JP7811488B2 - Driving history analysis system - Google Patents

Description

The present invention relates to a driving history analysis system.

Patent Document 1 describes a system that verifies whether a polygon belonging to a probe car's parking location is a site related to the destination, and if it is confirmed that there is a relationship between the two, registers the deviation position from the last traveled link and identifies the parking lot entrance based on this registered deviation position.

Patent No. 5437674

The system described in Patent Document 1 can identify parking lot entrances, but the parking lot polygon may differ from the actual shape of the parking lot. In such cases, the parking spot polygon may not be linked to the parking lot. For example, if a parking lot is divided into sections, polygon data may not be linked to a specific parking lot. This may result in the vehicle not being recognized as having been parked in the linked parking lot, even though it is actually parked in the linked parking lot. This may result in the inability to obtain parking-related information or driving history.

The present invention was made in consideration of the above-mentioned problems, and aims to provide a driving history analysis system that increases the likelihood of identifying information about parking, such as parking in a parking lot, even when parking lot data is insufficient.

To achieve the above objective, the present invention comprises a driving history acquisition unit that acquires the driving history of vehicles parked in parking lot polygon areas associated with parking lot attributes, an analysis target history acquisition unit that acquires analysis target history, which is the driving history of vehicles parked in non-parking lot polygon areas not associated with the parking lot attributes, and an attribute determination unit that determines the attributes of the non-parking lot polygons, and when the distance between the parking lot polygon indicated by the driving history and the non-parking lot polygon indicated by the analysis target history is less than a predetermined threshold, the attribute determination unit records the non-parking lot polygon as the parking lot polygon in a recording unit.

In other words, in the driving history analysis system, if the distance between a parking lot polygon associated with a parking lot attribute and a non-parking lot polygon not associated with a parking lot attribute is less than a predetermined threshold, the non-parking lot polygon is recorded as a parking lot polygon. This makes it possible to identify polygons not associated with a parking lot attribute that are likely to represent a parking lot as parking lots. As a result, even if parking lot data is insufficient, for example, it is possible to increase the likelihood of identifying information about parking, such as parking in a parking lot.

In order to achieve the above-mentioned objectives, the present invention comprises a driving history acquisition unit that acquires the driving history of vehicles parked in a parking lot polygon area associated with a parking lot attribute; an analysis target history acquisition unit that acquires an analysis target history, which is the driving history of vehicles parked in a non-parking lot polygon area not associated with the parking lot attribute; and an attribute determination unit that determines the attributes of the non-parking lot polygon, and the attribute determination unit records the non-parking lot polygon as the parking lot polygon in a recording unit if a first parking trajectory, which is the trajectory of a vehicle parked in the non-parking lot polygon area indicated by the analysis target history, and a second parking trajectory, which is the trajectory of a vehicle parked in the parking lot polygon area indicated by the driving history, satisfy a predetermined condition.

In other words, in the driving history analysis system, if the parking trajectory of a non-parking lot polygon that is not associated with a parking lot attribute (first parking trajectory) and the parking trajectory of a parking lot polygon that is associated with a parking lot attribute (second parking trajectory) satisfy a predetermined condition (e.g., the parking trajectories are similar), the non-parking lot polygon is recorded as a parking lot polygon. This makes it possible to identify polygons that are not associated with a parking lot attribute but are likely to represent a parking lot as parking lots. As a result, even if parking lot data is insufficient, the likelihood of identifying information about parking, such as parking in a parking lot, can be increased.

1 is a block diagram showing the configuration of a driving history analysis system according to an embodiment of the present invention; FIG. 10 is a diagram illustrating a configuration for determining the attributes of non-parking lot polygons based on the distance between polygons. FIG. 10 is another diagram illustrating a configuration for determining the attributes of non-parking lot polygons based on the distance between polygons. FIG. 10 is yet another diagram illustrating a configuration for determining the attributes of non-parking lot polygons based on the distance between polygons. 10 is a diagram illustrating a configuration for determining attributes of a non-parking lot polygon based on a parking lot locus. FIG. 4 is a flowchart illustrating an example of processing of the system according to the present embodiment. 10 is a subroutine in the processing of the system in this embodiment. 10 is another subroutine in the processing of the system in this embodiment.

Here, the embodiments of the present invention will be described in the following order.
(1) Configuration of driving history analysis system:
(2) Flowchart:
(3) Other embodiments:

(1) Configuration of driving history analysis system:
The driving history analysis system in this embodiment is a system that acquires and analyzes driving history information about which parking lots a vehicle has been parked in. Fig. 1 is a block diagram showing the configuration of a driving history analysis system 10 according to the present invention. The driving history analysis system 10 is a server that can communicate with a vehicle 100. Although not shown, the driving history analysis system 10 can communicate with multiple vehicles.

Vehicle 100 is a so-called probe car, and is equipped with a communication unit 110 and a positioning unit 120. Vehicle 100 also has a recording medium (not shown) that records map information 130 and driving history 140. Communication unit 110 is a wireless communication circuit for wireless communication with driving history analysis system 10. Positioning unit 120 is equipped with positioning sensors such as a GNSS (Global Navigation Satellite System) receiver, a vehicle speed sensor, and a gyro sensor (none of which are shown).

Map information 130 includes node data indicating the positions of nodes set on roads on which vehicles travel, shape interpolation point data indicating the positions of shape interpolation points used to identify the shape of roads between nodes, link data indicating the connections between nodes, and facility data indicating the positions of facilities on and around roads.

Facility data may include various types of information. This information may include, for example, polygon data that represents the two-dimensional shape of the facility. Specifically, for example, a polygon representing the outer perimeter of the facility is divided into triangles, and information representing each triangle (e.g., information representing the vertices) is associated with the facility data as polygon data. Note that the polygon data may not be divided into triangles, and information representing each vertex of the polygon may be associated with the facility data. Note that the facility data is associated with the name of the facility.

