Traffic volume estimation device

Abstract

To provide a traffic volume estimation device capable of accurately estimating traffic volumes based on remote sensing and data from connected cars.SOLUTION: A traffic volume estimation device includes: an overall traffic volume measurement unit 320 that pre-detects the number of all vehicles in a predetermined area based on remote sensing by analyzing satellite images; a partial traffic volume measurement unit 322 that pre-detects the number of connected vehicles in the predetermined area based on location information obtained from connected vehicles capable of V2X communication; and a traffic volume estimation unit 326 that estimates a traffic volume in an arbitrary area based on a CA ratio which is a ratio of the number of connected vehicles pre-detected in the predetermined area with respect to the number of all vehicles pre-detected in the predetermined area, and the number of connected vehicles in the arbitrary area near the predetermined area that is newly detected by the partial traffic volume measurement unit 322.SELECTED DRAWING: Figure 4

Description

[Technical Field]

[0001] The present invention relates to a traffic volume estimation device that accurately estimates traffic volume based on remote sensing and data from connected cars. [Background technology]

[0002] Remote sensing based on satellite imagery or aerial photographs may be used to estimate traffic volume within a designated area.

[0003] Patent Document 1 discloses an invention for a traffic situation observation system that estimates the traffic volume of a predetermined area by using a first image and a second image of a predetermined area acquired at different times, and searching for each vehicle extracted from the first image in the second image. [Prior art documents] [Patent Documents]

[0004] [Patent Document 1] Japanese Patent Publication No. 2005-56186 [Overview of the Initiative] [Problems that the invention aims to solve]

[0005] However, the invention described in Patent Document 1 requires multiple satellite images for each predetermined region. Since satellite images have a high unit cost per image, there was a problem with high operating costs.

[0006] Furthermore, since it is difficult to obtain images of the same area over a long period of time using satellite imagery and aerial photographs, it is not easy to obtain two images, a first and a second, of the same subject.

[0007] Furthermore, current satellite imagery does not have high resolution, making it difficult to identify individual vehicles from the images using image recognition. This has resulted in problems with the accuracy of traffic volume estimation.

[0008] Considering the above facts, the present invention aims to provide a traffic volume estimation device that accurately estimates traffic volume based on remote sensing and data from connected cars. [Means for solving the problem]

[0009] To achieve the above objective, the traffic volume estimation device described in claim 1 includes a first detection unit that pre-detects the total number of vehicles in a predetermined area by remote sensing, and a unit that transmits the vehicle's position information via wireless communication. and vehicle information The location information obtained from a connected vehicle capable of transmitting and the aforementioned vehicle information Based on the above, the connected vehicle in the predetermined area By vehicle type A second detection unit that pre-detects the number of vehicles, and the number of connected vehicles pre-detected in the predetermined area relative to the total number of vehicles pre-detected in the predetermined area. By vehicle type The connected vehicle ratio is the ratio of the number of vehicles, and the number of connected vehicles in an arbitrary area near the predetermined area newly detected by the second detection unit. By car model Based on the number of units, in the arbitrary area By vehicle type It includes a traffic volume estimation unit that estimates traffic volume.

[0010] Traffic volume as described in claim 1 Estimate According to the device, based on data from remote sensing and connected cars By car model Traffic volume can be estimated with high accuracy.

[0011] The traffic volume estimation device according to claim 2 is configured such that the predetermined area includes the connected vehicle.

[0012] According to the traffic volume estimation device described in claim 2, location information from connected vehicles can be reliably obtained.

[0013] In the traffic volume estimation device according to claim 3, the predetermined area and the arbitrary area are set by dividing map information into predetermined mesh units.

[0014] According to the traffic volume estimation device described in claim 3, by setting the predetermined area and the arbitrary area with square meshes, the computational processing load can be suppressed.

[0015] The traffic volume estimation device described in claim 4, wherein the remote sensing is at least any one of image analysis of satellite images, image analysis of image data acquired by a fixed-point camera, detection of traffic volume by a traffic counter, and analysis of communication records by an electronic toll collection system equipped in a vehicle.

