicle's starting timing. Having such features makes it possible to detect the vehicle's starting operation.

[0018] Furthermore, in a driving state analysis program having the above-described features, the program may also perform the following steps: reading the stored speed information via the calculation means, and determining the timing at which the vehicle's speed falls below a predetermined threshold and no further speed increase occurs for a predetermined period of time based on the speed information, as the vehicle's stopping timing. Having these features makes it possible to detect the vehicle's stopping action.

[0019] Furthermore, in a driving state analysis program having the above-described characteristics, the program may also detect the timing at which acceleration occurs in the vehicle in the opposite direction to the direction of travel during the stopping timing, and, based on the speed information, detect whether or not the vehicle's speed has increased, and execute a process to determine the time information at which both coincide as the timing of the vehicle's reverse movement. Having such characteristics makes it possible to detect whether or not the vehicle is performing a reverse movement.

[0020] Furthermore, in a driving state analysis program having the above-described characteristics, the information acquired by the in-vehicle device includes a reverse gear signal associated with the time information, and when this signal is acquired by the storage means along with the position information, if an increase in the vehicle's speed is detected at the timing when the reverse gear signal rises, the program may perform a process to detect this as the timing for the vehicle to reverse. Even with such characteristics, it is possible to detect whether or not the vehicle is performing a reverse maneuver.

[0021] Furthermore, a driving state analysis program having the above-described characteristics may also perform a process to detect the time period during which the vehicle is traveling straight, based on the vehicle's starting timing and turning timing, or the vehicle's stopping timing and turning timing. Having such characteristics makes it possible to understand the length and distance of the time the vehicle is traveling straight. [Effects of the Invention]

[0022] According to the driving state analysis method and program having the above characteristics, it is possible to detect the presence or absence of an operation with a risk factor while suppressing the amount of data to be determined. In addition, when determining the driving state based on video data, it is not necessary to view all of the long-time video, so that the labor required for the determiner can be reduced.

Brief Description of the Drawings

[0023] [Figure 1] It is a figure which shows the structural example of the vehicle-mounted device and computer for implementing the driving state analysis method which concerns on embodiment. [Figure 2] It is a figure which shows the flow at the time of implementing the driving state analysis method which concerns on embodiment. [Figure 3] It is a figure for demonstrating the flow of data preprocessing at the time of implementing a driving state analysis method. [Figure 4] It is a figure for demonstrating the flow of each driving state detection at the time of implementing a driving state analysis method. [Figure 5] It is a figure for demonstrating the analysis process for determining whether or not a right or left turn operation has occurred in a vehicle. [Figure 6] It is a figure for demonstrating how to obtain the local curvature radius r and the global curvature radius R regarding the right / left turn analysis of a vehicle when n1 = 1 and n2 = 3, and how to determine the threshold value thereof. [Figure 7] It is a figure for demonstrating the analysis process for determining whether or not a start operation has occurred in a vehicle. [Figure 8] It is a figure for demonstrating the analysis process for determining whether or not a stop operation has occurred in a vehicle. [Figure 9] It is a figure for demonstrating the analysis process for determining whether or not a reverse operation has occurred in a vehicle. [Figure 10] It is a figure which shows the structural example when a vehicle-mounted device is equipped with a vehicle signal detection means. [Figure 11] It is a figure which shows the change of a detected value at the time of determining whether or not a start operation has occurred in a vehicle using a vehicle speed pulse. [Figure 12]This figure shows the change in the detected value when determining whether or not a vehicle has stopped using a vehicle speed pulse. [Figure 13] This diagram shows the rise time (input) of the reverse gear signal and the change in the detected value when determining whether the vehicle is moving in reverse based on the change in vehicle speed. [Figure 14] This diagram shows an example of how to determine the vehicle's straight-ahead driving timing based on changes in the vehicle speed pulse, the vehicle's starting and stopping timings, and the vehicle's right and left turning timings. [Modes for carrying out the invention]

[0024] Hereinafter, embodiments of the operating state analysis method and operating state analysis program of the present invention will be described in detail with reference to the drawings. The driving state analysis method according to this embodiment involves inputting information (log information) detected by the in-vehicle device 10 into the computer 20 and executing a program that detects the timing of vehicle actions (events) such as turning left or right, starting, stopping, and reversing. In this embodiment, the in-vehicle device 10 can be any device that has a GPS receiver and an acceleration sensor that can be fixed to the vehicle body, such as a drive recorder, digital tachograph, GPS logger, or smartphone. In the following embodiments, the in-vehicle device 10 is mainly described using a drive recorder as an example.

