JP2022046120A - Location information furnishing method to on-vehicle sensor data, system and program for the same - Google Patents

Abstract

To provide a method by which a high accuracy location information is furnished to data acquired through an on-vehicle sensor and facilities along a traveling track of a vehicle are accurately referred with a ledger.SOLUTION: This location information furnishing method comprises a time-distance conversion step to convert a sampling number of on-vehicle sensor data sampled per an even time interval and wave form data per an even time interval into a sampling number of on-vehicle sensor data sampled per an even distance interval and a wave form data per an even distance interval based on a distance-time correspondence table, a location information furnishing step to convert a sampling number of on-vehicle sensor data of an even distance interval into a sampling number of on-vehicle sensor data sampled per even distance interval calibrated based on a location correspondence table between before and after the location calibrations, and an on-vehicle sensor data conversion step to convert an on-vehicle sensor data with an even time interval into an on-vehicle sensor data corresponding to location information expressed by distance columns based on a sampling number of an on-vehicle sensor data per an even distance interval after a location calibration.SELECTED DRAWING: Figure 1

Description

The present disclosure relates to a method, a system and a program for imparting position information to in-vehicle sensor data.

Conventionally, in the railway field, in order to detect changes in the environment around the track over time and to find obstacles such as buildings and vegetation that may interfere with the running of railway vehicles, etc. Train patrols have been carried out in which workers are on board the train to visually check the situation. In recent years, a technique has been proposed in which images around a track taken by a camera installed on a railroad vehicle traveling on the same line at different dates and times are compared.

In such a technique, among images shot at different dates and times, by comparing frames shot at the same place, a distance marker is used for the frame in order to detect changes in buildings, vegetation, etc. at the place. It is necessary to be provided with data indicating the position information that can be expressed from (see, for example, Patent Documents 1 and 2). It should be noted that a technique for imparting accurate kilometer information to waveform data such as track displacement and sway acceleration acquired by a railway vehicle or the like has already been proposed (see, for example, Patent Documents 3 and 4).

Japanese Unexamined Patent Publication No. 2015-141698 Japanese Unexamined Patent Publication No. 2017-001638 Japanese Unexamined Patent Publication No. 2018-0632225 Japanese Unexamined Patent Publication No. 2020-003247

However, in the above-mentioned conventional techniques (Patent Documents 1 and 2), when each frame of an image taken by a camera installed in a railroad vehicle cannot be accurately associated with a kilometer, it is possible to compare images at two different periods. , False detection occurs at the stage of detecting changes in buildings and vegetation around the track.

For example, as a method of adding a kilometer to each frame of an image taken by a camera installed in a railroad car, the latitude / longitude information acquired at the same time as the shooting using a satellite positioning system (GNSS) and a known known image created in advance. There is a method of collating with the kilometer-latitude-longitude information of, and a method of calculating the distance measured by the speed generator provided in the railroad vehicle. However, in the former case, if there is a tall building along the railway line or in the tunnel, the positioning accuracy will decrease, and in the latter case, the wheels will slip or slide, causing a positional deviation from the actual kilometer. An error will occur in the kilometer added to each frame of the image. Further, for example, a method of collating the feature points extracted by image analysis with the known kilometer-feature point information created in advance and adding the kilometer information can be considered, but in this method, similar scenes can be considered. In the case of a frame in which is continuous, an erroneous correspondence will occur and a positional deviation from the actual kilometer will occur. Further, due to these misalignments, erroneous detection occurs when the differences are detected by associating the frames of the images taken at two different times with each other.

Here, by solving the problems of the conventional technology and making it possible to add highly accurate position information to the data acquired by the in-vehicle sensor, the equipment in various parts of the vehicle's travel route can be accurately used as a ledger. A method, system, and position information addition method, system, and position information to the in-vehicle sensor data that can be collated and can detect the difference between the data acquired by the in-vehicle sensor at two different times and extract the deformed parts in various parts of the vehicle's travel route. The purpose is to provide a program.

Therefore, in the method of adding position information to the in-vehicle sensor data, the in-vehicle sensor data at equal time intervals, the waveform data at equal time intervals, and the equal time are used by the measuring device mounted on the vehicle traveling along the traveling path. A distance-time correspondence table is created based on the measurement process for measuring data related to the speed or distance of the interval and the data related to the speed or distance at the equal time interval, and the equal time is based on the distance-time correspondence table. The time-distance conversion step of converting the sampling number of the vehicle-mounted sensor data at intervals and the waveform data at equal time intervals into the sampling number of the vehicle-mounted sensor data at equal distance intervals and the waveform data at equal distance intervals, and the above. Perform waveform matching between the waveform data of the distance interval and the waveform data of the equidistant interval with known position information to create a post-correction-pre-correction position correspondence table, and based on the post-correction-pre-correction position correspondence table. Based on the position information imparting step of converting the sampling number of the in-vehicle sensor data of the equidistant interval into the sampling number of the in-vehicle sensor data of the position-corrected equidistant interval and the sampling number of the in-vehicle sensor data of the position-corrected equidistant interval. The vehicle-mounted sensor data conversion step of converting the vehicle-mounted sensor data at equal time intervals into vehicle-mounted sensor data corresponding to position information that can be represented from a distance marker is provided. Although it is preferable to use the methods described in Patent Documents 3 and 4 for waveform matching in the position information imparting step, other methods may be used.

The method of imparting position information to other in-vehicle sensor data further includes measurement of latitude and longitude data in the measurement step, and the time-distance conversion step includes equidistant latitude and longitude data of the latitude and longitude data. Pre-position correction-pre-correction position correspondence table (preliminary position correction-pre-correction position correspondence) by collating the equidistant interval latitude / longitude data with known position-latitude / longitude data. A table) is created, and based on the pre-position corrected-pre-correction position correspondence table, the sampling numbers of the in-vehicle sensor data at equidistant intervals and the waveform data at equidistant intervals are pre-position corrected equidistant, respectively. Further provided with a pre-position information imparting step of converting the sampling number of the vehicle-mounted sensor data of the interval and the waveform data of the pre-position corrected equidistant interval, and in the position information imparting step, the pre-position corrected equidistant interval Perform waveform matching between the waveform data and equidistant interval waveform data with known position information to create a post-correction-pre-correction position correspondence table, and based on the post-correction-pre-correction position correspondence table, the pre-position. The sampled numbers of the corrected equidistant in-vehicle sensor data are converted into the sampling numbers of the position-corrected equidistant in-vehicle sensor data.

Further, in another method of adding position information to the in-vehicle sensor data, in the time 1 and the time 2 in which a predetermined period elapses or is retroactive from the time 1 by the measuring device mounted on the vehicle traveling along the traveling path. , The measurement process at time 1 and time 2 for measuring the in-vehicle sensor data at equal time intervals, the waveform data at equal time intervals, and the data related to the speed or distance at equal time intervals, and the equal time at time 1 and time 2. A distance-time correspondence table is created based on the data related to the speed or distance of the interval, and based on the distance-time correspondence table, the sampling numbers of the in-vehicle sensor data at equal time intervals in the time 1 and the time 2 and the above-mentioned Time 1 to convert equidistant waveform data in time 1 and time 2 into sampling numbers of in-vehicle sensor data at equal distance intervals in time 1 and time 2 and waveform data at equal distance intervals in time 1 and time 2. And the time-distance conversion step in time 2 and the position correspondence table after position correction-before correction by performing waveform matching between the waveform data at equal distance intervals in time 1 and the waveform data at equal distance intervals for which position information is known. Based on the post-correction-pre-correction position correspondence table created, the sampling number of the in-vehicle sensor data at equal distance intervals converted from the in-vehicle sensor data at equal time intervals in the time 1 and the equal time in the time 1 The equidistant interval waveform data converted from the interval waveform data is used as the sampling number of the position-corrected equidistant in-vehicle sensor data in time 1 and the position-corrected equidistant interval waveform data in time 1, respectively. The position at the time 1 was used as a reference by performing waveform matching between the position information imparting step at the time 1 to be converted and the position-corrected equidistant interval waveform data at the time 1 and the equidistant interval waveform data at the time 2. A post-correction-pre-correction position correspondence table is created, and based on the post-correction-pre-correction position correspondence table, sampling of in-vehicle sensor data at equal distance intervals is converted from the in-vehicle sensor data at equal time intervals in the time period 2. The position information imparting step in time 2 for converting the number into the sampling number of the in-vehicle sensor data at the equidistant interval corresponding to the position in time 1, and the sampling number of the in-vehicle sensor data in the position-corrected equidistant interval in the time 1. And, based on the sampling number of the in-vehicle sensor data of the position in the time 1 and the equidistant interval of the corresponding time 2, the time in the time 1 and the time 2 is equal. The in-vehicle sensor data conversion process in time 1 and time 2 for converting the in-vehicle sensor data of the interval into the in-vehicle sensor data corresponding to the position information that can be represented from the distance markers in time 1 and time 2, and the distance marker in time 1 and time 2. The present invention includes a vehicle-mounted sensor data difference detection step of detecting a difference in the vehicle-mounted sensor data corresponding to the position information that can be expressed from the above and extracting the change points of the data in the time 1 and the time 2.