[0016] According to the traffic volume estimation device described in claim 4, by utilizing various remote sensing, the number of all vehicles in a predetermined area can be accurately detected.

[0017] The traffic volume estimation device described in claim 5, wherein the number of all vehicles in the predetermined area by the first detection unit, and the connected vehicles in the predetermined area by the second detection unit By car model The number of each is acquired in advance according to weekdays, holidays, days of the week, and time zones, and the traffic volume estimation unit acquires in advance the connected vehicle ratio in the predetermined area according to weekdays, holidays, days of the week, and time zones. The number of all vehicles in the predetermined area and the connected vehicles in the predetermined area By car model The number is calculated by the number of all vehicles in the predetermined area and the number of connected vehicles in the predetermined area, and the connected vehicles in the arbitrary area newly detected by the second detection unit according to weekdays, holidays, days of the week, and time zones By car model Based on the number of vehicles and the connected vehicle ratio according to weekdays, holidays, days of the week, and time zones corresponding to the newly acquired number of connected vehicles, the any Traffic volume in the area is estimated.

[0018] According to the traffic volume estimation device described in claim 5, by calculating the traffic volume based on the connected vehicle ratio calculated in advance according to weekdays, holidays, days of the week, and time zones and the number of connected vehicles newly acquired according to weekdays, holidays, days of the week, and time zones, the accuracy of the estimated traffic volume can be improved. Vehicle-specific stands By calculating the traffic volume based on the number of vehicles, the accuracy of the estimated traffic volume can be improved.

Advantages of the Invention

[0019] As described above, the traffic volume estimation device according to the present invention can accurately estimate traffic volume based on remote sensing and data from connected cars. [Brief explanation of the drawing]

[0020] [Figure 1] This is a schematic diagram showing an example of a specific configuration of the traffic volume estimation device according to this embodiment. [Figure 2] This is a block diagram showing an example of a processing server configuration. [Figure 3] This is a block diagram showing an overview of the processing of the traffic volume estimation device according to this embodiment. [Figure 4] This is an example of a CPU function block diagram during program execution. [Figure 5] This is a flowchart illustrating an example of the processing performed by the processing server. [Modes for carrying out the invention]

[0021] The traffic volume estimation device 100 according to this embodiment will be described below with reference to Figure 1. Figure 1 is a schematic diagram showing an example of a specific configuration of the traffic volume estimation device 100 according to this embodiment. As shown in Figure 1, the traffic volume estimation device 100 according to this embodiment includes a satellite 70 capable of acquiring satellite images of the ground on which vehicles 80A, 80B, and 80C are traveling, a satellite information server 54 that receives the satellite images acquired by satellite 70, a V2X server 56 that can communicate with vehicle 80C, which is a so-called connected vehicle, via V2X communication and acquires location information and speed information of vehicle 80C, a map information DB server 58 equipped with a map information database (DB), and a processing server 10 that estimates traffic volume based on information acquired via the network 52 from each of the satellite information server 54, V2X server 56, and map information DB server 58.

[0022] Figure 2 is a block diagram showing an example of the configuration of the processing server 10. The processing server 10 includes a computer 30. The computer 30 is equipped with a CPU 32, ROM 34, RAM 36, and input / output ports 38. As an example, it is desirable that the computer 30 be a model capable of executing advanced computational processing at high speed, such as an engineering workstation or a supercomputer.

[0023] In the computer 30, the CPU 32, ROM 34, RAM 36, and input / output ports 38 are connected to each other via various buses such as an address bus, a data bus, and a control bus. Various input / output devices are connected to the input / output ports 38, including a display 40, a mouse 42, a keyboard 44, a hard disk drive (HDD) 46, and a disk drive 50 that reads information from various disks (e.g., CD-ROMs and DVDs) 48.

