JP2019218035A - Obstacle detection system and obstacle detection method - Google Patents

Abstract

To improve detection performance of an obstacle under a double-track environment where a plurality of tracks are adjacent to each other.SOLUTION: An obstacle detection system 1 for detecting an obstacle under a double-track environment including a first track and a second track that are adjacent to each other, includes an external sensor (side monitoring camera 110) that is installed on a first train (train 100 traveling on a track) traveling on the first track at an angle to an advancing axis of the train and has a sensor region including a region over the second track, an obstacle detection device 120 for detecting the presence/absence of an obstacle on the second track by using sensor data of the side monitoring camera 110, and a communication unit 130 for transmitting and receiving data or signals between the obstacle detection device 120 and the outside.SELECTED DRAWING: Figure 1

Description

  The present invention relates to an obstacle detection system and an obstacle detection method, and is suitably applied to an obstacle detection system and an obstacle detection method for detecting an obstacle in a double track environment.

  Trains running on the track cannot avoid obstacles on the track by steering, so it is important to recognize the presence of obstacles on the track to improve train safety and operability. is there.

  Conventionally, in the case of a manned driving system in which a driver drives a train, the presence or absence of an obstacle on a track is visually confirmed by the driver. On the other hand, in the case of unmanned driving systems where there is no driver, measures are generally taken to provide a dedicated track and cut off other traffic to suppress the occurrence of obstacles on the track, and safety measures in open tracks No unmanned operation system has been realized.

  In view of such a background, in recent years, in order to realize an unmanned operation system on an open track, an external sensor is mounted on the train, and based on information acquired from the external sensor, whether there is an obstacle on the track in front of the train. Systems have been devised to detect the For example, according to the on-orbit obstacle detection system disclosed in Patent Document 1, a forward monitoring camera is installed in front of a train, and after detecting a track area in an image obtained from the camera, the vehicle travels in the same place in the past. The presence or absence of an obstacle on the orbit is detected by comparing the recorded image with the recorded image.

JP-A-2006-052849

  However, when trying to detect the presence or absence of an obstacle on the orbit in an open orbit by the conventional technology, there are the following problems.

  First, when constructing an obstacle detection function for trains, a technology that accurately detects obstacles farther away than a car is required, but it was difficult for conventional technology to meet such demands . Specifically, a train requires a longer braking distance than a car. Specifically, for example, the braking distance of a vehicle traveling at a speed of 100 km / h is about 100 m, whereas the braking distance of a train traveling at the same speed is about 400 m. The difference in braking performance depends on the difference in the brake structure between the car and the train.However, in the case of a train, it is necessary to transport the standing passenger safely, so rapid deceleration is not possible and a long braking distance is required. Become.

  In recent years, in the automotive field, external sensors such as cameras, millimeter wave radars, and LiDARs (Light Detection and Ranging) have been developed for driving assistance or autonomous driving. The maximum length is about 200 m. Therefore, it is difficult to accurately detect distant obstacles simply by using the obstacle detection function using an external sensor for automobiles in the train driving environment, and the technology required for the obstacle detection function for trains is difficult. We cannot meet the standard.

  Another problem is that when a train runs on a curve or slope, the track is shielded by a structure adjacent to the track, resulting in blind spots from an external sensor mounted on the vehicle. There is a problem that there is a traveling section where the obstacle detection cannot be sufficiently performed.

  The present invention has been made in view of the above points, and aims to propose an obstacle detection system and an obstacle detection method that improve obstacle detection performance in a double track environment where a plurality of tracks are adjacent. is there.

  In order to solve such a problem, the present invention provides an obstacle detection system that detects an obstacle in a double track environment including an adjacent first trajectory and a second trajectory. This obstacle detection system is installed on a first train traveling on the first track at an angle to a traveling axis of the first train, and an external field sensor including the second track on a sensor area. And, using the sensor data of the external sensor, transmitting / receiving data or a signal between the obstacle detection device and the outside, and an obstacle detection device that detects the presence or absence of an obstacle on the second track. And a communication unit.