Facility data may also include information indicating the attributes of the facility. In this embodiment, if the attribute of a facility is a parking lot, information indicating that the attribute of the facility is a parking lot is associated with the facility data. When an attribute indicating that the facility is a parking lot is associated with facility data, the polygon is said to have an attribute indicating that the parking lot is associated with it, and such a polygon is defined as a parking lot polygon. On the other hand, if the attribute of the facility is not associated with information indicating that it is a parking lot, the polygon associated with the facility data is defined as a non-parking lot polygon. Note that when the attribute of a facility is a parking lot, a name such as "XX Parking Lot" or "XX Park" is associated as the facility name.

The driving history 140 includes, as probe information, at least information regarding the parking of the vehicle 100. Specifically, the driving history is associated with the identification information of the vehicle 100 and includes information indicating the vehicle's position over time (at regular intervals or at regular times), which is the vehicle's trajectory. In this embodiment, the driving history is a single history that covers the period from the start of vehicle 100's driving to its end, with the end point of the driving being the parking position. Each vehicle position is associated with information indicating whether map matching was performed. If map matching was performed, the vehicle's position is on a road; if map matching was not performed, the vehicle is off the road (for example, in a facility such as a parking lot).

The vehicle 100 then transmits probe information including the driving history 140 to the driving history analysis system 10 via the communication unit 110 at regular time intervals.

Next, the driving history analysis system 10 will be described. The driving history analysis system 10 includes a control unit 20 equipped with a CPU, RAM, ROM, etc., a recording unit 30, and a communication unit 40. The recording unit 30 records various programs and data. The communication unit 40 is a wireless communication circuit for wireless communication with the vehicle 100. The control unit 20 executes various programs stored in the recording unit 30 and ROM. As an example of this program, the control unit 20 can execute a driving history analysis program 21, and in this embodiment, can modify information associated with polygons in map information based on the driving history.

The recording unit 30 records map information 30a and driving history 30b. The map information 30a includes node data, shape interpolation point data, link data, facility data, etc. The map information 30a in the driving history analysis system 10 is updated based on probe information from each probe car, including the vehicle 100. In other words, the accuracy of the map information 30a is improved by updating the map information based on the probe information. The driving history 30b, like the driving history 140 of the vehicle 100, includes at least parking-related information, such as the vehicle's trajectory. The driving history 30b also includes probe information from each probe car, including the vehicle 100. In other words, the driving history 30b records a large number of driving histories corresponding to the identification information of a large number of vehicles.

The relationship between facility data and polygon data will now be explained. As mentioned above, facility data may include polygon data that represents the two-dimensional shape of the facility. Polygon data associated with facility data may not have facility attributes associated with it. This is because it is difficult to accurately define polygons and attributes for all facilities on a map, and some facility data may be inaccurate or insufficient. In other words, even if polygon data exists, facility data may also exist, but attributes may not be associated with that facility data. Furthermore, a single parking lot may be represented by multiple polygons (in other words, multiple facility data). In this case, parking lot attributes may be associated with some polygons, while parking lot attributes may not be associated with the remaining polygons. This embodiment is configured to determine the attributes of polygons that do not have parking lot attributes associated with them, i.e., non-parking lot polygons.

The control unit 20 can execute various programs stored in the recording unit 30 or ROM. In this embodiment, the control unit 20 can execute a driving history analysis program 21 as one example of such a program. The driving history analysis program 21 is a program that causes the control unit 20 to function as a driving history acquisition unit 21a, an analysis target history acquisition unit 21b, and an attribute determination unit 21c. Note that, below, the processes described as being performed by the driving history acquisition unit 21a, the analysis target history acquisition unit 21b, and the attribute determination unit 21c are processes implemented by the control unit 20.

The driving history acquisition unit 21a acquires the driving history of vehicles parked in the area of a parking lot polygon associated with parking lot attributes. Specifically, the driving history acquisition unit 21a acquires the driving history recorded in the recording unit 30. At this time, the driving history acquisition unit 21a identifies the parking spot of the vehicle and, with reference to map information 30a, determines whether the parking spot is included in the polygon based on the parking spot. If the parking spot is included in the polygon, the driving history acquisition unit 21a refers to map information 30a and identifies the attribute of the facility corresponding to the polygon. Then, the driving history acquisition unit 21a acquires the driving history of vehicles for which the facility attribute has been identified as a parking lot.

The analysis target history acquisition unit 21b acquires the analysis target history, which is the driving history of vehicles parked in non-parking lot polygon areas that are not associated with parking lot attributes. Specifically, the analysis target history acquisition unit 21b acquires the driving history recorded in the recording unit 30. At this time, the analysis target history acquisition unit 21b identifies the vehicle's parking location and, with reference to map information 30a, determines whether the parking location is included in a polygon based on the parking location. If the parking location is included in a polygon, the analysis target history acquisition unit 21b references map information 30a and acquires one driving history for which the attribute of the facility corresponding to the polygon is not a parking lot attribute, and treats that driving history as the analysis target.

The attribute determination unit 21c determines the attributes of non-parking lot polygons. Specifically, it determines the attributes of non-parking lot polygons acquired by the analysis target history acquisition unit 21b. The determination of the attributes of non-parking lot polygons is explained in detail below.