[0024] Furthermore, the input / output port 38 is connected to a network 52, enabling the exchange of information with various devices connected to the network 52. In this embodiment, a satellite information server 54, a V2X server 56, and a map information DB server 58 are connected to the network 52.

[0025] The HDD 46 of the computer 30 has installed the following programs: a total traffic volume measurement program that extracts vehicles from satellite images and measures traffic volume; a partial traffic volume measurement program that estimates the number of vehicles 80C in a predetermined range based on location information and speed information of vehicles 80C capable of V2X communication; a vehicle ratio calculation program that calculates the ratio between the number of vehicles calculated by the total traffic volume measurement program and the number of vehicles calculated by the partial traffic volume measurement program; and a target traffic volume estimation program that estimates the traffic volume in a predetermined range using the calculation results from the partial traffic volume measurement program and the vehicle ratio calculated by the vehicle ratio calculation program.

[0026] In this embodiment, the CPU 32 executes a total traffic volume measurement program, which initiates machine learning using satellite images and other data as training data for mathematical models such as CNNs (Convolutional Neural Networks), and constructs a total traffic volume estimation model, which is a trained model based on machine learning.

[0027] Furthermore, when the CPU 32 executes the target traffic volume estimation program, machine learning is initiated, and a traffic volume estimation model, which is a trained model based on machine learning, is constructed. The CPU 32 displays the processing results of the program on the display 40.

[0028] There are several ways to install the program according to this embodiment onto the computer 30. For example, the program can be stored on a CD-ROM or DVD along with a setup program, the disk can be inserted into the disk drive 50, and the setup program can be executed on the CPU 32 to install the program onto the HDD 46. Alternatively, the program can be installed onto the HDD 46 by communicating with other information processing equipment connected to the computer 30 via a public telephone line or network 52.

[0029] Figure 3 is a block diagram showing an overview of the processing of the traffic volume estimation device 100 according to this embodiment. In step S300, the number of connected vehicles is calculated and vehicle recognition is performed using remote sensing with satellite images. The number of connected vehicles is calculated based on information such as the vehicle's CAN (Controller Area Network) data or probes that indicate the actual position and speed at which the vehicle traveled. Vehicle recognition from satellite images involves extracting vehicles from satellite images within a predetermined range and measuring the total traffic volume of all vehicles, including connected vehicles, within that predetermined range.

[0030] In step S302, in step S302A, the ratio of connected vehicles to all vehicles is calculated from the number of connected vehicles and the total traffic volume calculated in step S300. In step S302B, a population estimation model is constructed by statistically registering, as a database (DB), the ratio of connected vehicles to all vehicles in a plurality of regions (a plurality of ranges). The ratio of connected vehicles to all vehicles in step S302 and the population estimation model are statistically calculated from data obtained in advance from connected vehicles and data such as satellite images.

[0031] In step S304, based on the traffic volume estimation model, traffic volume data in a specific region or an area near the specific region, which is the target of traffic volume estimation, is calculated and output from the data registered in the population estimation model and the number of connected vehicles in the specific region or an area near the specific region newly acquired.

[0032] The CA ratio, which is the ratio of connected vehicles to all vehicles, is the number of connected vehicles V c0 , the number of all vehicles being V a0 , and is shown, for example, by the following formula (1). CA ratio = V c0 / V a0 …(1)

[0033] Also, the number of connected vehicles newly acquired in the specific region is V c1 , the traffic volume in the specific region is V a1 , and assuming that V c1 / V a1 is equal to the above CA ratio, the traffic volume V a1 in the specific region is calculated by the following formula (3). [[ID=I33]] V c0 / V a0 = V c1 / V a1 …(2) V a1 = V a0 * V c1 / V c0 …(3)