  Further, in order to solve this problem, in the present invention, an obstacle detection method by an obstacle detection system that detects an obstacle on an adjacent track in a double track environment including an adjacent first track and a second track. Provided. In this obstacle detection method, the obstacle detection system is installed on the first train traveling on the first track at an angle to the traveling axis of the first train, and An external field sensor including in the sensor area, an obstacle detection device that detects the presence or absence of an obstacle on the second track using sensor data of the external field sensor, and between the obstacle detection device and the outside. A communication unit for transmitting and receiving data or signals.

  ADVANTAGE OF THE INVENTION According to this invention, the detection performance of an obstacle can be improved in the double track environment with which several tracks are adjacent.

1 is a conceptual diagram of an obstacle detection device according to a first embodiment of the present invention. 1 is a block diagram illustrating a configuration example of an obstacle detection system according to a first embodiment of the present invention. FIG. 4 is a diagram (part 1) for describing an example of arrangement of side monitoring cameras. FIG. 9 is a diagram (part 2) for describing an example of the arrangement of the side monitoring cameras. It is a figure which shows an example of the camera video image | photographed by the side surveillance camera. FIG. 6 is a diagram for explaining an adjacent trajectory region in the camera image shown in FIG. 5. FIG. 6 is a diagram for explaining a building limit area on an adjacent track in the camera image shown in FIG. 5. It is a flowchart which shows an example of the processing procedure of an obstacle detection process. It is a figure which shows the example of an image which superimposed the building limit area | region on an adjacent track | track on a camera image. FIG. 10 is a diagram illustrating a change example of the image illustrated in FIG. 9 over time. FIG. 7 is a conceptual diagram of an obstacle detection device according to a second embodiment of the present invention. It is a block diagram showing an example of composition of an obstacle detection system concerning a 2nd embodiment of the present invention. It is a figure for explaining an example of arrangement of a side surveillance camera in a 2nd embodiment. 13 is a flowchart illustrating an example of a procedure of an obstacle detection process according to the second embodiment.

  Hereinafter, embodiments of the present invention will be described in detail with reference to the drawings. In each of the drawings, the same components are denoted by the same reference numerals, and repeated description will be omitted.

(1) First Embodiment FIG. 1 is a conceptual diagram of an obstacle detection device according to a first embodiment of the present invention. FIG. 2 is a block diagram illustrating a configuration example of the obstacle detection system according to the first embodiment of the present invention. The configuration and functions of the obstacle detection system 1 according to the first embodiment of the present invention will be described with reference to FIGS.

  The obstacle detection system 1 according to the present embodiment is suitable for use in a double track environment where a plurality of tracks are adjacent. Specifically, it is assumed that there are a first track and a second track laid adjacent to each other as a double track environment. It is not necessary that the first track and the second track are always laid in parallel with a fixed distance therebetween. It is assumed that the distance, the altitude difference, and the like change between them.

  As shown in FIG. 2, in the present embodiment, the on-track traveling train 100 traveling on the first track and traveling on the second track (adjacent track) laid adjacent to the first track The adjacent on-track train 300 and a control system 200 for managing the location of each train are communicably connected via a communication unit 130. The obstacle detection system 1 is configured to include at least the side monitoring camera 110 of the on-track traveling train 100, the obstacle detection device 120, and the communication unit 130. The configuration of the obstacle detection system 1 may include the train location management unit 210 of the control system 200.

  In the obstacle detection system 1, the obstacle detection device 120 detects an obstacle on an adjacent track, and determines permission / non-permission of the traveling train 300 on the adjacent track traveling on the adjacent track based on the detection result. The traveling of the train 300 on the adjacent track is controlled based on the determination result.

  Hereinafter, each component will be described in detail.

(1-1) On-track running train 100
As shown in FIGS. 1 and 2, the on-track traveling train 100 is configured to detect an obstacle on the adjacent track based on a side monitoring camera 110 that captures an adjacent track and an image captured by the side monitoring camera 110. And an obstacle detection device 120 for detecting.