The control unit 20 first determines whether a parking lot polygon exists among polygons adjacent to the non-parking lot polygon. That is, the control unit 20 determines whether a polygon associated with a parking lot attribute exists around a non-parking lot polygon to which a parking lot attribute is not associated. If the control unit 20 determines that a parking lot polygon exists around the non-parking lot polygon, it determines whether the distance between the non-parking lot polygon and the parking lot polygon is less than a predetermined threshold. If the control unit 20 determines that the distance between the non-parking lot polygon and the parking lot polygon is less than the threshold, it further determines whether a vehicle parked in the area of the non-parking lot polygon passed through the parking lot polygon. Specifically, the control unit 20 references the analysis target history and determines whether the vehicle passed through other polygons before entering the polygon containing the parking spot based on the driving position and map information 30a. If a vehicle parked in a non-parking lot polygon passed through other polygons before entering the polygon containing the parking spot, the control unit 20 identifies the polygon through which the vehicle passed immediately before entering the polygon containing the parking spot. If the control unit 20 identifies the polygon through which the vehicle traveled immediately before entering the polygon containing the parking spot and determines that the polygon through which the vehicle traveled immediately before entering is a parking lot polygon, it assigns a parking lot attribute to the non-parking lot polygon. Additionally, the control unit 20 associates the non-parking lot polygon with the same name as the name associated with the parking lot polygon. In other words, the control unit 20 regards the non-parking lot polygon as an adjacent parking lot polygon. The control unit 20 then records the non-parking lot polygon as a parking lot polygon in the recording unit 30.

With reference to Figure 2, the determination of the attributes of non-parking lot polygons will be specifically described. In the example shown in Figure 2, there is a parking lot that includes three polygons P1, P2, and P3 as a whole, and these three polygons P1, P2, and P3 include parking lot polygons and non-parking lot polygons. In Figure 2, the symbol E indicates the entrance/exit of the parking lot, the symbol O indicates a point on the road where map matching was not possible, and the symbol P indicates the parking point of the vehicle. A road R and the like exist around the polygons P1, P2, and P3. The road R is a road with intersections _I1 and _I2 as its endpoints (nodes). In Figure 2, the dashed lines indicate an example of a vehicle's parking trajectory, and the dashed-dotted lines indicate links connecting the nodes.

In the example in Figure 2, of polygons P1, P2, and P3, only polygon P2 is a parking lot polygon. In other words, only polygon P2 is associated with a parking lot attribute. In the example shown in Figure 2, a "☆" is added to the area of polygon P2 to indicate that polygon P2 is associated with a parking lot attribute and is a parking lot polygon. On the other hand, polygons P1 and P3 are non-parking lot polygons, meaning that polygons P1 and P3 are not associated with a parking lot attribute.

The control unit 20 determines whether the adjacent polygon, polygon P2, is a parking lot polygon because the parking spot is within polygon P1 and polygon P1 is a non-parking lot polygon. As described above, polygon P2 is a parking lot polygon, so it determines whether the distance L between polygon P1 and polygon P2 is less than a threshold value. Note that the distance L between polygons P1 and P2 is the length of the perpendicular line connecting the points where the perimeter of one polygon (polygon P1) is closest to the perimeter of the other polygon (polygon P2). In the example shown in Figure 2, one side of polygon P1 and polygon P2 overlaps, and the length of the perpendicular line (i.e., the distance L between the polygons) is "0". Therefore, it can be said that the distance L is less than the threshold value.

The control unit 20 further determines whether a vehicle parked in the area of polygon P1 passed through polygon P2. Specifically, the control unit 20 references the analysis target history and determines whether the vehicle passed through other polygons (e.g., polygons P2 and P3) before entering polygon P1, which includes the parking spot. If a vehicle parked in polygon P1 passed through other polygons P2 and P3 before entering polygon P1, the control unit 20 identifies the polygon through which the vehicle traveled immediately before entering polygon P1 and determines whether a parking lot attribute is associated with the identified polygon. In the example shown in Figure 2, the polygon through which the vehicle traveled immediately before entering polygon P1, which includes the parking spot, is polygon P2, and polygon P2 is a parking lot polygon.

If the polygon traveled through immediately before entering the polygon containing the parking spot is a parking lot polygon, the control unit 20 assigns a parking lot attribute to the non-parking lot polygon. Therefore, in the example shown in Figure 2, the control unit 20 assigns a parking lot attribute to polygon P1. Furthermore, in this embodiment, when the control unit 20 assigns a parking lot attribute to a non-parking lot polygon, it assigns a name associated with the adjacent parking lot polygon. In other words, the control unit 20 regards the non-parking lot polygon as a parking lot polygon and associates the name associated with polygon P2 with polygon P1.

On the other hand, as shown in Figure 3, if the distance between polygon P1 and polygon P2 is large, i.e., if the distance L between polygon P1 and polygon P2 is equal to or greater than the threshold, the control unit 20 considers it difficult to conclude whether the vehicle parking spot P is a parking lot or not, and does not associate a parking lot attribute with the facility data of polygon P1.

Next, as shown in Figure 4, an example will be described in which, of polygons P1, P2, and P3, polygon P3 is a parking lot polygon and polygons P1 and P2 are non-parking lot polygons. In the example shown in Figure 4, the vehicle's parking spot P is within polygon P1. When parking within polygon P1, the vehicle passes through polygon P3, which is a parking lot polygon, but polygon P2, through which the vehicle traveled immediately before entering polygon P1, is a non-parking lot polygon and is not associated with a parking lot attribute. Therefore, the control unit 20 deems it difficult to conclude whether the vehicle's parking spot P is a parking lot or not, and does not associate a parking lot attribute with the facility data of polygon P1.

In this way, in this embodiment, if the distance between a parking lot polygon associated with a parking lot attribute and a non-parking lot polygon not associated with a parking lot attribute is less than a predetermined threshold, the non-parking lot polygon is assigned a parking lot attribute and recorded as a parking lot polygon. This makes it possible to identify polygons not associated with a parking lot attribute that are likely to represent a parking lot as parking lots. As a result, even if parking lot data is insufficient, the likelihood of identifying information about parking, such as parking in a parking lot, can be increased.