[0034] Figure 4 is an example of a functional block diagram of the CPU 32 during program execution. The CPU 32 has a total traffic volume measurement function that measures traffic volume by extracting vehicles from satellite images, etc., by executing a total traffic volume measurement program; a partial traffic volume measurement function that estimates the number of vehicles 80C in a predetermined range based on the location information and speed information of vehicles 80C capable of V2X communication by executing a partial traffic volume measurement program; a vehicle ratio calculation function that calculates the ratio between the number of vehicles calculated by the total traffic volume measurement program and the number of vehicles calculated by the partial traffic volume measurement program by executing a vehicle ratio calculation program; and a traffic volume estimation function that estimates the traffic volume in a predetermined range using the calculation results from the partial traffic volume measurement program and the vehicle ratio calculated by the vehicle ratio calculation program by executing a target traffic volume estimation program. When the CPU 32 executes a program having each of these functions, the CPU 32 functions as a total traffic volume measurement unit 320, a partial traffic volume measurement unit 322, a vehicle ratio calculation unit 324, and a traffic volume estimation unit 326.

[0035] The overall traffic volume measurement unit 320 includes a remote data acquisition unit 320A that acquires remote data such as satellite images from satellite 70 via the network 52, an overall traffic volume measurement program 320B that extracts vehicles from satellite images and measures traffic volume, and an overall traffic volume storage unit 320C that stores the overall traffic volume obtained by executing the overall traffic volume measurement program in an HDD 46 or the like.

[0036] The partial traffic volume measurement unit 322 includes a connected vehicle data acquisition unit 322A that acquires location information and speed information of vehicles 80C from vehicles 80C capable of V2X communication via the network 52, a partial traffic volume measurement program 322B that estimates the number of vehicles 80C present in a predetermined range based on the location information and speed information of the vehicles 80C capable of V2X communication, and a partial traffic volume holding unit 322C that stores the partial traffic volume calculated by the partial traffic volume measurement program 322B in an HDD 46 or the like.

[0037] The vehicle ratio calculation unit 324 includes a vehicle ratio calculation program 324A that calculates the ratio between the number of vehicles calculated by the overall traffic volume measurement program and the number of vehicles calculated by the partial traffic volume measurement program, and a vehicle ratio data storage unit 324B that stores the calculated vehicle ratio data in an HDD 46 or the like.

[0038] The traffic volume estimation unit 326 includes a target traffic volume estimation program 326A that estimates traffic volume within a predetermined range using the calculation results from the partial traffic volume measurement program and the vehicle ratio calculated by the vehicle ratio calculation program, and an estimation result holding unit 326B that stores the estimation results from the target traffic volume estimation program in an HDD 46 or the like.

[0039] In this embodiment, in calculating traffic volume in the overall traffic volume measurement unit 320, the partial traffic volume measurement unit 322, the vehicle ratio calculation unit 324, and the traffic volume estimation unit 326, map information stored in the map information DB server 58 is referenced as appropriate.

[0040] Figure 5 is a flowchart showing an example of processing by the processing server 10. The periodic execution thread is run periodically (for example, annually, monthly, or weekly) for the purpose of statistical sample collection. In step S100, the remote data acquisition unit 320A of the overall traffic volume measurement unit 320 determines a predetermined range 1 to be imaged. Map information stored in the map information DB server 58 is referenced to determine the predetermined range 1.

[0041] The following parallel processing is performed, where steps S102, S104, and S106 are processed in parallel with steps S108 and S110. In this embodiment, the processing results in steps S102, S104, and S106 are stored in the HDD 46 or the like by the overall traffic volume storage unit 320C, and the processing results in steps S108 and S110 are stored in the HDD 46 or the like by the partial traffic volume storage unit 322C. If the processing results are to be stored, the procedures in steps S102, S104, and S106 and steps S108 and S110 may be executed in a single thread.

[0042] In step S102, the remote data acquisition unit 320A acquires satellite images of a predetermined range 1 via the satellite information server 54 and the network 52.

[0043] In step S104, the overall traffic volume measurement program 320B of the overall traffic volume measurement unit 320 detects vehicles in the acquired satellite image. Then, in step S106, the number of vehicles detected in the satellite image is counted to calculate the total traffic volume within a predetermined range 1.