  3 and 4 are diagrams (Nos. 1 and 2) for explaining an example of the arrangement of the side monitoring cameras. 3 and 4, the on-track traveling train 100 traveling on the first track 401 in the upward direction in the figure and the second track (adjacent track) 402 on the downward direction in the figures are shown. And a train 300 running on an adjacent track that travels. In FIGS. 3 and 4, the side monitoring camera shooting area 111 is indicated by a broken line as a shooting range of the side monitoring camera 110. An obstacle 400 is shown as an example of an irregular small obstacle on the adjacent track 402.

  As shown in FIG. 3, the side monitoring camera 110 is installed at an angle with respect to the traveling axis (traveling direction) of the on-track traveling train 100, for example, on the leading vehicle of the on-track traveling train 100. In the present embodiment, as shown in FIG. 4, the side monitoring camera 110 is installed perpendicularly to the traveling axis of the on-track traveling train 100 and at an angle of depression. By satisfying such installation conditions, the adjacent track 402 and its surrounding area (which may be simply referred to as “adjacent track 402” for simplicity) are included in the side monitoring camera shooting area 111 (FIG. 4). reference).

  In the present description, a camera (side monitoring camera 110) is used as an example of an external sensor that captures the adjacent trajectory 402. However, the type of the external sensor that can be used in the present embodiment is not limited to a camera. Types of sensors may be used. However, in consideration of the purpose of detecting an irregular small obstacle (obstacle 400) existing on the adjacent orbit 402, the use of a camera or LiDAR is effective. Also, the vehicle on which the side monitoring camera 110 is installed in the on-track traveling train 100 is not limited to the leading vehicle illustrated.

  Next, the obstacle detection device 120 will be described. The obstacle detection device 120 is a computer that can provide various functions mainly by executing a program. As shown in FIGS. 1 and 2, the obstacle detection device 120 includes a train track route acquisition unit 121, a train track position calculation unit 122, an on-track building limit area calculation unit 123, and an on-track obstacle detection unit 124. , And a progress permission determining unit 125 as a functional configuration.

  The train track route acquisition unit 121 acquires travel route information between stations where the train 100 travels on the track. Here, the travel route information may be, specifically, the position, gradient, curvature, branch information, and the like of the travel route based on the departure station. The train track route acquisition unit 121 sends the acquired travel route information to the train location position calculation unit 122.

  The train on-rail position calculating unit 122 obtains the running route information of the on-track traveling train 100 from the train track route obtaining unit 121, and calculates the current on-rail position (on-rail position information) on the running route of the on-track running train 100. Then, the train on-line position calculation unit 122 transmits the calculated on-rail position information to the on-track building limit area calculation unit 123 and also transmits to the train on-line position management unit 210 of the control system 200 via the communication unit 130. .

  In addition, the method by which the train on-rail position calculating unit 122 calculates the on-rail position is not particularly limited. Specifically, for example, a method is conceivable in which a GPS (Global Positioning System) is mounted on the on-track traveling train 100 and calculation is performed based on GPS information obtained from the GPS. Further, for example, a method of calculating the on-rail position by comparing a camera video (image) captured by the side monitoring camera 110 when the vehicle traveled on the same travel route in the past with a current camera video (image) is also considered. Can be Further, for example, a sensor capable of acquiring distance information to a target such as LiDAR is installed in the on-track traveling train 100, and three-dimensional position information acquired when the vehicle traveled on the same traveling route in the past and current three-dimensional position information are acquired. A method of calculating the on-rail position by collating with.

  In addition, in the obstacle detection device 120, in parallel with the processing by the train track route acquisition unit 121 and the train on-rail position calculation unit 122, the building limit area calculation unit 123 on the adjacent track, the obstacle detection unit 124 on the adjacent track, Then, an obstacle detection process is performed by the progress permission determination unit 125. Since the details of the obstacle detection processing will be described later with reference to FIG. 8, only the outline of the obstacle detection processing will be described below with reference to FIGS.

  The on-track building limit area calculation unit 123 receives the camera image 403 (see FIG. 5) captured by the side monitoring camera 110 and calculates an adjacent track area 404 (see FIG. 6) in the camera image 403. Here, the adjacent trajectory region means a region in a predetermined range including the adjacent trajectory.