Next, we will explain an example of a case where a vehicle does not pass through a parking lot polygon before entering a polygon containing a parking spot. As described above, if a vehicle passes through a parking lot polygon before entering a polygon containing a parking spot, the control unit 20 assigns a parking lot attribute to the non-parking lot polygon. On the other hand, if a vehicle does not pass through a parking lot polygon before entering a polygon containing a parking spot, the control unit 20 determines the attribute of the polygon containing the parking spot by comparing the parking trajectory based on the driving history (the history to be analyzed) with the parking trajectory based on the driving history of parking in the area of the parking lot polygon, in addition to making a judgment based on the distance between adjacent polygons. Note that in this embodiment, if the distance between the non-parking lot polygon and the parking lot polygon is equal to or greater than a threshold, it is considered difficult to consider the parking spot as a parking lot, as described above, and the parking lot attribute is not assigned to the non-parking lot polygon (i.e., the process of considering the non-parking lot polygon as a parking lot polygon is not performed). Therefore, we will explain an example of a case where a vehicle does not pass through a parking lot polygon before entering a polygon containing a parking spot, assuming that the distance between the polygon containing the parking spot and the parking lot polygon is less than a threshold.

If a vehicle has not passed through a parking lot polygon before entering a polygon containing a parking spot, assuming that the distance between the parking spot polygon and the parking lot polygon is less than a threshold (i.e., they are adjacent), the polygon containing the parking spot is assumed to be located in front of the parking lot polygon. However, even if the polygon containing the parking spot is located in front of the parking lot polygon, this does not necessarily mean that the parking spot is a parking lot. As in the example above, once a parking lot polygon has been passed, there is little error in treating a non-parking lot polygon as a parking lot polygon. However, if the vehicle has not yet passed through or entered the parking lot polygon, there may be an error in treating the polygon containing the parking spot as a parking lot polygon. Therefore, in this embodiment, if a vehicle has not passed through a parking lot polygon before entering the polygon containing the parking spot, the attributes of the parking spot polygon are determined based on the parking trajectory.

Here, we will explain the parking trajectory of a vehicle. The parking trajectory is information that indicates the time series changes in the vehicle's position, and is calculated by vehicle 100, which is a probe car. Specifically, as described above, vehicle 100 is equipped with a GNSS receiver, a vehicle speed sensor, and a gyro sensor as positioning sensors. The GNSS receiver receives radio waves from navigation satellites and outputs a signal for calculating the vehicle's current location via an interface (not shown). The vehicle speed sensor outputs a signal corresponding to the rotational speed of the vehicle's wheels. The gyro sensor detects the angular acceleration of the vehicle's turning in a horizontal plane, and outputs a signal corresponding to the vehicle's orientation.

The vehicle 100 acquires its current location within a facility such as a parking lot (i.e., outside a road) using a control unit (not shown) based on the self-contained navigation trajectory, which is a trajectory of the vehicle's position estimated based on signals output from the vehicle speed sensor and gyro sensor, and map information 130. If the target for acquiring the current location is a road, multiple possible comparison roads are set and the comparison roads are narrowed down based on the error circle of the GNSS signal acquired by the GNSS receiver. The vehicle 100 then references the map information 130 and performs a map matching process to estimate that the road on which the vehicle is traveling is the road whose shape most closely matches the self-contained navigation trajectory, among the narrowed-down comparison roads, and acquires the vehicle's current location on the road estimated by the map matching process. During the course of travel, the vehicle 100 performs a process of acquiring the vehicle's current location based on the output signals from the GNSS receiver, vehicle speed sensor, and gyro sensor at predetermined distance intervals or predetermined time intervals. The vehicle 100 then acquires information showing the history of the current location at predetermined intervals, i.e., information showing the changes in the current location, as a parking trajectory, and records the acquired parking trajectory in the driving history 140.

The parking trajectory calculated in this manner is transmitted from the vehicle 100 to the driving history analysis system 10 and recorded in the recording unit 30 of the driving history analysis system 10. Therefore, the control unit 20, using the function of the analysis target history acquisition unit 21b, references the recording unit 30 to acquire parking trajectories in which parking points are included in non-parking lot polygons.

In this embodiment, the recording unit 30 records a driving history 30b, which includes information about the vehicle's parking trajectory. The control unit 20 calculates a parking trajectory S to a parking spot P, using a point on the road that is outside of map matching as the origin O (i.e., the starting point of the parking trajectory). Specifically, the control unit 20 acquires one analysis target history in which the facility attribute corresponding to the polygon containing the parking spot is not a parking lot attribute, and acquires the parking trajectory S of that analysis target history. The control unit 20 acquires the parking trajectory S by referencing the driving history 30b recorded in the recording unit 30.