[0044] In step S108, the connected vehicle data acquisition unit 322A of the partial traffic volume measurement unit 322 extracts location information and speed data of connected vehicles (CVs) capable of V2X communication that are traveling within a predetermined range 1. Then, in step S110, the partial traffic volume measurement program 322B of the partial traffic volume measurement unit 322 counts the number of CVs within the predetermined range 1.

[0045] In step S112, the vehicle ratio calculation program 324A of the vehicle ratio calculation unit 324 calculates the CA ratio, which is the ratio of connected vehicles C to the total traffic volume A in a predetermined range 1.

[0046] In step S114, the vehicle ratio data holding unit 324B of the vehicle ratio calculation unit 324 stores the calculated CA ratio as a database in the HDD 46 or the like, in association with a location (predetermined range 1), and terminates the periodic execution. The calculated CA ratio is also used in step S202 of the sequential execution described later.

[0047] The sequential execution thread is executed as needed when it is desired to estimate the traffic volume in an arbitrarily determined range A. In step S200, the traffic volume estimation unit 326's target traffic volume estimation program 326A determines the range A for which the traffic volume is to be determined. Map information stored in the map information DB server 58 is referenced to determine range A. Range A is a region near a predetermined range 1 determined by the periodic execution thread, but it may also be the same region as predetermined range 1.

[0048] The following parallel processing is performed, where steps S202 and S204 and steps S206 and S208 are processed in parallel. In this embodiment, the processing results in steps S202 and S204 are stored in the HDD 46 or the like by the estimation result storage unit 326B, and the processing results in steps S206 and S208 are stored in the HDD 46 or the like by the partial traffic volume storage unit 322C. If the processing results are to be stored, the procedures in steps S202 and S204 and steps S206 and S208 may be executed in a single thread.

[0049] In step S202, the traffic volume estimation unit 326 uses the target traffic volume estimation program 326A to search for data within range A or near range A from the data calculated in step S114 of the periodic execution thread and registered in the DB.

[0050] In step S204, the traffic volume estimation unit 326 uses its target traffic volume estimation program 326A to extract and obtain the vehicle ratio A, which is the CA ratio for range A, from the search results of step S202.

[0051] In step S206, the connected vehicle data acquisition unit 322A of the partial traffic volume estimation unit 322 extracts location information and speed data of V2X communication-enabled CVs traveling within range A. Then, in step S208, the partial traffic volume measurement program 322B of the partial traffic volume measurement unit 322 counts the number of CVs A within range A.

[0052] In step S210, the traffic volume estimation program 326A of the traffic volume estimation unit 326 estimates the traffic volume in range A using the vehicle ratio A obtained in step S204, the number of CVs A obtained in step S208, and the above formula (3). Then, in step S212, the calculation result from step S210 is output as the estimated traffic volume within range A, and the estimation result holding unit 326B holds the estimated traffic volume within range A in the HDD 46 or the like, and the execution is terminated.

[0053] As explained above, according to this embodiment, traffic volume can be accurately estimated based on remote sensing and data from connected cars, using the CA ratio, which is the ratio of connected vehicles to the total traffic volume in each region calculated by remote sensing through prior statistical processing, and the newly acquired number of connected vehicles in each region.

[0054] While overall regional traffic volume can be calculated from satellite imagery, the high cost per image results in high operational costs. However, in this embodiment, regional traffic volume is calculated from the CA ratio statistically calculated in advance for each region by the periodic execution thread shown in Figure 5, and the number of connected vehicles for each region that has been newly acquired. Therefore, traffic volume can be estimated at a lower cost compared to calculating traffic volume solely from satellite imagery.

[0055] Furthermore, in this embodiment, the computational load can be reduced by dividing the map information into predetermined mesh units and calculating the CA ratio for each mesh. Alternatively, the CA ratio may be calculated for each road by referring to the map information. In addition, in order to ensure the accuracy of the CA ratio calculation, a specific range that is the target area for traffic volume estimation may be set so that connected vehicles are always included. Specifically, when setting the mesh, the map information is divided so that connected vehicles are always present in each mesh.