In the position information addition system to the in-vehicle sensor data, the in-vehicle sensor data at equal time intervals, the waveform data at equal time intervals, and the speed at equal time intervals are used by the measuring device mounted on the vehicle traveling along the traveling path. Alternatively, a distance-time correspondence table is created based on the measurement unit that measures the data related to the distance and the speed or the data related to the distance at the equal time interval, and the vehicle is mounted on the vehicle at the equal time interval based on the distance-time correspondence table. The time-distance conversion unit that converts the sampling number of the sensor data and the waveform data at equal time intervals into the sampling number of the in-vehicle sensor data at equal distance intervals and the waveform data at equal distance intervals, and the waveform at the equal distance intervals. Perform waveform matching between the data and the waveform data of the equidistant interval with known position information to create a post-correction-pre-correction position correspondence table, and based on the post-correction-pre-correction position correspondence table, the equidistant interval. Based on the distance axis data position information addition unit that converts the sampling number of the vehicle-mounted sensor data of the vehicle into the sampling number of the vehicle-mounted sensor data of the position-corrected equidistant interval, and the sampling number of the vehicle-mounted sensor data of the position-corrected equidistant interval. It is provided with an in-vehicle sensor data position information adding unit that converts the in-vehicle sensor data at equal time intervals into in-vehicle sensor data corresponding to the position information that can be represented from the distance marker.

In the position information addition system to other in-vehicle sensor data, in the time 1 and the time 2 in which a predetermined period elapses or is retroactive from the time 1 by the measuring device mounted on the vehicle traveling along the traveling path. A measuring unit that measures in-vehicle sensor data at equal time intervals, waveform data at equal time intervals, and data related to speed or distance at equal time intervals, and related to speed or distance at equal time intervals in time 1 and time 2. A distance-time correspondence table is created based on the data, and based on the distance-time correspondence table, the sampling numbers of the in-vehicle sensor data at equal time intervals in the time 1 and the time 2 and the sampling numbers in the time 1 and the time 2 etc. The time-distance conversion unit that converts the time-spaced waveform data into the sampling numbers of the in-vehicle sensor data at equal-distance intervals in time 1 and time 2, and the waveform data at equal-distance intervals in time 1 and time 2, and the time 1 above. A post-correction-pre-correction position correspondence table is created by performing waveform matching between the equidistant interval waveform data and the equidistant interval waveform data for which position information is known, and the post-correction-pre-correction position correspondence table is used. Based on this, the sampling number of the in-vehicle sensor data at the equal distance interval to which the in-vehicle sensor data at the equal time interval in the time 1 is converted, and the waveform data at the equal distance interval in which the waveform data at the equal time interval in the time 1 is converted. Are converted into the sampling number of the in-vehicle sensor data of the position-corrected equidistant interval in the period 1 and the waveform data of the position-corrected equidistant interval in the period 1, respectively, and the position-corrected equidistant interval in the period 1 is obtained. By performing waveform matching between the waveform data and the waveform data at equal distance intervals in the period 2, a post-correction-pre-correction position correspondence table is created based on the position in the period 1, and the post-correction-pre-correction position correspondence. Based on the table, the sampling number of the in-vehicle sensor data at the equal distance interval to which the in-vehicle sensor data at the equal time interval in the time 2 is converted is converted into the sampling number of the in-vehicle sensor data at the equivalent distance interval corresponding to the position in the time 1. Based on the distance axis data position information addition unit, the sampling number of the position-corrected equidistant vehicle-mounted sensor data in the time 1 and the sampling number of the vehicle-mounted sensor data corresponding to the position in the time 1 , The in-vehicle sensor data at equal time intervals in the time 1 and the time 2 is converted into the in-vehicle sensor data corresponding to the position information that can be represented by the distance marker in the time 1 and the time 2. The difference between the in-vehicle sensor data position information adding unit to be converted and the in-vehicle sensor data corresponding to the position information that can be represented from the distance markers in the time 1 and the time 2 is detected, and the change points of the data in the time 1 and the time 2 are extracted. It is equipped with an in-vehicle sensor data difference detection unit.

In the in-vehicle sensor data addition program, a computer is used to add position information to the in-vehicle sensor data, and the in-vehicle sensor data at equal time intervals is provided by a measuring device mounted on a vehicle traveling along the travel path. A distance-time correspondence table is created based on the waveform data at equal time intervals, the measuring unit that measures data related to speed or distance at equal time intervals, and the data related to speed or distance at equal time intervals, and the distance. -Based on the time correspondence table, the sampling numbers of the in-vehicle sensor data at equal time intervals and the waveform data at equal time intervals are converted into the sampling numbers of the in-vehicle sensor data at equal distance intervals and the waveform data at equal distance intervals. After position correction-after position correction-after position correction- Based on the pre-correction position correspondence table, the distance axis data position information addition unit that converts the sampling number of the vehicle-mounted sensor data at the same distance interval into the sampling number of the vehicle-mounted sensor data at the equidistant distance corrected, and the position-corrected Based on the sampling number of the in-vehicle sensor data at equal distance intervals, it functions as an in-vehicle sensor data position information adding unit that converts the in-vehicle sensor data at equal time intervals into in-vehicle sensor data corresponding to the position information that can be represented from the distance marker.

In the program for adding position information to other in-vehicle sensor data, a computer is used to add position information to the in-vehicle sensor data at time 1 and by a measuring device mounted on a vehicle traveling along the traveling path. A measuring unit that measures in-vehicle sensor data at equal time intervals, waveform data at equal time intervals, and data related to speed or distance at equal time intervals in the lapse of a predetermined period from time 1 or retroactive time 2, said time 1 And a distance-time correspondence table is created based on the data related to the speed or distance at equal time intervals in time 2, and based on the distance-time correspondence table, the in-vehicle sensor data at equal time intervals in time 1 and time 2 The sampling number and the waveform data of the equidistant intervals in the time 1 and the time 2 are the sampling numbers of the in-vehicle sensor data of the equidistant intervals in the time 1 and the time 2, and the waveform of the equidistant interval in the time 1 and the time 2. The time-distance conversion unit that converts to data creates a post-correction-pre-correction position correspondence table by performing waveform matching between the equidistant interval waveform data in time 1 and the equidistant interval waveform data for which position information is known. Based on the post-correction-pre-correction position correspondence table, the sampling number of the in-vehicle sensor data at the equal distance interval converted from the in-vehicle sensor data at the equal time interval in the time 1 and the equal time interval in the time 1 The waveform data of the equidistant interval to which the waveform data is converted is converted into the sampling number of the in-vehicle sensor data of the position-corrected equidistant interval in the period 1 and the waveform data of the position-corrected equidistant interval in the period 1, respectively. The position-corrected-pre-correction position correspondence table based on the position in time 1 is obtained by performing waveform matching between the position-corrected equidistant interval waveform data in time 1 and the equidistant interval waveform data in time 2. Based on the post-correction-pre-correction position correspondence table created, the sampling number of the equidistant in-vehicle sensor data converted from the equidistant in-vehicle sensor data in the time 2 has already corresponded to the position in the time 1. Distance axis data position information addition unit that converts to the sampling number of in-vehicle sensor data at equal distance intervals, position-corrected in-vehicle sensor data sampling number in the time 1 and the position in the time 1 Based on the sampling number of the in-vehicle sensor data at equal distance intervals in time 2, the in-vehicle sensor data at equal time intervals in time 1 and time 2 is obtained in time 1 and time. The in-vehicle sensor data position information adding unit that converts the in-vehicle sensor data corresponding to the position information that can be expressed from the distance marker in the period 2 and the position information that can be expressed from the distance indicator in the period 2 based on the position information of the period 1 and the period 1. It functions as an in-vehicle sensor data difference detection unit that detects the difference in the in-vehicle sensor data corresponding to the above and extracts the change points of the data in the time 1 and the time 2.