More specifically, the control unit 20 extracts coordinates at every fixed distance a [m] from the origin O to the parking point P, which are points outside the map matching, to obtain the parking trajectory S. For example, if the distance traveled from the origin O of the parking trajectory to the parking point P is b [m], the number of extracted points d is
d=b/a [pieces]
This becomes:

The control unit 20 calculates the coordinates for the number of extraction points d. If the coordinates of the start point of the parking trajectory are _S0 ( _X0 , _Y0 ), the parking trajectory S from the start point to the parking point (i.e., parking completion) can be expressed as follows:
S=(S ₀ (X ₀ , Y ₀ ), S ₁ (X ₁ , Y ₁ ), S ₂ (X ₂ , Y ₂ ),...S _d (X _d , Y _d ))

The control unit 20 performs the same process on the parking lot polygon, which is the comparison target history. As described above, the recording unit 30 includes the driving history 30b, which includes a parking locus. Therefore, the control unit 20, using the function of the driving history acquisition unit 21a, extracts, from the recorded parking history, a parking locus Sα whose parking point is included in the parking lot polygon and whose length is the same as that of the parking locus _S , which is the analysis target history. Specifically, the control unit 20 refers to the parking locus whose parking point is included in the parking lot polygon, and extracts the parking locus _Sα whose distance from the origin O is the same as that of the parking locus S. The acquired parking locus _Sα can be expressed as follows:
S _α = (S _α0 (X _α0 , Y _α0 ), S _α1 (X _α1 , Y _α1 ), ...S _αd (X _αd , Y _αd ))
The method for calculating the parking locus _Sα , which is the history to be compared, is the same as the method for calculating the parking locus S, which is the history to be analyzed, and therefore will not be described here.

Next, the control unit 20 compares the calculated parking locus S with the parking locus _Sα . Specifically, the control unit 20 calculates the distance between each corresponding point on the parking locus S and the parking locus _Sα , and calculates the deviation of the parking locus S from the parking locus _Sα . That is, the control unit 20 calculates the distance between each corresponding coordinate (e.g., _S0 and _Sα0 ) on the parking locus S and the parking locus _Sα . Then, the control unit 20 regards the sum of the calculated distances as the difference between the two loci, and determines whether it is less than a predetermined value. If the control unit 20 determines that the sum of the distances between each point on the parking locus S and the parking locus _Sα is less than the predetermined value, the control unit 20 regards the parking locus S and the parking locus _Sα as similar or the same. That is, the control unit 20 determines that the vehicle was parked in the non-parking lot polygon along a trajectory similar to the trajectory taken when the vehicle was parked inside the parking lot polygon, and that the two polygons are close to each other, so that the attribute of the facility associated with the non-parking lot polygon can be considered to be a parking lot. In other words, the control unit 20 considers the non-parking lot polygon to be a parking lot polygon.

If the sum of the distances between the parking locus S and the parking locus _Sα is equal to or greater than a predetermined value, the control unit 20 extracts the parking locus _Sβ in the parking data as the next comparison target and compares the parking locus S with the parking locus _Sβ . This process of comparing the parking loci is then repeated at least until it is determined that the distance between the parking loci is less than the predetermined value. In this embodiment, the parking locus S corresponds to the "first parking locus", the parking locus _Sα corresponds to the "second parking locus", and the fact that the parking loci are similar or the same corresponds to the "predetermined condition".

With reference to Figure 5, we will explain a specific example of a case where a vehicle does not pass through a parking lot polygon before entering a polygon containing a parking spot. In the example shown in Figure 5, the same symbols are used to indicate the same information as in Figure 2. In the example shown in Figure 5, the vehicle's parking spot P is within polygon P1. Of polygons P1, P2, and P3, polygon P2 is a parking lot polygon, and polygons P1 and P3 are non-parking lot polygons. In the example shown in Figure 5, the polygon containing the parking spot is located in front of parking lot polygon P3, and therefore polygon P3 is the polygon to be analyzed.

The control unit 20 calculates a parking locus S from the origin O, which is a point outside the map matching, to the parking point P. The control unit 20 also extracts a parking history to be compared recorded in the driving history 30b, and extracts and calculates a parking locus having the same length as the parking locus S in the comparison history. In the example shown in FIG. 5, comparison data α and comparison data β are extracted as comparison histories, and the parking locus S in the comparison data α is shown as a parking locus _Sα , and the parking locus S in the comparison data β is shown as a parking locus _Sβ . In FIG. 5, the parking locus _Sα is shown by a dashed line, and the parking locus _Sβ is shown by a dashed line.

The control unit 20 compares the parking locus S with each parking locus. Specifically, the control unit 20 compares the parking locus S with the parking locus _Sα , and the parking locus S with the parking locus _Sβ . As shown in FIG. 5, the parking locus S and the parking locus _Sα are parking loci that almost overlap each other, and the difference between them is small. Therefore, it can be said that the parking locus S and the parking locus _Sα are similar. In other words, the difference between the parking locus S and the parking locus _Sα is less than a predetermined value. On the other hand, even though the parking locus S and the parking locus _Sβ are loci of the same length, the end point of the parking locus S is within polygon P3, while the end point of the parking locus _Sβ is within polygon P2, and therefore the parking loci cannot be said to be similar. In other words, the difference between the parking locus S and the parking locus _Sβ is equal to or greater than a predetermined value.

As such, in the example shown in FIG. 5, the parking locus S and the parking locus _Sα are similar. In other words, it can be said that the parking locus S of a vehicle parked within polygon P1, which is a non-parking lot polygon, and the parking locus _Sα of a vehicle parked within polygon P2, which is a parking lot polygon, are loci obtained by the drivers of vehicles traveling with similar intentions. In other words, the parking locus S can be considered to be a locus resulting from traveling to park the vehicle in a parking lot. Therefore, in the example shown in FIG. 5, the attribute of the facility associated with polygon P3 can be considered to be a parking lot. In other words, the vehicle parked in polygon P3 can be considered to have been parked in the same parking lot as polygon P2.

In the example shown in Figure 5, the control unit 20 extracts and compares comparison data α and comparison data β as comparison data, but if at least one similar parking trajectory can be extracted, the non-parking lot polygon can be considered a parking lot polygon.

As described above, the driving history analysis system 10 configured in this manner determines the attributes of non-parking lot polygons based on the distance from the parking lot polygon and the parking trajectory. The processing executed by the driving history analysis system 10 of this embodiment is explained below using a flowchart.