[0025] The drive recorder, which corresponds to the in-vehicle device 10 according to this embodiment, comprises at least a shooting means 12, a location information acquisition means 14, an acceleration detection means 16, and a recording means 18. The shooting means 12 can be any device with camera functionality that can shoot video. The location information acquisition means 14 can be any so-called GPS function, and it is preferable that it can also calculate and record the distance traveled and speed based on the change in this location information. The acceleration detection means 16 can be any acceleration sensor. Furthermore, the recording means 18 can be any element that can record various information and video detected and calculated by the in-vehicle device 10 such as a drive recorder, and generally, external memory such as an SD card is used.

[0026] In the driving state analysis method according to this embodiment, the driving state is analyzed by inputting vehicle location information acquired by an in-vehicle device 10 such as a drive recorder, travel distance information obtained based on the location information and time information, speed information, and acceleration information into a computer 20. The computer 20 only needs to have at least a storage means 22, a calculation means 24, and a temporary storage means 26. The storage means 22 can be a large-capacity storage means such as an HDD or SSD, and is an element for storing various information recorded in the recording means 18 of the in-vehicle device 10. The input of various information can be done via an interface (not shown), and may be done via a dedicated slot, wired connection, or wireless communication.

[0027] The arithmetic means 24 is a means for reading various information stored in the memory means 22 and performing calculations according to instructions such as a program, and is generally referred to as the CPU. The temporary storage means 26 is an element for temporarily expanding the information read from the memory means 22 by the arithmetic means 24 and performing calculation processing, and can be any element equivalent to so-called internal memory.

[0028] [analysis] The analysis of the driving state by the computer 20 configured as described above will be explained with reference to the flow shown in Figures 2 to 4. First, various information recorded in the recording means 18 of the in-vehicle device 10 is acquired by the computer 20. The acquisition of information may be direct via a dedicated slot or wired connection, or indirectly via a wireless communication line. The various information acquired by the computer 20 is stored in the storage means 22 in association with time information (Step 10: Data acquisition).

[0029] The various data taken into the storage means 22 are preprocessed to improve the accuracy of the analysis. The preprocessing of the various data is performed as shown in Figure 3. Specifically, first, the calculation means 24 reads the various data from the storage means 22 and removes data that may cause a decrease in detection accuracy. In detail, the process excludes detected values ​​in situations where GPS location information is difficult to acquire, such as immediately after engine start or immediately before stop (for example, immediately after engine start, the accuracy of the detected location information decreases because GPS cannot be acquired. Also, immediately before engine stop, the engine may stop before the vehicle has come to a complete stop, which reduces the accuracy of location information detection). In this embodiment, locations where the interval between continuously recorded detection information is 1 minute or more are determined to be in an engine stop state, and the recorded information for 10 seconds before and after that is excluded (step 22: filtering).

[0030] Next, the influence of the installation status of detection devices such as the in-vehicle unit 10 on the detection data is removed. Specifically, the influence of gravitational acceleration in the X-axis and Y-axis directions, which occurs depending on the tilt of the in-vehicle unit 10 (sensors inside the in-vehicle unit 10), is removed. In this embodiment, the average value of the entire recorded information is subtracted from the detected value so that the fluctuation value of acceleration is clearly shown (Step 24: Bias removal).

[0031] Furthermore, for the data that has been filtered and bias-removed, information indicating that the vehicle is in an idling state (for example, data during a predetermined period when both acceleration and speed are zero) is cut off (Step 26: Idling Cutoff). By subjecting various data to filtering, bias removal, and idling cutoff in this way, it becomes possible to utilize the data for highly accurate analysis of the driving state (Step 20: Pretreatment of Data).

[0032] After pretreatment is performed on various data, detection regarding each driving state is carried out. The order of detection of each driving state is not particularly limited, but for example, as shown in FIG. 4, it is preferably carried out in the order of right / left turn (Step 32), start (Step 34), stop (Step 36), reverse (Step 38).

[0033] [Right / Left Turn: Step 32] First, the state detection regarding a right / left turn is detected based on the minimum moving distance L calculated based on the position information and time information of the vehicle (specifically, the in-vehicle device) (see FIG. 5(A)). The minimum moving distance L refers to the distance between detection points where the distance from a reference detection point among the position information of the vehicle detected every predetermined time (for example, every 1 second) is equal to or greater than a predetermined distance (for example, 10 m or more).