According to the present disclosure, it is possible to add highly accurate position information to the in-vehicle sensor data acquired by the in-vehicle sensor at equal time intervals, and to accurately collate the equipment in various parts of the vehicle's travel route with the ledger. It is possible to detect the difference in the data acquired by the in-vehicle sensor at two different times and extract the deformed parts in various parts of the traveling route of the vehicle.

It is a block diagram which shows the functional structure of the position information addition system to the vehicle-mounted sensor data in this embodiment. It is a figure explaining the outline of the operation of the position information addition system to the vehicle-mounted sensor data in this embodiment. It is a flowchart which shows the position information addition operation which partially simplified to the vehicle-mounted sensor data in this embodiment. It is a figure explaining the operation of each part of the sensing system in the position information addition operation partially simplified to the vehicle-mounted sensor data in this embodiment. It is a flowchart which shows the complete position information addition operation to the vehicle-mounted sensor data in this embodiment. It is a figure explaining the operation of each part of the sensing system in the operation of giving the complete position information to the vehicle-mounted sensor data in this embodiment. It is a flowchart which shows the operation which detects the difference of the vehicle-mounted sensor data acquired in two different time periods in this embodiment. It is a figure explaining the operation of each part of the sensing system in the operation of detecting the difference of the vehicle-mounted sensor data acquired in two different time periods in this embodiment.

Hereinafter, the present embodiment will be described in detail with reference to the drawings.

FIG. 1 is a block diagram showing a functional configuration of a position information imparting system to in-vehicle sensor data in the present embodiment, and FIG. 2 is a diagram illustrating an outline of an operation of the position information imparting system to in-vehicle sensor data in the present embodiment. Is. In FIG. 2, (a) is a diagram for explaining waveform matching, (b) is a diagram for explaining position correction of an image frame number (sampling number), and (c) is a diagram showing an image frame before position correction. (D) is a diagram showing an image frame after position correction.

In FIG. 1, reference numeral 10 denotes a sensing system as a position information imparting system to the vehicle-mounted sensor data according to the present embodiment, which is a kind of computer system used for imparting position information to the vehicle-mounted sensor data. The sensing system 10 is a computer system used to add highly accurate position information to data acquired by an in-vehicle sensor 11a mounted on a vehicle (not shown), and is a computer system such as a CPU, a magnetic disk, or a semiconductor memory. It is a computer system constructed in a computer including a storage device such as a keyboard, a mouse, an input device such as a touch panel, an output device such as a CRT, a liquid crystal display, and a printer, and a communication interface. The computer on which the sensing system 10 is built is, for example, a personal computer, a workstation, a server, a tablet computer, or the like, but any kind of computer that operates according to a program such as application software installed in the storage device. It may be a single computer, or it may be a group of computers in which a plurality of computers are connected so as to be able to communicate with each other via a network.

From the viewpoint of function, the sensing system 10 in the present embodiment includes a measurement unit 11, a time-distance conversion unit 12, a distance axis data approximate position information addition unit 13, a distance axis data position information addition unit 14, and an in-vehicle sensor. A data position information adding unit 15 and an in-vehicle sensor data difference detecting unit 16 are provided. The distance axis data approximate position information giving unit 13 may be omitted.

The measuring unit 11 is a portion mounted on a vehicle (not shown), and includes an in-vehicle sensor 11a, a waveform data measuring device 11b, a speed / distance data measuring device 11c, and a latitude / longitude measuring device 11d as measuring devices. And include. The latitude / longitude measuring device 11d may be omitted.

In the present embodiment, the vehicle is typically a railroad vehicle traveling on a railroad track, but of any type as long as it is a vehicle traveling along a track such as a railroad track or a track similar thereto. It may be a vehicle such as a railroad track, a new transportation system, a monorail, or a fixed route such as a bus used in a bus rapid transit (BRT) or a highway. It may be a traveling vehicle. Here, for convenience of explanation, the vehicle will be described as a railroad vehicle.

The in-vehicle sensor 11a is typically a camera capable of shooting a moving image such as a video camera installed in a vehicle such as a driver's cab of the vehicle, and is a device for capturing an image in front of the vehicle, but it is not always the case. The device is not limited to a device that captures a visible image, and is not limited to a device that captures an image of invisible information such as an infrared image or an ultrasonic image, or an optical sensor such as a laser scanner or LiDAR. It may be a device capable of sampling information such as point group data at equal time intervals with respect to the traveling direction of. The in-vehicle sensor 11a can be installed not only inside the vehicle but also outside the vehicle, for example, under the floor or on the roof of the vehicle. Here, for convenience of explanation, the vehicle-mounted sensor 11a will be described as a video camera installed in the driver's cab of the vehicle, and the vehicle-mounted sensor data as the data acquired by the vehicle-mounted sensor 11a will be described as image data.

Further, the waveform data measuring device 11b is a device that measures waveform data that correlates with the shape of a vehicle such as a track and a road in the longitudinal direction, and is, for example, a displacement in the vertical and horizontal directions of a rail. Measure waveform data such as track displacement, train sway (acceleration), and gyro signal (angular velocity, etc.).

Further, the speed / distance data measuring device 11c is a device for measuring vehicle speed data or equal distance pulse data, for example, an axle rotation speed detecting device using a speed generator, a rotary encoder, a railway vehicle such as a resolver, or the like. It is a speedometer that uses the rotational displacement detection device of the axle and the satellite positioning system (GNSS).

Further, the latitude / longitude measuring device 11d is a device that measures the latitude and longitude of a vehicle as equitime interval data, and is, for example, a device such as a vehicle navigation device that uses a satellite positioning system (GNSS).

The time-distance conversion unit 12 has a distance-time correspondence table creation unit 12a. Then, the measurement unit 11 acquires the measured data, the distance-time correspondence table creation unit 12a creates a distance-time correspondence table, and the data at equal time intervals is equal distance based on the distance-time correspondence table. Convert to interval data and output.

The distance axis data approximate position information giving unit 13 has a known position-latitude / longitude information storage unit 13a and a pre-position correction post-correction position correspondence table creation unit 13b. Then, the data output by the time-distance conversion unit 12 is acquired, collated with the known position-latitude / longitude information stored in the known position-latitude / longitude information storage unit 13a, and after the pre-position correction-before correction. The position correspondence table creation unit 13b creates a pre-position correction post-correction position correspondence table, and outputs pre-position-corrected equidistant interval data based on the pre-position correction post-correction position correspondence table.

The distance axis data position information adding unit 14 has a position information known equidistant interval waveform data storage unit 14a and a position correction post-correction position correspondence table creation unit 14b. Then, the data output by the distance axis data approximate position information giving unit 13 or the time-distance conversion unit 12 is acquired, and the equidistant information stored in the position information known equidistant interval waveform data storage unit 14a is known equidistant. After performing waveform matching with the interval waveform data, the post-correction-pre-correction position correspondence table creation unit 14b creates a post-correction-pre-correction position correspondence table, and based on the post-correction-pre-correction position correspondence table. Outputs position-corrected equidistant data.

The in-vehicle sensor data position information adding unit 15 acquires the data output by the distance axis data position information adding unit 14, adds position information to the data output by the in-vehicle sensor 11a, and is represented by a distance marker. The data of the vehicle-mounted sensor 11a corresponding to the position information is output. The "position information represented by the distance marker" is called "kilo" on railways and "kilopost" on expressways.