  FIG. 5 is a diagram illustrating an example of a camera image captured by the side monitoring camera. FIG. 5 illustrates a camera image 403 captured by the side monitoring camera 110, and an adjacent track 402 is captured. FIG. 6 is a diagram for explaining an adjacent trajectory region in the camera image shown in FIG. According to FIG. 6, in the camera image 403, a region near the adjacent trajectory 402 is shown as an adjacent trajectory region 404.

  Further, the adjacent track-based building limit area calculation unit 123 uses the calculated adjacent track area 404 and the adjacent track-based vehicle information (details will be described later) received from the train on-line position management unit 210 of the control system 200 to generate an adjacent track. The building limit area 405 (see FIG. 7) on the adjacent track of the vehicle traveling on the track 402 is calculated. Here, the construction limit area on the adjacent track is a predetermined range (building that may not obstruct the train operation in order to secure the safety of train operation on the adjacent track). Critical area).

  FIG. 7 is a diagram for describing a building limit area on an adjacent track in the camera image shown in FIG. According to FIG. 7, in the camera image 403, the adjacent track 402 and the area above the adjacent track 402 are shown as the building limit area 405 on the adjacent track. As a specific image, an area larger than the size of the adjacent track running train 300 traveling on the adjacent track 402 by a predetermined degree is set as the adjacent track limit area 405. The arrows shown in FIGS. 5 to 7 and FIGS. 9 and 10 to be described later mean the traveling direction (traveling direction) of the on-track traveling train 100 on which the side monitoring camera 110 is installed.

  The adjacent on-track obstacle detection unit 124 receives the adjacent on-track building limit area 405 calculated by the adjacent on-track building limit area calculation unit 123, and enters the adjacent on-track building limit area 405 into the adjacent on-track building limit area 405. It detects whether there is an obstacle that hinders traveling. For example, in the case of FIG. 7, there is no obstacle. On the other hand, in FIG. 9 and FIG. 10 described later, the obstacle 400 exists in the building limit area 405 on the adjacent track.

  The traveling permission determining unit 125 determines whether traveling of the train 300 traveling on the adjacent track 402 is permitted or not based on the result of the detection of the obstacle by the neighboring track detecting unit 124. Specifically, when an obstacle is detected, it is determined that the traveling of the adjacent train 300 is not permitted (not permitted), and a stop signal is transmitted to the control system 200 or the adjacent train 300. . On the other hand, if no obstacle is detected, it is determined that the traveling of the train 300 running on the adjacent track is permitted, and a progress permission signal is transmitted to the control system 200 and the train 300 running on the adjacent track. At this time, the progress permission determination unit 125 can also transmit the detection result of the obstacle by the on-orbit obstacle detection unit 124 to the control system 200.

  Further, as a modification of the obstacle detection system 1 according to the present embodiment, the train on-line position management unit 210 of the control system 200 may determine whether to permit or prohibit the traveling of the train 300 on the adjacent track. Good. In this case, the traveling permission determination unit 125 does not need to perform the traveling permission / non-permission determination, and the train on-line position management unit 210 of the control system 200 determines whether the traveling Based on the determination, the permission / prohibition of the traveling of the train 300 on the adjacent track may be determined, and a progress permission signal or a stop signal may be transmitted to the train 300 on the adjacent track according to the determination result.

(1-2) Control System 200
As shown in FIGS. 1 and 2, the control system 200 includes a train location control unit 210. The traffic control system 200 is a system for controlling the operation of a train, and has a functional unit other than the train location management unit 210, but is not shown in the drawings because it is not related to the present description.

  The train on-line position management unit 210 uses vehicle information (adjacent on-track vehicle information) on the nearest adjacent on-track running train 300 that runs on the adjacent track 402 as viewed from the on-track running train 100 using an operation diagram of the target route. The acquired information is transmitted to the obstacle detection device 120 of the on-track traveling train 100 (the adjacent on-track building limit area calculation unit 123) via the communication unit 130.

  When the nearest adjacent on-track train 300 overtakes or passes by the on-track train 100, the train on-rail position management unit 210 updates the on-track vehicle information to the next overtaking or passing vehicle information. And retransmit to the on-track traveling train 100.