[0034] The operation detection of a right / left turn is such that the arithmetic means 24 reads out the position information (associated with time information) stored in the storage means 22 and detects five detection points whose distances from each detection point satisfy the interval defined as the minimum moving distance L. After detecting the five detection points, the arithmetic means 24 designates the detection point located at the center of the five detection points as the reference detection point P0, and designates a pair of detection points that are n1L away from the reference detection point P0 as the adjacent detection points P1, P2. Further, the arithmetic means 24 designates the detection points located on the opposite side of the reference detection point P0 via the adjacent detection points P1, P2 (detection points that are n2L away from the reference detection point P0) as the start detection point P3 and the end detection point P4, respectively. Here, n1 and n2 are integers indicating distances in units of the minimum moving distance L, and the relationship between the two satisfies n1 < n2.

[0035] The calculation means 24 determines each detection point (P0 to P4) and then calculates the local radius of curvature r, which draws a circle passing through the reference detection point P0 and the two adjacent detection points P1 and P2 (the reference detection point P0 and the two adjacent detection points P1 and P2 are located on the circumference), and the global radius of curvature R, which draws a circle passing through the reference detection point P0, the starting detection point P3, and the ending detection point P4 (the reference detection point P0, the starting detection point P3, and the ending detection point P4 are located on the circumference) (see Figures 5(B) and (C)). Here, threshold values ​​are set for the local radius of curvature r and the global radius of curvature R, respectively. The calculation means 24 then compares the local radius of curvature r and the global radius of curvature R with thresholds, and if both are below the respective thresholds, it determines that a right or left turn was made at the point where the reference detection point P0 is located, and identifies the time information of the detection of the reference detection point P0 as the timing of the vehicle's right or left turn.

[0036] Here, for example, if the vehicle is moving straight, it is not possible to draw a circle passing through the three detection points, as shown in Figure 5(A). In contrast, if the vehicle is turning right or left (turning right if the right side in the figure is the direction of travel), both the local radius of curvature r and the global radius of curvature R will be relatively small, as shown in Figure 5(B). On the other hand, if the vehicle's action is not a right or left turn but a lane change, etc. (including cases where there is an actual change in the vehicle's action, as well as cases where errors due to noise etc. occur in the GPS detection data), as shown in Figure 5(C), the local radius of curvature r will be a relatively small value, but the global radius of curvature R will be a larger value than that shown in (B). Therefore, when both the local radius of curvature r and the global radius of curvature R are below the threshold, it is possible to determine that the vehicle has performed a right or left turn.

[0037] In the above explanation, adjacent detection points P1 and P2 are defined as points 1L (n1=1) away from the reference detection point P0, and the starting detection point P3 and ending detection point P4 are defined as points 2L (n2=2) away from the reference detection point P0, respectively. However, n1 and n2 do not necessarily have to be consecutive values; for example, n1=1 and n2=3 may be defined. Furthermore, the threshold for the local radius of curvature r (rmax) and the threshold for the global radius of curvature R (Rmax) should be set to values ​​greater than the maximum curvature when each detection point is considered to have undergone a right or left turn.

[0038] For example, in the case where n1=1 and n2=3, as shown in Figure 6, it can be recognized that a right or left turn may have occurred if the reference detection point P0 and the adjacent detection point P2 satisfy at least the following relationship: that is, the sum of the angles θy and θx formed at the intersection of the line L connecting the two (P0 and P2), the line Y connecting two or more detection points following the reference detection point P0 on the entry side, and the line X connecting two or more detection points following the adjacent detection point P2 on the exit side is 90 degrees or more.

[0039] In the example shown in Figure 6, θy and θx are both set to 45 degrees, and lines Y and X intersect at a 90-degree angle. Therefore, the reference detection point P0 and the adjacent detection point P2 are located at positions L / √2 away from the intersection point O. In this relationship, if θ1 is the angle between line Y connecting adjacent detection point P1 and reference detection point P0 and line L1 connecting adjacent detection points P1 and adjacent detection points P2, and θ2 is the angle between line Y connecting starting detection point P3 and reference detection point P0 and line L2 connecting starting detection point P3 and ending detection point P4, then the local radius of curvature r and the global radius of curvature R can be expressed as shown in Equations 1 and 2, respectively.

number

number

[0040] Here, sinθ1 and sinθ2 can be shown by equations 3 and 4, respectively.

number

number

[0041] The threshold values ​​rmax for the local radius of curvature r and Rmax for the global radius of curvature R need to be set to values ​​larger than the calculated values ​​of r and R, respectively, considering the differences in angles at intersections where right and left turns occur, the detection error of GPS position information, and the timing difference in the detection of detected coordinates. Taking these factors into account, in the above embodiment, rmax and Rmax can be set to values ​​slightly less than twice the calculated values, for example, rmax = 20 and Rmax = 50.