The vehicle-mounted sensor data difference detection unit 16 detects and outputs a difference in data acquired by the vehicle-mounted sensor 11a at two different periods.

Then, the sensing system 10 roughly performs position correction by distance sampling and waveform matching as follows.

In the distance sampling, time sampling data is obtained from data related to the speed (output data of the speed / distance data measuring device 11c or the latitude / longitude measuring device 11d) acquired in synchronization with the image which is the data acquired by the in-vehicle sensor 11a. Create a distance-time correspondence table, which is a list for distance sampling. According to the distance-time correspondence table, the image frame number (sampling number) of the image which is the image data acquired by the in-vehicle sensor 11a and the data of the waveform data measuring device 11b acquired in synchronization are sampled by distance, and the kilometer information is obtained. Is given.

Next, in the position correction by the waveform matching, as shown in FIG. 2A, the data of the waveform data measuring device 11b obtained by distance sampling and the accurate kilometer information acquired by the track inspection vehicle or the like are used. Performs position correction by waveform matching with waveform data provided with. Further, in conjunction with this, as shown in FIG. 2B, the position of the distance sampled image frame number (sampling number) is corrected. As a result, as shown in FIGS. 2 (c) and 2 (d), a position-corrected image frame can be obtained. By performing such waveform matching processing on two images taken at two different periods, it is possible to accurately associate the images taken at the two periods.

In addition, a temporary kilometer correction process is performed by collating the distance sampled latitude / longitude information with the known latitude / longitude-kilometer correspondence data between the distance sampling and the position correction by waveform matching. It is also possible to improve the accuracy of position correction.

Next, the operation of the sensing system 10 having the above configuration will be described. First, a partially simplified position information addition operation to the in-vehicle sensor data will be described.

FIG. 3 is a flowchart showing a partially simplified position information addition operation to the in-vehicle sensor data in the present embodiment, and FIG. 4 is a partially simplified position information addition operation to the in-vehicle sensor data in the present embodiment. It is a figure explaining the operation of each part of a sensing system.

First, when the sensing system 10 starts operating, the measuring unit 11 executes the measuring step in step S1. In this case, the vehicle-mounted sensor 11a is a video camera installed in the driver's cab of the vehicle, and a plurality of image frames are measured as vehicle-mounted sensor data at equal time intervals. Further, the waveform data measuring device 11b is a device that measures waveform data that correlates with the shape of the vehicle's travel path in the longitudinal direction, such as track displacement such as displacement of the rail in the vertical and horizontal directions, and train sway. It is assumed that waveform data such as (acceleration) and gyro signal (angular velocity, etc.) are measured at equal time intervals. Further, the speed / distance data measuring device 11c is a device for measuring vehicle speed data or equidistant pulse data, and measures data related to speed or distance at equal time intervals.

Next, in step S2, the time-distance conversion unit 12 executes a time-to-distance conversion step as a time-distance conversion step. In this step, the time-distance conversion unit 12 first acquires the data measured by the vehicle-mounted sensor 11a, the waveform data measuring device 11b, and the speed / distance data measuring device 11c from the measuring unit 11. Specifically, the sampling numbers of the vehicle-mounted sensor data at equal time intervals and the vehicle-mounted sensor data at equal time intervals are acquired from the vehicle-mounted sensor 11a, and the waveform data at equal time intervals are obtained from the waveform data measuring device 11b. It is acquired, and data related to the speed or distance at equal time intervals is acquired from the speed / distance data measuring device 11c. These data acquired from the measuring unit 11 are the same time axis data.

Subsequently, the distance-time correspondence table creation unit 12a represents the correspondence relationship between the time axis and the distance axis based on the data related to the speed or distance at equal time intervals acquired from the speed / distance data measuring device 11c-. Create a time correspondence table.

Subsequently, the time-distance conversion unit 12 has the sampling numbers of the vehicle-mounted sensor data acquired from the vehicle-mounted sensor 11a at equal time intervals and the equal time intervals acquired from the waveform data measuring device 11b based on the distance-time correspondence table. The waveform data is subjected to distance conversion processing, and is converted into a sampling number of in-vehicle sensor data at equal distance intervals and waveform data at equal distance intervals. The sampling number of the vehicle-mounted sensor data at equidistant intervals and the waveform data at equidistant intervals are the same distance axis data.

Next, in step S3, the distance axis data position information giving unit 14 executes the distance axis data position information giving step as the position information giving step. In this step, the distance axis data position information adding unit 14 first acquires the sampling number of the vehicle-mounted sensor data at equidistant intervals and the waveform data at equidistant intervals from the time-distance conversion unit 12. Then, waveform matching is performed between the equidistant interval waveform data and the equidistant interval waveform data with known position information stored in the equidistant interval waveform data storage unit 14a with known position information as accurate waveform data. Then, the post-correction-pre-correction position correspondence table creation unit 14b creates a post-correction-pre-correction position correspondence table as a correction distance comparison table.

Subsequently, the distance axis data position information adding unit 14 determines the sampling number of the equidistant interval in-vehicle sensor data acquired from the time-distance conversion unit 12 and the equidistant interval based on the post-correction-pre-correction position correspondence table. Position correction processing as kilometer correction processing is performed on the waveform data of. Thereby, as the position-corrected distance data, the sampling number of the vehicle-mounted sensor data of the position-corrected equidistant interval and the waveform data of the position-corrected equidistant interval can be obtained.

Next, in step S4, the vehicle-mounted sensor data position information adding unit 15 executes the position information imparting step to the vehicle-mounted sensor data as the vehicle-mounted data conversion step, and ends the process. In the process, the vehicle-mounted sensor data position information imparting unit 15 is measured by the time-distance conversion unit 12 based on the sampling number of the vehicle-mounted sensor data of the position-corrected equidistant interval acquired from the distance axis data position information imparting unit 14. The vehicle-mounted sensor data at equal time intervals acquired from the vehicle-mounted sensor 11a is converted into vehicle-mounted sensor data corresponding to the position information that can be represented by the distance marker, and this is output.

Next, the flowchart will be described.
Step S1 The measuring unit 11 executes the measuring step.
Step S2 The time-distance conversion unit 12 executes a time-to-distance conversion step.
Step S3 The distance axis data position information addition unit 14 executes the distance axis data position information addition step.
Step S4 The vehicle-mounted sensor data position information adding unit 15 executes the position information adding process to the vehicle-mounted sensor data, and ends the process.

In this way, by performing the position information giving operation in steps S1 to S4, the position information that can be represented from the high-precision distance marker is given to the in-vehicle sensor data such as the video data, so that it exists arbitrarily along the traveling route of the vehicle. It is possible to search for equipment in. For example, by associating a kilometer with an image frame of video data taken by a railroad vehicle, a maintenance vehicle, or the like, it is possible to specify an accurate image frame corresponding to the designated kilometer. This makes it possible to search for an accurate image of any equipment by collating it with a ledger or the like of the track equipment. In addition, the image corresponding to the orbital displacement can be searched by collating with the kilometer of the orbital displacement data.

Next, an unsimplified and complete position information addition operation to the in-vehicle sensor data will be described.

FIG. 5 is a flowchart showing a complete position information addition operation to the vehicle-mounted sensor data in the present embodiment, and FIG. 6 is an operation of each part of the sensing system in the complete position information addition operation to the vehicle-mounted sensor data in the present embodiment. It is a figure explaining.

First, when the sensing system 10 starts operating, the measuring unit 11 executes the measuring step in step S11. In this case, the vehicle-mounted sensor 11a is a video camera installed in the driver's cab of the vehicle, and a plurality of image frames are measured as vehicle-mounted sensor data at equal time intervals. Further, the waveform data measuring device 11b is a device that measures waveform data that correlates with the shape of the vehicle's travel path in the longitudinal direction, such as track displacement such as displacement of the rail in the vertical and horizontal directions, and train sway. It is assumed that waveform data such as (acceleration) and gyro signal (angular velocity, etc.) are measured at equal time intervals. Further, the speed / distance data measuring device 11c is a device for measuring vehicle speed data or equidistant pulse data, and measures data related to speed or distance at equal time intervals. Further, the latitude / longitude measuring device 11d is a device for measuring the latitude and longitude of the vehicle, and shall measure the latitude / longitude data.