  Here, the adjacent on-track vehicle information is, for example, the vehicle structure (vehicle width, vehicle height, number of trains, and the like) of the nearest adjacent on-track train 300. As will be described later, when the on-track running train 100 and the adjacent on-track running train 300 directly communicate with each other via the communication unit 130, the IP address and the like assigned to the adjacent on-track running train 300 are also included. .

  Also, when the adjacent on-track traveling train 300 has the same configuration as the on-track traveling train 100 (in particular, when it has a function corresponding to the on-track position calculation unit 122), the on-track position management unit 210 The current on-rail position information of the adjacent on-track train 300 can also be transmitted to the on-track train 100 as adjacent on-track vehicle information.

  In addition, the train-on-rail position management unit 210 receives the detection result of the obstacle by the adjacent on-track obstacle detection unit 124 from the obstacle detection device 120 (progress permission determination unit 125) of the on-track traveling train 100, The traveling permission signal / stop signal of the traveling train 300 is received. Here, when the progress permission signal / stop signal of the adjacent track running train 300 is received, for example, the control signal to the adjacent track running train 300 is finally determined by the administrator of the control system 200. Is also good. In this case, a progress permission signal or a stop signal is transmitted to the train control drive unit 310 of the train 300 running on the adjacent track according to the final decision of the manager.

  In the present embodiment, the traffic control system 200 is provided to improve the safety of the obstacle detection system 1, but the traffic control system 200 is not an essential component in the obstacle detection system of the present invention. For example, an access point (corresponding to the communication unit 130) may be provided on the track, and the running control may be performed using only the inter-vehicle communication between the on-track running train 100 and the adjacent on-track running train 300. At this time, the traveling permission determining unit 125 of the obstacle detection device 120 in the on-track traveling train 100 transmits a traveling permission signal or a stop signal to the train control driving unit 310 of the adjacent on-track traveling train 300, The traveling control of the traveling train 300 on the adjacent track can be realized.

(1-3) Train 300 on adjacent track
As shown in FIG. 2, the train 300 running on an adjacent track includes a train control drive unit 310 that controls and drives the own vehicle.

  The train control drive unit 310 continues or stops the running of the train 300 on the adjacent track based on the progress permission signal or the stop signal input via the communication unit 130. More specifically, the train control drive unit 310 has two states: an inactive state in which a brake is applied to the adjacent on-track train 300, and an active state in which the brake is released. In the present embodiment, for example, it is configured to be normally inactive and to be active only while receiving the progress permission signal. This makes it possible to realize a so-called fail-safe configuration in which the inactive state to which the brake is applied is set to be on the safe side, and the safety of train control can be enhanced.

  Although not shown in FIG. 2, as a modified example of the present embodiment, the adjacent on-track traveling train 300 corresponds to the side monitoring camera 110 and the obstacle detection device 120 similarly to the on-track traveling train 100. A configuration may be provided. In the case of such a configuration, both the on-track traveling train 100 and the adjacent on-track traveling train 300 can realize forward monitoring (obstruction detection) of each other.

(1-4) Obstacle Detection Processing As described above in (1-1), in the obstacle detection device 120, the on-orbit building limit area calculation unit 123, the on-orbit obstacle detection unit 124, and the progress permission determination An obstacle detection process is performed by the unit 125. Hereinafter, the obstacle detection processing according to the present embodiment will be described in detail with reference to FIGS.

  FIG. 8 is a flowchart illustrating an example of a processing procedure of the obstacle detection processing. The obstacle detection processing according to the present embodiment is realized by executing the processing procedure illustrated in FIG. 8 at a constant cycle.

  According to FIG. 8, when the obstacle detection process is started, first, the building limit area calculation unit 123 on the adjacent track acquires the camera image 403 captured by the side monitoring camera 110 (step S11). Specifically, the camera image 403 illustrated in FIG. 5 is obtained.

  Next, the adjacent orbital building limit area calculation unit 123 calculates the adjacent orbital area 404 in the camera image 403 input in step S11 (step S12). When the calculation processing of the adjacent trajectory region 404 ends, the process proceeds to step S13.

  Here, an example of a method of calculating the adjacent trajectory region 404 in step S12 will be described. In this example, it is basically assumed that the trajectory is linear without interruption.