[0042] [Starting: Step 34] Next, the state detection related to starting (forward movement) is detected based on changes in the vehicle's speed information and changes in acceleration information. Here, thresholds (upper limit of stopping speed: arbitrary value, lower limit of forward acceleration: arbitrary value) are defined for both the speed information and the acceleration information, as shown in Figure 7. The calculation means 24 reads the vehicle's speed information and acceleration information (both of which are associated with time information) from the storage means 22 and detects changes in each piece of information.

[0043] Specifically, the system detects when the vehicle's speed exceeds a threshold (upper limit of stopping speed) and starts moving, and when the vehicle's acceleration exceeds a threshold (lower limit of forward acceleration), it determines that a starting action has been performed. Time information is then detected as the timing of the start, and this is recognized as the moment the vehicle started moving.

[0044] [Stop: Step 36] Next, the stopping state is detected based on changes in the vehicle's speed information and the elapsed time. Here, the speed information has a threshold (upper limit of stopping speed) set for determining whether the vehicle has stopped, similar to when it is starting (see Figure 8). It is preferable that the speed information threshold for starting detection and the speed information threshold for stopping detection be the same value.

[0045] The calculation means 24 determines that a vehicle has stopped if the vehicle speed remains below a threshold (upper limit of stopping speed) for a predetermined minimum stopping time (e.g., 3 seconds) or longer. This is because if the determination were based solely on the change in speed, the vehicle would be incorrectly determined to have stopped even if the speed information locally falls below the threshold due to the influence of noise or braking. Furthermore, even if the calculation means 24 detects a moving speed exceeding the threshold (upper limit of stopping speed) within a predetermined minimum moving time (e.g., 3 seconds) after determining that the vehicle has stopped, it determines that the stopped state is maintained if the moving speed subsequently decreases to below the threshold. This is to account for changes in the detected value due to the influence of noise or other factors. The calculation means 24 then detects the time information when these stopping operations occurred and determines the timing at which the vehicle stopped or was in a stopped state.

[0046] [Backward: Step 38] Finally, the detection of the reverse state is performed based on changes in vehicle speed information and acceleration information, similar to the detection of the starting state. Here, as shown in Figure 9, a threshold (reverse acceleration lower limit: arbitrary value) is defined for acceleration information in the opposite direction to the direction of travel, i.e., in the negative direction (Step 30: Each driving detection).

[0047] The calculation means 24 determines that the vehicle has reversed when it is determined that the vehicle is in a stopped state, and when it detects acceleration information in the negative direction that falls below a threshold (lower limit of backward acceleration), and at this time, it detects an increase in speed information (the speed information does not need to reach the upper limit of the stopping speed), it detects the time information, and determines that this is the timing when the vehicle reversed.

[0048] In step 30, after each operating state has been detected, the time information of the detected operation is output (step 40: timing output).

[0049] [effect] This driving state analysis method and program allows for the detection of the presence and timing of actions involving risk factors while keeping the amount of data to be analyzed to a minimum. Therefore, investigations of driving states based on dashcam footage can be conducted by picking out the timing of right / left turns, starting, stopping, and reversing, which are detected as driving actions involving risk factors. This makes it possible to improve the efficiency of the assessment and reduce the workload required of the assessor.

[0050] [Application Example 1] In the above embodiment, the detection of a vehicle turning right or left is determined by whether the local radius of curvature r and the global radius of curvature R, based on the detection point, are below a threshold. However, in addition to the above determination elements, the accuracy of determining whether a right or left turn has occurred may be improved by detecting the presence or absence of lateral acceleration (the presence or absence of acceleration exceeding a threshold). In this case, if both the local radius of curvature r and the global radius of curvature R are below a threshold, and the lateral acceleration exceeds a threshold, it will be determined that a right or left turn has been performed on the vehicle.