(2) Flowchart:
6 is a flowchart showing an example of the parking lot attribute determination process executed by the control unit 20. In this embodiment, the parking lot attribute determination process is repeatedly executed at predetermined regular intervals (for example, once a week). Specific control contents will be described below.

First, the control unit 20 reads the driving history acquired after the previous execution of the parking attribute determination process (step S1). That is, the control unit 20 references the driving history 30b and acquires all of the driving history acquired after the previous execution of the parking attribute determination process.

The control unit 20 then begins the following loop processing (steps S2 to S5) for the driving history read in step S1. That is, the control unit 20 selects one of the driving histories acquired in step S1 as the processing target and performs the loop processing, sequentially repeating the loop processing for all driving histories acquired in step S1 until the loop processing is completed. Specifically, the control unit 20 determines whether a polygon exists at the parking spot (step S2). That is, the control unit 20 determines whether a polygon exists at the parking spot, in other words, whether the parking spot is within the polygon area (step S3). Note that steps S2 and S3 may be executed simultaneously, or the order of steps S2 and S3 may be reversed. Furthermore, since steps S2 and S3 are essentially the same, one of the steps may be omitted.

For example, if the parking spot is not within a facility, such as on a road, or is within a facility to which no polygon is associated, the control unit 20 determines that no polygon exists at the parking spot and makes a negative determination in step S3. If a negative determination is made in step S3, i.e., if it is determined that no polygon exists at the parking spot, the control unit 20 ends the loop processing for the driving history being processed and starts the loop processing for the next driving history.

On the other hand, if step S3 returns a positive determination, i.e., if it is determined that a polygon exists within the parking spot, the control unit 20 determines whether a parking lot attribute is associated with the polygon containing the parking spot (step S4). That is, the control unit 20 determines whether the polygon at the parking spot is a parking lot polygon. In other words, the control unit 20 determines whether the attribute of the facility associated with the polygon containing the parking spot is a parking lot attribute. If step S4 returns a positive determination, i.e., if it is determined that a parking lot attribute is associated with the polygon at the parking spot, the control unit 20 ends the loop processing for the driving history being processed without performing any special processing, because the polygon at the parking spot already has the parking lot attribute and the facility name associated with it, and then starts the loop processing for the next driving history.

On the other hand, if step S4 returns a negative determination, i.e., if it is determined that no parking lot attributes are associated with the parking spot polygon, the control unit 20 performs adjacent polygon-based parking lot attribute determination processing to determine parking lot attributes based on polygons adjacent to the parking spot (step S5). In this case, since the driving history shows a trajectory of parking in a non-parking lot polygon, the driving history is the history to be analyzed, and step S5 is executed for the history to be analyzed.

Figure 7 shows a subroutine for the parking lot attribute determination process based on adjacent polygons in step S5. When this process is performed, the polygon containing the parking spot is a non-parking lot polygon to which no parking lot attribute is assigned due to a negative determination in step S4 described above. The control unit 20 first determines whether any adjacent polygons have a parking lot attribute associated with them (step S50). That is, the control unit 20 determines whether any parking lot polygons exist within a predetermined distance around the polygon containing the parking spot. If the determination in step S50 is negative, i.e., if it is determined that no adjacent polygons have a parking lot attribute associated with them, the control unit 20 considers it difficult to conclude whether the polygon containing the parking spot is a parking lot or not, and does not associate a parking lot attribute with the facility data of the parking spot (step S51).

On the other hand, if the answer to step S50 is affirmative, i.e., if it is determined that there is an adjacent polygon to which a parking lot attribute is associated, the control unit 20 will initiate the following loop processing (steps S52 to S55) for each polygon to which a parking lot attribute is associated. In other words, the control unit 20 will perform the loop processing for all parking lot polygons among the adjacent polygons.

Specifically, the control unit 20 determines whether the distance between the polygon containing the parking spot and the parking lot polygon is less than a predetermined threshold (step S52). As explained above in FIG. 2, the control unit 20 determines that the distance between the polygons is less than the threshold if, for example, part of the perimeter of the polygon containing the parking spot overlaps with part of the perimeter of the parking lot polygon. On the other hand, as explained above in FIGS. 3 and 4, the control unit 20 determines that the distance between the polygons is greater than or equal to the threshold if the shortest distance between the perimeter of the polygon containing the parking spot and the perimeter of the parking lot polygon is greater than or equal to the threshold. Of course, this method of determining the distance is just one example, and the determination may also be made based on the distance between the centers of gravity of the polygons, etc.

If the answer to this step S52 is negative, i.e., if it is determined that the distance between the polygon containing the parking spot and the parking lot polygon is greater than or equal to the threshold, the control unit 20 starts a loop process for the relationship with the next polygon to which the parking lot attributes are associated.

On the other hand, if the answer to step S52 is affirmative, i.e., if the distance between the polygon containing the parking spot and the parking lot polygon is determined to be less than the threshold, the control unit 20 determines whether the vehicle passed through the parking lot polygon immediately before entering the polygon containing the parking spot (step S53). That is, the control unit 20 determines whether the polygon containing the parking spot is located further forward in the direction of travel than the parking lot polygon, and whether the parking lot polygon is the polygon immediately before (in the reverse direction of travel) the polygon containing the parking spot.

If the answer to this question in step S53 is affirmative, i.e., if it is determined that the vehicle passed through a parking lot polygon immediately before entering a polygon containing a parking spot, the control unit 20 assigns a parking lot attribute to the non-parking lot polygon (step S54). That is, the attribute of the polygon containing the parking spot is set to "parking lot." Additionally, the control unit 20 associates the polygon containing the parking spot with the same name as the name associated with the parking lot polygon.