Next, in step S12, the time-distance conversion unit 12 executes a time-to-distance conversion step. In this step, the time-distance conversion unit 12 first acquires the data measured by the in-vehicle sensor 11a, the waveform data measuring device 11b, the speed / distance data measuring device 11c, and the latitude / longitude measuring device 11d from the measuring unit 11. do. Specifically, the sampling numbers of the vehicle-mounted sensor data at equal time intervals and the vehicle-mounted sensor data at equal time intervals are acquired from the vehicle-mounted sensor 11a, and the waveform data at equal time intervals is obtained from the waveform data measuring device 11b. It is acquired, data related to speed and distance at equal time intervals are acquired from the speed / distance data measuring device 11c, and latitude / longitude data is acquired from the latitude / longitude measuring device 11d. These data acquired from the measuring unit 11 are the same time axis data.

Subsequently, the distance-time correspondence table creation unit 12a represents the correspondence relationship between the time axis and the distance axis based on the data related to the speed and distance at equal time intervals acquired from the speed / distance data measuring device 11c. Create a time correspondence table.

Subsequently, the time-distance conversion unit 12 determines the sampling number of the vehicle-mounted sensor data acquired from the vehicle-mounted sensor 11a and the waveform data of the equal-time interval acquired from the waveform data measuring device 11b based on the distance-time correspondence table. , And, the latitude and longitude data acquired from the latitude and longitude measuring device 11d is subjected to distance conversion processing, and the sampling number of the in-vehicle sensor data at equal distance intervals, the waveform data at equal distance intervals, and the latitude and longitude data at equal distance intervals are obtained. Convert. The sampling number of the vehicle-mounted sensor data at equidistant intervals, the waveform data at equidistant intervals, and the latitude / longitude data at equidistant intervals are the same distance axis data.

Next, in step S13, the distance axis data rough position information giving unit 13 executes a rough position information giving step of the distance axis data or a traveling path detection / discrimination step as a pre-position information giving step. In this step, the distance axis data approximate position information adding unit 13 first obtains a sampling number of in-vehicle sensor data at equidistant intervals, waveform data at equidistant intervals, and latitude / longitude data at equidistant intervals from the time-distance conversion unit 12. get. Then, the latitude / longitude data at the same distance interval is collated with the known position-latitude / longitude information stored in the known position-latitude / longitude information storage unit 13a as the correspondence data between the known kilometer and the latitude / longitude. Then, the post-correction-pre-correction position correspondence table creation unit 13b creates the pre-position correction post-correction position correspondence table as the first correction distance comparison table.

Subsequently, the distance axis data approximate position information adding unit 13 has a sampling number of equidistant interval in-vehicle sensor data acquired from the time-distance conversion unit 12 based on the pre-position correction-pre-correction position correspondence table, and the like. Pre-position correction processing as the first kilometer correction processing is performed on the waveform data of the distance interval. Thereby, as the pre-position corrected equidistant axis data, the sampling number of the vehicle-mounted sensor data of the pre-position corrected equidistant interval and the waveform data of the pre-position corrected equidistant interval can be obtained.

Next, in step S14, the distance axis data position information addition unit 14 executes the distance axis data position information addition step. In this step, the distance axis data position information addition unit 14 first receives pre-position-corrected equidistant interval in-vehicle sensor data, which is data that has undergone pre-position correction processing, from the distance-axis data approximate position information addition unit 13. The sampling number and the pre-position corrected equidistant interval waveform data are acquired. Then, the pre-position corrected equidistant interval waveform data and the equidistant interval waveform data with known position information stored in the equidistant interval waveform data storage unit 14a with known position information as waveform data with accurate position information. After waveform matching, the post-correction-pre-correction position correspondence table creation unit 14b creates a post-correction-pre-correction position correspondence table as a correction distance comparison table.

Subsequently, the distance axis data position information giving unit 14 is the pre-position corrected equidistant interval in-vehicle sensor data acquired from the distance axis data approximate position information giving unit 13 based on the post-correction-pre-correction position correspondence table. The sampling number and the pre-position corrected equidistant interval waveform data are subjected to position correction processing as a kilometer correction process. Thereby, as the position-corrected distance data, the sampling number of the vehicle-mounted sensor data of the position-corrected equidistant interval and the waveform data of the position-corrected equidistant interval can be obtained.

Next, in step S15, the vehicle-mounted sensor data position information imparting unit 15 executes the position information imparting step to the vehicle-mounted sensor data, and ends the process. In the process, the vehicle-mounted sensor data position information imparting unit 15 is measured by the time-distance conversion unit 12 based on the sampling number of the vehicle-mounted sensor data of the position-corrected equidistant interval acquired from the distance axis data position information imparting unit 14. The vehicle-mounted sensor data at equal time intervals acquired from the vehicle-mounted sensor 11a is converted into vehicle-mounted sensor data corresponding to the position information that can be represented by the distance marker, and this is output.

Next, the flowchart will be described.
Step S11 The measuring unit 11 executes the measuring step.
Step S12 The time-distance conversion unit 12 executes a time-to-distance conversion step.
Step S13 The distance axis data approximate position information addition unit 13 executes the approximate position information addition step of the distance axis data or the travel path detection / determination step.
Step S14 The distance axis data position information addition unit 14 executes the distance axis data position information addition step.
Step S15 The vehicle-mounted sensor data position information adding unit 15 executes the position information adding process to the vehicle-mounted sensor data, and ends the process.

In this way, by performing the position information giving operation in steps S11 to S15, the position information that can be represented from the extremely high-precision distance marker is given to the in-vehicle sensor data such as the video data, so that it exists along the traveling route of the vehicle. Highly accurate search of any equipment is possible. For example, by associating a kilometer with an image frame of video data taken by a railroad vehicle, a maintenance vehicle, or the like, it is possible to specify an accurate image frame corresponding to the designated kilometer. This makes it possible to search for an accurate image of any equipment by collating it with a ledger or the like of the track equipment. In addition, the image corresponding to the orbital displacement can be searched by collating with the kilometer of the orbital displacement data.

Next, the operation of detecting the difference between the in-vehicle sensor data acquired at two different periods will be described. The position information addition operation to the in-vehicle sensor data included in the operation may be a partially simplified position information addition operation or an unsimplified, complete position information addition operation. Although it is good, here, for convenience of explanation, it is assumed that it is a complete position information giving operation.

FIG. 7 is a flowchart showing an operation of detecting a difference in in-vehicle sensor data acquired at two different periods in the present embodiment, and FIG. 8 shows a difference in in-vehicle sensor data acquired in two different periods in the present embodiment. It is a figure explaining the operation of each part of a sensing system in the operation.

Here, the sensing system 10 performs a position information addition operation to the in-vehicle sensor data acquired in the first period, the period 1, and then in the second period after a predetermined period has elapsed or retroactively from the period 1. It will be described as assuming that the difference between the in-vehicle sensor data acquired in the two periods is detected after the position information is added to the in-vehicle sensor data acquired in the certain period 2.

First, at time 1, when the sensing system 10 starts operating, the measuring unit 11 executes the measuring step in step S11-1. Since the operation in step S11-1 is the same as the operation in step S11 shown in FIG. 5, the description thereof will be omitted.

Next, in step S12-1, the time-distance conversion unit 12 executes a time-to-distance conversion step. Since the operation in step S12-1 is the same as the operation in step S12 shown in FIG. 5, the description thereof will be omitted.

Next, in step S13-1, the distance axis data approximate position information addition unit 13 executes a schematic position information addition step or a travel path detection / determination step of the distance axis data. Since the operation in step S13-1 is the same as the operation in step S13 shown in FIG. 5, the description thereof will be omitted.

Next, in step S14-1, the distance axis data position information addition unit 14 executes the distance axis data position information addition step. As a result, it is possible to obtain waveform data at equidistant intervals with position correction at time 1. The position-corrected equidistant interval waveform data at time 1 is stored in a storage device (not shown). Since the operation in step S14-1 is the same as the operation in step S14 shown in FIG. 5, the description thereof will be omitted.