[0051] [Application Example 2] In the above embodiment, it is stated that the vehicle speed information is calculated based on changes in location information obtainable by the location information acquisition means 14 (GPS function). However, the on-board unit 10 may also acquire the vehicle speed pulse and use it as speed information. If such a method is adopted, as shown in Figure 10, the on-board unit 10 should be equipped with a vehicle signal detection means 28 to enable the acquisition of vehicle speed pulses. In such a configuration, the vehicle signal detection means 28 performs a process of recording the acquired vehicle speed pulse in the recording means 18 in association with time information.

[0052] With this configuration, as shown in Figures 11 and 12, the speed information is represented by a rectangular waveform, reducing noise and errors in the speed information compared to the speed information shown in Figures 7 and 8. Therefore, it becomes easier to determine whether the vehicle is moving or not compared to when speed information is calculated based on position information (for example, the examples shown in Figures 7 and 8). As a result, it becomes possible to set a lower threshold for stopping speed (upper limit of stopping speed). Consequently, it becomes possible to more accurately determine when the vehicle's speed crosses the threshold (upper limit of stopping speed), that is, whether the vehicle is moving or not. Therefore, the accuracy of detecting the timing of starting and stopping can be improved.

[0053] [Application Example 3] Furthermore, the above embodiment describes detecting the vehicle's reversal based on changes in speed information relative to the vehicle's direction of travel and changes in acceleration information. However, if the in-vehicle unit 10 is configured to include a vehicle signal detection means 28, this vehicle signal detection means 28 can also be configured to detect the vehicle's reverse gear signal. In such a configuration, as shown in Figure 13, the timing of the vehicle's reversal can be accurately detected by the rising edge (input) of the reverse gear signal and the change in speed information (increase in the vehicle speed pulse).

[0054] Furthermore, it becomes possible to accurately detect the grace period from the rise of the reverse gear signal to the rise of the vehicle speed pulse (start of reverse), as well as the maximum speed and distance traveled during reverse. Therefore, by setting thresholds for grace period, maximum speed, and distance traveled, it becomes possible to detect actions involving risk factors during vehicle reverse, and the time period in which such actions occurred. Thus, the efficiency of detection of judgment points by judges can be improved, and the workload of judges can be reduced.

[0055] [Application Example 4] Furthermore, by detecting vehicle speed pulses via the vehicle signal detection means 28, the vehicle's starting and stopping timings can be accurately detected. For example, as shown in Figure 14, the vehicle's starting timing and turning timing, or the vehicle's turning timing and stopping timing, or the detection interval from the starting timing to the stopping timing, i.e., the interval between events that could be hazardous, can be identified as the vehicle's straight-ahead timing.

[0056] [Application Example 5] Furthermore, when analyzing driving conditions, it becomes possible to detect traffic violations and identify the date and time of such violations based on events that could be risk factors, location information where those events occurred, map information that can reflect that location information, traffic regulation information attached to the map information (e.g., stop signs, one-way streets, speed limit information, etc.), and speed information. [Explanation of Symbols]

[0057] 10...In-vehicle device, 12...Photography means, 14...Location information acquisition means, 16...Acceleration detection means, 18...Recording means, 20...Computer, 22...Storage means, 24...Calculation means, 26...Temporary storage means, 28...Vehicle signal detection means.

Claims

1. A method for analyzing the driving state of a vehicle based on information obtainable by an in-vehicle device, The information acquired by the in-vehicle device is stored in the computer's storage means. The information stored in the storage means is read out. The computer's calculation means predetermines a minimum travel distance calculated based on the vehicle's position information, detects five detection points using the minimum travel distance as the unit of distance, defines the detection point located in the center of the five detection points as the reference detection point, the detection points located before and after the reference detection point as adjacent detection points, the detection point located on the opposite side of the reference detection point via the adjacent detection points as the starting detection point and the ending detection point, respectively. The local radius of curvature is determined by drawing a circle that passes through the reference detection point and the two adjacent detection points, and the global radius of curvature is determined by drawing a circle that passes through the reference detection point, the starting detection point, and the ending detection point. A driving state analysis method characterized in that, when both the local radius of curvature and the global radius of curvature are below a predetermined threshold, the time information of the detection of the reference detection point is detected as the right or left turning timing of the vehicle.