On the other hand, if the answer to step S53 is negative, i.e., if it is determined that the vehicle did not pass through a parking lot polygon immediately before entering a polygon containing the parking spot, the control unit 20 performs parking lot attribute determination processing based on the parking trajectory (step S55). This parking lot attribute determination processing based on the parking trajectory will be described later.

The control unit 20 ends the loop processing when it has completed the processing of step S54 or step S55 for the number of polygons to which parking lot attributes have been associated. The control unit 20 then determines whether parking lot attributes have been associated with the driving history (analysis target history) (step S56). That is, the control unit 20 determines whether parking lot attributes have been associated with each polygon containing a parking spot through the above-mentioned loop processing. If parking lot attributes have not been associated through the loop processing, a negative determination is made in step S56, and in that case, the control unit 20 proceeds to step S51. That is, the control unit 20 considers it difficult to conclude whether each polygon containing a parking spot is a parking lot, and does not associate parking lot attributes with the facility data of the parking spot.

On the other hand, if the answer to step S56 is affirmative, i.e., if the loop processing determines that parking lot attributes are associated with the driving history, the control unit 20 terminates the subroutine shown in Figure 7.

Here, we will explain the parking lot attribute determination process based on the parking trajectory in step S55. Figure 8 shows a subroutine for the parking lot attribute determination process based on the parking trajectory in step S55. The control unit 20 performs the following loop processing (processing of steps S550 to S554) for all driving histories in which the vehicle parked in adjacent parking lot polygons. That is, the analysis target driving history, which is the driving history to be processed, indicates a trajectory in which the vehicle parked in a non-parking lot polygon, and the loop processing of steps S52 to S55 in Figure 7 is performed for a specific parking lot polygon adjacent to the non-parking lot polygon. Because step S55 is performed during this process, the processing shown in Figure 8 is performed in a state in which a specific parking lot polygon is adjacent to a non-parking lot polygon. Therefore, the control unit 20 references driving history 30b, extracts driving histories in which the vehicle parked in a parking lot polygon adjacent to a non-parking lot polygon, and performs the loop processing of steps S550 to S554 for each of these driving histories. The parking lot attribute determination process based on the parking trajectory shown in Figure 8 is performed when the vehicle does not pass through the parking lot polygon before entering the polygon containing the parking spot.

Specifically, the control unit 20 calculates the parking trajectory S from the point where the road is no longer within map matching to the parking spot (step S550). As described above in FIG. 5, the control unit 20 references the driving history 30b and, based on the analysis target history of parking in non-parking lot polygons, extracts coordinates for each fixed driving distance from the point where the road is no longer within map matching to the parking spot to obtain the parking trajectory S. Note that step S550 may be executed before the loop processing. In other words, after step S550 is executed, steps S551 to S554 may be repeated as the target of the loop processing.

The control unit 20 also acquires a parking locus _Sα in the comparison data α to be compared with the parking locus S (step S551). Specifically, the control unit 20 refers to the driving history 30b, and extracts coordinates at intervals of a certain driving distance from the point outside the map matching to the parking point, based on the driving history that is the processing target of the loop processing of steps S550 to S554. Then, the control unit 20 extracts a parking locus _Sα having the same length as the parking locus S, based on the extracted coordinates.

Next, the control unit 20 calculates the sum of the distances between the parking locus S and each point (coordinate) on the parking locus _Sα for the parking locus acquired in step S550 and step S551 (step S552). That is, the control unit 20 calculates the distance between each corresponding point on each parking locus and acquires the sum of the distances to calculate the difference between the parking locus S and the parking locus _Sα . Then, the control unit 20 determines whether the sum of the differences between the parking locus S and the parking locus _Sα is less than a predetermined value (step S553). That is, the control unit 20 determines whether the parking locus S is similar to the parking locus _Sα .

If the determination in step S553 is negative, i.e., if the difference between the parking locus S and the parking locus _Sα is equal to or greater than a predetermined value, the control unit 20 determines that the parking locus S and the parking locus _Sα are not similar, and compares them with the parking locus _Sβ , which is the next comparison data. On the other hand, if the determination in step S553 is positive, i.e., if the difference between the parking locus S and the parking locus _Sα is less than the predetermined value, the control unit 20 sets the attribute of the polygon containing the parking spot to "parking lot" (step S554). That is, the control unit 20 determines that the parking locus S and the parking locus _Sα are similar, and regards the non-parking lot polygon as a parking lot polygon. The control unit 20 also associates the name associated with the facility data of the comparison data α with the polygon containing the parking spot.

If the difference between the parking locus S and the parking locus _Sα is equal to or greater than a predetermined value and therefore the answer to step S553 is negative, the control unit 20 sequentially extracts comparison data (comparison data β, comparison data γ, ...) from adjacent polygons and performs the loop process in Fig. 8 until at least a similar parking locus is found. If the control unit 20 has extracted and compared all the target data and has not found a parking locus similar to the parking locus S, it ends the loop process and proceeds to step S56 in Fig. 7. If step S554 is executed, the control unit 20 may exit the loop process of steps S550 to S554 and end the subroutine.

As described above, in this embodiment, if the distance between a parking lot polygon associated with a parking lot attribute and a non-parking lot polygon not associated with a parking lot attribute is less than a predetermined threshold, the non-parking lot polygon is assigned a parking lot attribute and recorded as a parking lot polygon. Furthermore, in this embodiment, the attributes of the non-parking lot polygon are determined by comparing the parking trajectory of the non-parking lot polygon with the parking trajectory of the polygon associated with a parking lot attribute. This makes it possible to ascertain the attributes of the non-parking lot polygon, thereby increasing the likelihood of identifying information about parking, such as parking in a parking lot, even if, for example, parking lot data is insufficient. In other words, parking history can be obtained even if the polygon is not organized to match the actual shape of the parking lot. Furthermore, by being able to obtain the attributes of such non-parking lot polygons, information such as "which vehicle types can park in which parking lots" can be accumulated as parking history.