Next, in step S15-1, the vehicle-mounted sensor data position information adding unit 15 executes the position information imparting step to the vehicle-mounted sensor data, and ends the process at the time 1. As a result, it is possible to output the in-vehicle sensor data corresponding to the position information that can be represented from the distance marker at the time 1. The vehicle-mounted sensor data corresponding to the position information that can be represented from the distance marker in the time 1 is stored in a storage device (not shown). Since the operation in step S15-1 is the same as the operation in step S15 shown in FIG. 5, the description thereof will be omitted.

Next, when the sensing system 10 starts operating at the time when a predetermined period elapses or retroactively from the time 1, the measurement unit 11 executes the measurement step in step S11-2. Since the operation in step S11-2 is the same as the operation in step S11 shown in FIG. 5, the description thereof will be omitted.

Next, in step S12-2, the time-distance conversion unit 12 executes a time-to-distance conversion step. Since the operation in step S12-2 is the same as the operation in step S12 shown in FIG. 5, the description thereof will be omitted.

Next, in step S13-2, the distance axis data approximate position information addition unit 13 executes a schematic position information addition step or a travel path detection / determination step of the distance axis data. Since the operation in step S13-2 is the same as the operation in step S13 shown in FIG. 5, the description thereof will be omitted.

Next, in step S14-2, the distance axis data position information giving unit 14 executes the position information giving step of the distance axis data of the time 2 based on the position information of the time 1. In this step, the distance axis data position information addition unit 14 first receives pre-position-corrected equidistant interval in-vehicle sensor data, which is data that has undergone pre-position correction processing, from the distance-axis data approximate position information addition unit 13. The sampling number and the pre-position corrected equidistant interval waveform data are acquired. Further, the distance axis data position information adding unit 14 acquires the position-corrected equidistant interval waveform data obtained in step S14-1 and stored in a storage device (not shown) at time 1. Then, this time, that is, waveform matching is performed between the waveform data of the pre-position corrected equidistant interval in the period 2 and the waveform data of the position-corrected equidistant interval in the period 1, and the corrected distance comparison table is used in the period 1. Create a post-correction-pre-correction position correspondence table based on the position.

Subsequently, the distance axis data position information giving unit 14 is the pre-position corrected equidistant interval in-vehicle sensor data acquired from the distance axis data approximate position information giving unit 13 based on the post-correction-pre-correction position correspondence table. Position correspondence processing is performed on the sampling number. As a result, as the position-corresponding distance data, the sampling number of the vehicle-mounted sensor data at the equidistant interval between the position in the time 1 and the time 2 can be obtained.

Next, in step S15-2, the vehicle-mounted sensor data position information adding unit 15 executes a position information adding step to the vehicle-mounted sensor data of the time 2 based on the position information of the time 1. In the process, the in-vehicle sensor data position information adding unit 15 is based on the sampling number of the in-vehicle sensor data of the position corrected equidistant interval corresponding to the position in the time 1 acquired from the distance axis data position information adding unit 14. The distance conversion unit 12 converts the vehicle-mounted sensor data at equal time intervals acquired from the vehicle-mounted sensor 11a of the measurement unit 11 into vehicle-mounted sensor data corresponding to the position information that can be represented from the distance marker in the time period 2 based on the position information of the time period 1. can do.

Next, in step S16, the vehicle-mounted sensor data difference detection unit 16 executes the vehicle-mounted sensor data difference detection step for two periods, and ends the process. In the process, the vehicle-mounted sensor data difference detection unit 16 acquires vehicle-mounted sensor data corresponding to the position information that can be represented by the distance marker at the time 2 from the vehicle-mounted sensor data position information imparting unit 15, and is obtained in step S15-1. The in-vehicle sensor data corresponding to the position information that can be represented from the distance marker at the time 1 stored in the storage device (not shown) is acquired. Then, the difference between the in-vehicle sensor data corresponding to the position information represented by the distance marker in the time 2 based on the position information in the time 1 and the in-vehicle sensor data corresponding to the position information represented by the distance marker in the time 1 is detected and 2 Extract the change points of the data at the time and output this.

Next, the flowchart will be described.
Step S11-1 In time 1, the measuring unit 11 executes the measuring step.
Step S12-1 The time-distance conversion unit 12 executes a time-to-distance conversion step.
Step S13-1 The distance axis data approximate position information addition unit 13 executes the approximate position information addition step of the distance axis data or the travel path detection / determination step.
Step S14-1 The distance axis data position information addition unit 14 executes the distance axis data position information addition step.
Step S15-1 The vehicle-mounted sensor data position information adding unit 15 executes a position information adding step to the vehicle-mounted sensor data.
Step S11-2 At time 2, the measuring unit 11 executes the measuring step.
Step S12-2 The time-distance conversion unit 12 executes a time-to-distance conversion step.
Step S13-2 The distance axis data approximate position information addition unit 13 executes the outline position information addition step of the distance axis data or the travel path detection / discrimination step.
Step S14-2 The distance axis data position information adding unit 14 executes the position information giving step of the distance axis data of the time 2 based on the position information of the time 1.
Step S15-2 The vehicle-mounted sensor data position information adding unit 15 executes the position information adding step to the vehicle-mounted sensor data of the time 2 based on the position information of the time 1.
Step S16 The vehicle-mounted sensor data difference detection unit 16 executes the vehicle-mounted sensor data difference detection step of two periods, and ends the process.

In this way, by performing the operation of detecting the difference between the in-vehicle sensor data acquired in the two different periods of steps S11-1 to S16, the in-vehicle sensor data such as the video data acquired in the different two periods is highly accurate. Since the position information that can be expressed from the distance marker is given, it becomes possible to calculate the difference between the in-vehicle sensor data of the two periods having the same position information and extract the deformed part. For example, by applying position correction processing by waveform matching between waveform data acquired in synchronization with image frames of video data taken by railroad cars or maintenance cars at two different times, the same kilometer or so. Images taken at (location) can be extracted. In this way, by comparing the images of two different periods, it can be applied to the difference detection for detecting the deformed part of both.

As described above, the method of adding the position information to the in-vehicle sensor data in the present embodiment is the in-vehicle sensor data at equal time intervals and the waveform data at equal time intervals by the measuring device mounted on the vehicle traveling along the travel path. , And, a distance-time correspondence table is created based on the measurement process that measures the data related to the speed or distance at equal time intervals, and the data related to the speed or distance at equal time intervals, and based on the distance-time correspondence table. Time-distance conversion step of converting the sampling number of the in-vehicle sensor data at equal time intervals and the waveform data of the equal time interval into the sampling number of the in-vehicle sensor data of the equal distance interval and the waveform data of the equal distance interval. Perform waveform matching between the waveform data at equal distance intervals and the waveform data at equal distance intervals for which position information is known to create a post-correction-pre-correction position correspondence table, and based on the post-correction-pre-correction position correspondence table. Based on the position information giving process that converts the sampling number of the in-vehicle sensor data of the equidistant interval into the sampling number of the in-vehicle sensor data of the position-corrected equidistant interval, and the sampling number of the in-vehicle sensor data of the position-corrected equidistant interval, etc. It includes an in-vehicle sensor data conversion step of converting in-vehicle sensor data at time intervals into in-vehicle sensor data corresponding to position information that can be represented from a distance marker. As a result, highly accurate position information can be added to the data acquired by the in-vehicle sensor 11a, so that the equipment in various parts of the traveling route of the vehicle can be accurately collated with the ledger.

The measurement process also includes the measurement of latitude and longitude data, and the time-distance conversion process includes the conversion of equidistant interval latitude and longitude data to equidistant interval latitude and longitude data and known positions. -After pre-position correction by collating with latitude and longitude data-Create a pre-correction position correspondence table, and based on the pre-position correction-pre-correction position correspondence table, the sampling number of equidistant interval in-vehicle sensor data, and Further provided with a pre-position information imparting step of converting equidistant equidistant waveform data into pre-position corrected equidistant in-vehicle sensor data sampling numbers and pre-position corrected equidistant interval waveform data, respectively. In the information imparting process, waveform matching is performed between pre-position-corrected equidistant interval waveform data and equidistant interval waveform data with known position information to create a post-correction-pre-correction position correspondence table and position correction. After-Converts the pre-corrected equidistant interval in-vehicle sensor data sampling numbers to the position-corrected equidistant interval in-vehicle sensor data sampling numbers based on the pre-correction equidistant position correspondence table. As a result, highly accurate position information can be added to the data acquired by the in-vehicle sensor 11a, so that the equipment in various parts of the traveling route of the vehicle can be more accurately collated with the ledger.