2. The driving state analysis method according to claim 1, characterized in that the calculation means detects the vehicle's starting timing based on time information in which the timing at which the vehicle's speed rises above a predetermined threshold coincides with the timing at which the vehicle's acceleration exceeds a predetermined threshold.

3. The driving state analysis method according to claim 2, characterized in that the calculation means detects the timing at which the vehicle's speed falls below a predetermined threshold and no further speed increase occurs for a predetermined period of time or longer as the stopping timing of the vehicle.

4. The driving state analysis method according to claim 3, characterized in that, when the calculation means detects an increase in the vehicle's speed at the timing when the acceleration in the opposite direction to the direction of travel occurs in the vehicle during the stopping timing, it detects this as the vehicle's reversing timing.

5. The driving state analysis method according to claim 4, characterized in that the calculation means detects the time period during which the vehicle is driving straight, based on the vehicle's starting timing and the vehicle's turning timing, or the vehicle's stopping timing and the vehicle's turning timing, or the vehicle's starting timing and the vehicle's stopping timing.

6. The driving state analysis method according to claim 2, characterized in that the in-vehicle device acquires the vehicle's reverse gear signal, and when it detects an increase in the vehicle's speed at the timing when the reverse gear signal rises, it detects this as the vehicle's reverse timing.

7. The driving state analysis method according to claim 6, characterized in that the calculation means detects the time period during which the vehicle is driving straight, based on the vehicle's starting timing and the vehicle's turning timing, or the vehicle's stopping timing and the vehicle's turning timing, or the vehicle's starting timing and the vehicle's stopping timing.

8. The driving state analysis method according to any one of claims 2 to 7, characterized in that the in-vehicle device acquires the vehicle speed pulse signal of the vehicle and incorporates it into the storage means as vehicle speed information.

9. A program that is executed using vehicle location information acquired by an in-vehicle device and stored in a storage means in association with time information, A process of reading the location information stored in the storage means via the calculation means, A minimum travel distance calculated based on the aforementioned position information is predetermined, five detection points are detected using the minimum travel distance as the unit of distance, the detection point located in the center of the five detection points is defined as the reference detection point, the detection points located before and after the reference detection point are defined as adjacent detection points, the detection point located on the opposite side of the reference detection point via the adjacent detection points is defined as the starting detection point and the ending detection point, respectively. A process for determining the local radius of curvature of a circle that passes through the aforementioned reference detection point and two of the aforementioned adjacent detection points, and the global radius of curvature of a circle that passes through the aforementioned reference detection point, the starting detection point, and the ending detection point. A driving state analysis program characterized by performing the following: when both the local radius of curvature and the global radius of curvature are below a predetermined threshold, the program determines that the time information of the detection of the reference detection point is the right or left turning timing of the vehicle.

10. The information acquired by the in-vehicle device includes speed information and acceleration information associated with the time information, and when this information is taken into the storage means together with the position information, A process of reading velocity information and acceleration information stored in the storage means via the calculation means, The driving state analysis program according to claim 9, characterized by performing the following processes: detecting the timing at which the speed of the vehicle rises above a predetermined threshold based on the speed information, and detecting the timing at which the acceleration of the vehicle exceeds a predetermined threshold based on the acceleration information, and certifying the time information at which both coincide as the vehicle's starting timing.

11. A process of reading the stored speed information via the calculation means, The driving state analysis program according to claim 10, characterized in that it performs a process to determine the timing at which the vehicle stops if, after the vehicle's speed falls below a predetermined threshold based on the speed information, no further speed increase occurs for a predetermined period of time or longer.

12. The driving state analysis program according to claim 11, characterized in that, during the aforementioned stopping timing, it detects the timing at which acceleration occurs in the vehicle in the direction opposite to the direction of travel, and based on the speed information, it detects whether or not the vehicle's speed has increased, and performs a process to determine the time information at which both coincide as the timing for the vehicle to reverse.

13. The information acquired by the in-vehicle device includes a reverse gear signal associated with the time information, and when this is taken into the storage means together with the position information, The driving state analysis method according to claim 11, characterized in that when an increase in the vehicle's speed is detected at the timing when the reverse gear signal rises, a process is executed to detect this as the timing for the vehicle to reverse.

14. The driving state analysis program according to claim 12 or 13, characterized in that it performs a process to detect the time period during which the vehicle is moving straight, based on the vehicle's starting timing and the vehicle's right or left turning timing, or the vehicle's stopping timing and the vehicle's right or left turning timing.

Images