[0056] Furthermore, by obtaining the total traffic volume and the number of connected vehicles separately for weekdays, holidays, days of the week, and time of day, and calculating the CA ratio for each, and then calculating the traffic volume based on the newly obtained number of connected vehicles separately for weekdays, holidays, days of the week, and time of day, and the corresponding CA ratio for the time of day, the accuracy of the estimated traffic volume can be improved.

[0057] Furthermore, connected vehicles can also transmit data specific to each vehicle type, such as light vehicles, passenger cars, and commercial vehicles. Therefore, by multiplying the traffic volume estimated in step S210 of Figure 5 by the ratio of the number of connected vehicles of each vehicle type to the total number of connected vehicles, it is possible to estimate the traffic volume for each vehicle type.

[0058] Furthermore, in this embodiment, the total traffic volume was estimated by remote sensing using image analysis of satellite images, but instead of satellite images, image analysis of image data acquired by a fixed-point camera, detection of traffic volume using a traffic counter, or communication records from an electronic toll collection system (ETC2.0) installed in the vehicle may be used. Moreover, remote sensing using satellite images, fixed-point cameras, traffic counters, and detection results from ETC2.0 may be used. fruit By using each method in combination, the accuracy of calculating overall traffic volume can be improved. [Explanation of Symbols]

[0059] 10 Processing Servers 30 Computers 32 CPU 54 Satellite Information Server 56 V2X Servers 58 Map Information DB Server 70 satellites 80A, 80B, 80C vehicles 100 Traffic volume estimation device 320 Overall traffic measurement section 322 Partial traffic measurement section 324 Vehicle Ratio Calculation Unit 326 Traffic Estimation Department

Claims

1. A first detection unit that pre-detects the total number of vehicles in a predetermined area using remote sensing, A second detection unit that pre-detects the number of connected vehicles of each type in a predetermined area based on the location information and vehicle type information obtained from a connected vehicle capable of transmitting its own location information and vehicle type information via wireless communication, A traffic volume estimation unit estimates the traffic volume by vehicle type in the arbitrary area based on the connected vehicle ratio, which is the ratio of the number of connected vehicles by vehicle type detected in the predetermined area to the total number of vehicles detected in the predetermined area, and the number of connected vehicles by vehicle type in an arbitrary area near the predetermined area newly detected by the second detection unit. A traffic volume estimation device that includes [this component].

2. The traffic volume estimation device according to claim 1, wherein the predetermined area is set to include the connected vehicle.

3. The traffic volume estimation device according to claim 2, wherein the predetermined area and the arbitrary area are set by dividing map information into predetermined mesh units.

4. The traffic volume estimation device according to claim 3, wherein the remote sensing is at least one of the following: image analysis of satellite images, image analysis of image data acquired by a fixed-point camera, detection of traffic volume using a traffic counter, and analysis of communication records from an electronic toll collection system installed in a vehicle.

5. The first detection unit obtains the total number of vehicles in the predetermined area, and the second detection unit obtains the number of connected vehicles by type in the predetermined area, separately for weekdays, holidays, days of the week, and time of day. The traffic volume estimation device according to any one of claims 1 to 4, wherein the traffic volume estimation unit calculates the connected vehicle ratio in the predetermined area using the total number of vehicles in the predetermined area and the number of connected vehicles by type in the predetermined area, which have been previously acquired separately for weekdays, holidays, days of the week, and time of day, and estimates the traffic volume in the arbitrary area based on the number of connected vehicles by type in the arbitrary area newly detected by the second detection unit separately for weekdays, holidays, days of the week, and time of day, and the connected vehicle ratio for each weekday, holiday, day of the week, and time of day corresponding to the newly acquired number of connected vehicles.

Images