Further, in the method of adding position information to other in-vehicle sensor data in the present embodiment, a predetermined period elapses or retroactively from time 1 and time 1 by a measuring device mounted on a vehicle traveling along the traveling path. The measurement steps in Time 1 and Time 2 and the measurement steps in Time 1 and Time 2 for measuring in-vehicle sensor data at equal time intervals, waveform data at equal time intervals, and data related to speed or distance at equal time intervals, and Time 1 and Time 2 Create a distance-time correspondence table based on the data related to the speed or distance at equal time intervals in, and based on the distance-time correspondence table, the sampling numbers of the in-vehicle sensor data at equal time intervals in time 1 and time 2, and Time 1 to convert equidistant waveform data in time 1 and time 2 into sampling numbers of in-vehicle sensor data at equal distance intervals in time 1 and time 2 and waveform data at equal distance intervals in time 1 and time 2. And, the time-distance conversion process in time 2 and the waveform matching between the waveform data at equal distance intervals in time 1 and the waveform data at equal distance intervals whose position information is known are performed to create a position correspondence table after position correction-before correction. Then, based on the post-correction-pre-correction position correspondence table, the sampling number of the in-vehicle sensor data at equal distance intervals converted from the in-vehicle sensor data at equal time intervals in time 1, and the waveform data at equal time intervals in time 1. The converted equidistant interval waveform data is converted into the sample number of the position-corrected equidistant interval in-vehicle sensor data in time 1 and the position-corrected equidistant interval waveform data in time 1, respectively. After position correction based on the position in time 1 by performing waveform matching between the position information giving process in A position correspondence table is created, and the sampling number of the in-vehicle sensor data at equal distance intervals converted from the in-vehicle sensor data at equal time intervals in time 2 is set as the position in time 1 based on the position correspondence table after position correction-before correction. Corresponding to the position information giving process in the time 2 to be converted into the sampling number of the in-vehicle sensor data of the corresponding equidistant interval, the sampling number of the in-vehicle sensor data of the position-corrected equidistant interval in the time 1, and the position in the time 1. Based on the sampling number of the in-vehicle sensor data at the same distance interval, the in-vehicle sensor data at the same time interval in the time 1 and the time 2 is transferred to the time 1 and the time 2. The difference between the in-vehicle sensor data conversion process at time 1 and time 2 to convert to the in-vehicle sensor data corresponding to the position information that can be expressed from the distance marker and the in-vehicle sensor data corresponding to the position information that can be expressed from the distance marker in time 1 and time 2. It is provided with a difference detection step of in-vehicle sensor data for detecting and extracting the change points of the data in the time 1 and the time 2. As a result, it is possible to detect the difference in the data acquired by the in-vehicle sensor 11a at two different periods and extract the deformed parts in various parts of the traveling path of the vehicle.

It should be noted that the disclosure herein describes features relating to suitable and exemplary embodiments. Various other embodiments, modifications and modifications within the scope and purpose of the claims attached herein can be naturally conceived by those skilled in the art by reviewing the disclosure of the present specification. be.

The present disclosure can be applied to methods, systems and programs for imparting position information to in-vehicle sensor data.

10 Sensing system 11 Measuring unit 11a In-vehicle sensor 11b Waveform data measuring device 11c Speed / distance data measuring device 11d Latitude / longitude measuring device 12 Time distance conversion unit 14 Distance axis data position information adding unit 15 In-vehicle sensor data position information adding unit 16 In-vehicle sensor Data difference detector

Claims (7)