(3) Other embodiments:
The above embodiment is one example for implementing the present invention, and various other embodiments are also possible. For example, at least some of the components constituting the driving history analysis system 10 may be separated into multiple devices or systems. That is, at least some of the driving history acquisition unit 21a, analysis target history acquisition unit 21b, and attribute determination unit 21c constituting the driving history analysis system 10 may be separated into multiple devices. Furthermore, some of the components of the above embodiment may be omitted, and the order of processing may be changed or omitted.

In addition, in the above-described embodiment, the control unit 20 was configured to assign parking lot attributes to non-parking lot polygons when certain conditions are met based on the distance between polygons and parking trajectories. However, the control unit 20 may also assign parking lot attributes or the name of the facility as a parking lot to parking history (i.e., parking data). Furthermore, in the above-described embodiment, an attribute indicating that the non-parking lot polygon is a parking lot and the name of the adjacent parking lot are associated with the facility data associated with the non-parking lot polygon. However, it is also possible to associate only one of them. If an attribute indicating that the polygon is a parking lot is associated with facility data, it is possible to identify the area within the polygon as a parking lot. If the name of the adjacent parking lot is associated with facility data, it is possible to identify that the polygon is the same parking lot as the adjacent polygon.

In addition, in the above-described embodiment, when determining the attributes of a polygon containing a parking spot to be analyzed, the system is configured to determine the attributes based on either the distance between the polygons or the parking trajectory, depending on the positional relationship between the parking spot and the polygon associated with the parking spot attribute. That is, if the point where the vehicle is parked passes through the parking spot polygon, the attributes of the polygon containing the parking spot are determined based on the distance between the polygons. If the point where the vehicle is parked is located in front of the parking spot polygon but not through it, the attributes of the polygon containing the parking spot are determined based on the parking trajectory. However, in this embodiment, it is sufficient to be able to determine the attributes of the polygon containing the parking spot to be analyzed. Therefore, regardless of the positional relationship between the polygon containing the parking spot and the polygon associated with the parking spot attribute (i.e., whether the polygon containing the parking spot is in front of or behind the parking spot polygon in the direction of vehicle travel), the attributes may be determined based on the distance between the polygons or the parking trajectory of each polygon.

In other words, even if a polygon containing a parking spot is located in the reverse direction (near side) of the parking lot polygon, the attributes of the polygon containing the parking spot may be determined based on the distance between the polygons. Similarly, even if a polygon containing a parking spot is located in the forward direction (far side) of the parking lot polygon, the attributes of the polygon containing the parking spot may be determined based on the parking trajectory.

Furthermore, it is sufficient to determine the attributes of a polygon containing a parking spot using at least one of the distance between polygons and the parking trajectory, and therefore the attributes of a polygon containing a parking spot may be determined using both the distance between polygons and the parking trajectory.

The present invention can also be applied as a program or method. The above-mentioned systems, programs, and methods may be realized as stand-alone devices or may be realized using shared components, and include a variety of configurations. For example, it is possible to provide a method or program realized by the above-mentioned system. It can also be modified as appropriate, with some being software and some being hardware. Furthermore, the invention can also be realized as a recording medium for a program that controls a device. Of course, the recording medium for the software may be a magnetic recording medium or a semiconductor memory, and any recording medium developed in the future can be considered in exactly the same way.

10... Driving history analysis system, 20... Control unit, 21... Driving history analysis program, 21a... Driving history acquisition unit, 21b... Analysis target history acquisition unit, 21c... Attribute determination unit, 30... Recording unit, 30a... Map information, 30b... Parking history, 40... Communication unit, 100... Vehicle (probe car), 110... Communication unit, 120... Positioning unit, 130... Map information, 140... Parking history.

Claims (3)

a driving history acquisition unit that acquires driving histories of vehicles parked in areas of parking lot polygons associated with parking lot attributes;
an analysis target history acquisition unit that acquires an analysis target history, which is the travel history of a vehicle parked in an area of a non-parking lot polygon to which the parking lot attribute is not associated;
an attribute determination unit that determines the attributes of the non-parking lot polygons;
The attribute determination unit
If the distance between the parking lot polygon indicated by the driving history and the non-parking lot polygon indicated by the analysis target history is less than a predetermined threshold, the non-parking lot polygon is recorded as the parking lot polygon in a recording unit.
Driving history analysis system. The attribute determination unit
When the vehicle passes through the parking lot polygon in the travel history of the vehicle parked in the area of the non-parking lot polygon and the distance is less than the threshold, the non-parking lot polygon is recorded as the parking lot polygon in the recording unit.
The driving history analysis system according to claim 1 . a driving history acquisition unit that acquires driving histories of vehicles parked in areas of parking lot polygons associated with parking lot attributes;
an analysis target history acquisition unit that acquires an analysis target history, which is the travel history of a vehicle parked in an area of a non-parking lot polygon to which the parking lot attribute is not associated;
an attribute determination unit that determines the attributes of the non-parking lot polygons;
The attribute determination unit
When a first parking locus, which is a locus of a vehicle parked in the area of the non-parking lot polygon indicated by the analysis target history, and a second parking locus, which is a locus of a vehicle parked in the area of the parking lot polygon indicated by the driving history, satisfy a predetermined condition, the non-parking lot polygon is recorded as the parking lot polygon in a recording unit;
Extracting a trajectory having the same length as the first parking trajectory from the second parking trajectory;
determining that the predetermined condition is satisfied when a difference between the first parking locus and the extracted second parking locus is less than a predetermined value;
Driving history analysis system.