A measurement process that measures in-vehicle sensor data at equal time intervals, waveform data at equal time intervals, and data related to speed or distance at equal time intervals by a measuring device mounted on a vehicle traveling along the travel path.
A distance-time correspondence table is created based on the data related to the speed or distance at the equidistant interval, and the sampling number of the in-vehicle sensor data at the equidistant interval and the equidistant interval based on the distance-time correspondence table. A time-distance conversion process for converting waveform data into equidistant interval in-vehicle sensor data sampling numbers and equidistant interval waveform data.
Perform waveform matching between the equidistant interval waveform data and equidistant interval waveform data for which position information is known to create a post-correction-pre-correction position correspondence table, and use the post-correction-pre-correction position correspondence table. Based on this, the position information giving step of converting the sampling number of the equidistant equidistant in-vehicle sensor data into the sampling number of the position-corrected equidistant in-vehicle sensor data,
An in-vehicle sensor data conversion process that converts the equidistant in-vehicle sensor data into in-vehicle sensor data corresponding to the position information that can be represented from the distance marker based on the sampling number of the in-vehicle sensor data with the equidistant interval corrected.
A method of imparting position information to in-vehicle sensor data, which comprises. The measurement step includes measurement of latitude / longitude data.
The time-distance conversion step comprises converting the latitude / longitude data to equidistant interval latitude / longitude data.
The equidistant interval latitude / longitude data is collated with the known position-latitude / longitude data to create a pre-position corrected-pre-correction position correspondence table, and based on the pre-position corrected-pre-correction position correspondence table. The sampling numbers of the equidistant in-vehicle sensor data and the equidistant waveform data are the pre-position corrected equidistant interval in-vehicle sensor data sampling numbers and the pre-position corrected equidistant interval waveforms, respectively. Further equipped with a pre-position information addition process to convert to data,
In the position information imparting step, waveform matching is performed between the pre-position corrected equidistant interval waveform data and the equidistant interval waveform data whose position information is known, and a post-correction-pre-correction position correspondence table is created. According to claim 1, the sampling number of the pre-position corrected equidistant interval in-vehicle sensor data is converted into the sampling number of the position-corrected equidistant in-vehicle sensor data based on the position-corrected-pre-correction position correspondence table. A method of adding position information to the described in-vehicle sensor data. In-vehicle sensor data at equal time intervals and waveform data at equal time intervals at time 1 and time 2 when a predetermined period has elapsed or retroactively from time 1 by a measuring device mounted on a vehicle traveling along the traveling path. , And the measurement steps at time 1 and time 2 for measuring data related to speed or distance at equal time intervals.
A distance-time correspondence table is created based on the data related to the speed or distance at equal time intervals in the time 1 and time 2, and the vehicle is mounted on the vehicle at equal time intervals in the time 1 and time 2 based on the distance-time correspondence table. The sampling number of the sensor data and the equidistant waveform data in the time 1 and the time 2 are the sampling numbers of the in-vehicle sensor data at the time 1 and the time 2 and the equidistant distance in the time 1 and the time 2. Time-distance conversion steps in time 1 and time 2 to convert to interval waveform data,
Perform waveform matching between the equidistant interval waveform data in time 1 and the equidistant interval waveform data for which position information is known to create a post-correction-pre-correction position correspondence table, and the post-correction-pre-correction position. Based on the correspondence table, the sampling number of the equidistant in-vehicle sensor data in which the equidistant in-vehicle sensor data in the time 1 is converted, and the equidistant interval in which the equidistant waveform data in the time 1 is converted. The waveform data of is converted into the sampling number of the in-vehicle sensor data of the position-corrected equidistant interval in the time 1 and the waveform data of the position-corrected equidistant interval in the time 1, respectively, and the position information giving step in the time 1.
Perform waveform matching between the position-corrected equidistant interval waveform data in time 1 and the equidistant interval waveform data in time 2, and create a post-correction-pre-correction position correspondence table based on the position in time 1. Then, based on the post-correction-pre-correction position correspondence table, the sampling number of the equidistant in-vehicle sensor data converted from the equidistant in-vehicle sensor data in the time 2 is already corresponded to the position in the time 1. The position information giving process at the time 2 of converting to the sampling number of the in-vehicle sensor data of the equidistant interval, and
Based on the sampling number of the position-corrected equidistant vehicle-mounted sensor data in the time 1 and the sampling number of the position-corresponding time 2 equidistant vehicle-mounted sensor data in the time 1 and the time. The in-vehicle sensor data conversion process in time 1 and time 2 for converting the in-vehicle sensor data at equal time intervals in 2 into the in-vehicle sensor data corresponding to the position information represented by the distance markers in time 1 and time 2.
The in-vehicle sensor data difference detection step of detecting the difference in the in-vehicle sensor data corresponding to the position information represented by the distance marker in the time 1 and the time 2 and extracting the change part of the data in the time 1 and the time 2.
A method of imparting position information to in-vehicle sensor data, which comprises. A measuring unit that measures in-vehicle sensor data at equal time intervals, waveform data at equal time intervals, and data related to speed or distance at equal time intervals by a measuring device mounted on a vehicle traveling along the travel path.
A distance-time correspondence table is created based on the data related to the speed or distance at the equidistant interval, and the sampling number of the in-vehicle sensor data at the equidistant interval and the equidistant interval based on the distance-time correspondence table. A sampling number of equidistant in-vehicle sensor data and a time-distance conversion unit that converts equidistant wave data into equidistant interval waveform data.
Perform waveform matching between the equidistant interval waveform data and equidistant interval waveform data for which position information is known to create a post-correction-pre-correction position correspondence table, and use the post-correction-pre-correction position correspondence table. Based on this, a distance axis data position information adder that converts the sampling number of the equidistant equidistant in-vehicle sensor data into a position-corrected equidistant in-vehicle sensor data sampling number.
An in-vehicle sensor data position information adding unit that converts the equidistant in-vehicle sensor data into in-vehicle sensor data corresponding to the position information that can be represented from the distance marker based on the sampling number of the in-vehicle sensor data with the equidistant interval corrected.
A system for adding location information to in-vehicle sensor data. In-vehicle sensor data at equal time intervals and waveform data at equal time intervals at time 1 and time 2 when a predetermined period has elapsed or retroactively from time 1 by a measuring device mounted on a vehicle traveling along the traveling path. , And a measuring unit that measures data related to speed or distance at equal time intervals.
A distance-time correspondence table is created based on the data related to the speed or distance at equal time intervals in the time 1 and time 2, and the vehicle is mounted on the vehicle at equal time intervals in the time 1 and time 2 based on the distance-time correspondence table. The sampling number of the sensor data and the equidistant waveform data in the time 1 and the time 2 are the sampling numbers of the in-vehicle sensor data at the time 1 and the time 2 and the equidistant distance in the time 1 and the time 2. A time-distance converter that converts interval waveform data,
Equidistant interval waveform data in time 1 and equidistant interval waveform data with known position information are matched to create a post-correction-pre-correction position correspondence table, and the post-correction-pre-correction position. Based on the correspondence table, the sampling number of the equidistant in-vehicle sensor data to which the equidistant in-vehicle sensor data in the time 1 is converted, and the equidistant interval in which the equidistant waveform data in the time 1 is converted. The waveform data of is converted into the sampling number of the position-corrected equidistant interval in-vehicle sensor data in time 1 and the position-corrected equidistant interval waveform data in time 1, respectively, and the position-corrected equidistant interval in time 1 etc. Perform waveform matching between the equidistant interval waveform data and equidistant interval waveform data at time 2 to create a post-correction-pre-correction position correspondence table based on the position at time 1 and then post-correction-correction. Based on the front position correspondence table, the sampling number of the equidistant equidistant in-vehicle sensor data converted from the equidistant in-vehicle sensor data in the time 2 is used as the sampling number of the equidistant in-vehicle sensor data corresponding to the position in the time 1. Equidistant axis data position information adder to convert to number,
Based on the sampling number of the in-vehicle sensor data of the position-corrected equidistant interval in the time 1 and the sampling number of the in-vehicle sensor data of the equidistant interval corresponding to the position in the time 1, the equal time in the time 1 and the time 2 An in-vehicle sensor data position information adding unit that converts the in-vehicle sensor data of the interval into in-vehicle sensor data corresponding to the position information that can be expressed from the distance markers in the period 1 and the period 2.
An in-vehicle sensor data difference detection unit that detects the difference in the in-vehicle sensor data corresponding to the position information that can be expressed from the distance marker in the time 1 and the time 2 and extracts the changed part of the data in the time 1 and the time 2.
A system for adding location information to in-vehicle sensor data. A computer to add location information to in-vehicle sensor data,
A measuring unit that measures in-vehicle sensor data at equal time intervals, waveform data at equal time intervals, and data related to speed or distance at equal time intervals by a measuring device mounted on a vehicle traveling along the travel path.
A distance-time correspondence table is created based on the data related to the speed or distance at the equal time interval, and the sampling number of the in-vehicle sensor data at the equal time interval and the equal time interval based on the distance-time correspondence table. A time-distance converter that converts waveform data into sampling numbers for in-vehicle sensor data at equal distance intervals and waveform data at equal distance intervals.
Perform waveform matching between the equidistant interval waveform data and equidistant interval waveform data for which position information is known to create a post-correction-pre-correction position correspondence table, and use the post-correction-pre-correction position correspondence table. Based on this, the distance axis data position information addition unit that converts the sampling number of the equidistant equidistant in-vehicle sensor data into the sampling number of the position-corrected equidistant in-vehicle sensor data, and
As an in-vehicle sensor data position information adding unit that converts the in-vehicle sensor data at equal time intervals into in-vehicle sensor data corresponding to the position information that can be represented from the distance marker, based on the sampling number of the in-vehicle sensor data at the equal distance interval that has been corrected. A program for adding location information to in-vehicle sensor data, which is characterized by functioning. A computer to add location information to in-vehicle sensor data,
In-vehicle sensor data at equal time intervals and waveform data at equal time intervals at time 1 and time 2 when a predetermined period has elapsed or retroactively from time 1 by a measuring device mounted on a vehicle traveling along the traveling path. , And a measuring unit that measures data related to speed or distance at equal time intervals,
A distance-time correspondence table is created based on the data related to the equidistant speed or distance in the time 1 and the time 2, and the vehicle is mounted on the vehicle at the equidistant time in the time 1 and the time 2 based on the distance-time correspondence table. The sampling numbers of the sensor data and the equidistant waveform data in the time 1 and the time 2 are the sampling numbers of the in-vehicle sensor data at the time 1 and the time 2 equidistant, and the equidistant in the time 1 and the time 2. Time-distance converter, which converts to equidistant waveform data,
Equidistant interval waveform data in time 1 and equidistant interval waveform data with known position information are matched to create a post-correction-pre-correction position correspondence table, and the post-correction-pre-correction position. Based on the correspondence table, the sampling number of the equidistant in-vehicle sensor data to which the equidistant in-vehicle sensor data in the time 1 is converted, and the equidistant interval in which the equidistant waveform data in the time 1 is converted. The waveform data of is converted into the sampling number of the in-vehicle sensor data of the position-corrected equidistant interval in time 1 and the waveform data of the position-corrected equidistant interval in time 1, respectively, and the position is corrected in the time 1 and the like. Perform waveform matching between the equidistant interval waveform data and equidistant interval waveform data at time 2 to create a post-correction-pre-correction position correspondence table based on the position at time 1 and then post-correction-correction. Based on the front position correspondence table, the sampling number of the equidistant equidistant in-vehicle sensor data converted from the equidistant in-vehicle sensor data in the time 2 is used as the sampling number of the equidistant in-vehicle sensor data corresponding to the position in the time 1. Equidistant axis data position information adder to convert to number,
Based on the sampling number of the position-corrected equidistant vehicle-mounted sensor data in the time 1 and the sampling number of the position-corresponding time 2 equidistant vehicle-mounted sensor data in the time 1 and the time. The in-vehicle sensor data position information addition unit that converts the in-vehicle sensor data at equal time intervals in 2 into the in-vehicle sensor data corresponding to the position information that can be represented from the distance markers in time 1 and time 2, and
It functions as an in-vehicle sensor data difference detection unit that detects the difference in the in-vehicle sensor data corresponding to the position information that can be expressed from the distance marker in the time 1 and the time 2 and extracts the changed part of the data in the time 1 and the time 2. A program for adding location information to in-vehicle sensor data.

Images