  First, the adjoining on-track construction limit area calculation unit 123 includes: an installation position and a posture of the side monitoring camera 110; travel route information of the on-track traveling train 100 acquired by the train track route acquisition unit 121; Based on the on-rail position information calculated by 122, a rough orbit existing area in the camera image 403 is estimated.

  Next, the on-track adjacent building limit area calculation unit 123 performs image processing on the estimated rough track existence area to extract edge information in the image. Here, the edge information indicates, for example, a line segment area where similar luminance values are continuous in an image.

  Then, by acquiring a straight line area along the gradient of the track from the extracted edge information, the adjacent-track building limit area calculation unit 123 can calculate the adjacent track area 404 indicating the peripheral area of the adjacent track 402. (See FIG. 6).

  In step S13, the adjacent-track building limit area calculation unit 123 determines whether or not the adjacent track area 404 exists in the camera image 403 of the side monitoring camera 110. Specifically, for example, it is determined whether or not the adjacent trajectory region 404 calculated in step S12 exists in the camera image 403 obtained in step S11 over a certain area. In step S13, when the adjacent orbital region 404 exists over a certain area, it is determined that the adjacent orbital region 404 exists (YES in step S13), and the process proceeds to step S14. On the other hand, when the adjacent track area 404 does not exist over a certain area, it is determined that the adjacent track area 404 does not exist (NO in step S13), and the process proceeds to step S19.

  In step S14, the on-track building limit area calculation unit 123 acquires the on-track vehicle information from the train on-rail position management unit 210 of the control system 200 via the communication unit 130, and proceeds to step S15.

  In step S15, the adjacent on-track building limit area calculation unit 123 calculates the adjacent on-track building limit area 405 based on the adjacent track area 404 calculated in step S12 and the adjacent on-track vehicle information acquired in step S14. I do.

  Here, the calculation of the building limit area 405 on the adjacent track in step S15 will be described in detail.

  First, as illustrated in FIG. 6, the adjacent trajectory regions 404 calculated in step S12 are two linear regions each including one trajectory (adjacent trajectory 402). Further, the width of the two adjacent tracks 402 is fixed, and basically, the interval does not change in the middle of the route. In addition, the two trajectories are on the same plane.

  From the above, the adjacent orbital building limit area calculation unit 123 calculates the installation position of the side monitoring camera 110, the camera parameters of the side monitoring camera 110 (settings related to imaging), and the pixels between the two adjacent track areas 404. Based on the number, the distance in the real space per pixel constituting the ground in the camera image 403 can be calculated.

  On the other hand, since the adjacent on-track vehicle information acquired in step S14 includes information indicating the vehicle width and the vehicle height of the adjacent on-track running train 300, the adjacent on-track building limit area calculation unit 123 sets By using these pieces of information and the distance in the real space per pixel calculated in advance, the building limit area 405 on the adjacent track in the camera image 403 can be calculated (see FIG. 7). As described above, a specific image of the adjacent railroad building limit area 405 is an area larger than the size of the adjacent railroad running train 300 running on the adjacent railroad track 402 by a predetermined amount. In the case of FIG. It is calculated in the area.

  When the calculation process of the adjacent on-track construction limit area 405 in step S15 ends, the adjacent on-track construction limit area calculation unit 123 converts the camera image 403 and the adjacent on-track construction limit area 405 into the adjacent on-track obstacle detection unit 124. , And the on-orbit obstacle detection unit 124 performs the process of step S16 with the input as a trigger.

  FIG. 9 is a diagram illustrating an example of an image in which a building limit area on an adjacent track is superimposed on a camera image. By inputting the camera image 403 and the building limit area 405 on the adjacent track, the obstacle detecting unit 124 on the adjacent track can acquire an image as shown in FIG. 9 at the start of step S16.

  In step S16, the adjacent on-track obstacle detection unit 124 determines whether there is an obstacle that hinders the running of the adjacent on-track train 300 in the adjacent on-track building limit area 405 calculated in step S15. Is detected (obstacle detection).

  As a method of obstacle detection in step S16, for example, a method in which the calculation processing of the adjacent trajectory region 404 in step S12 is extended can be considered. That is, in the camera image 403 (particularly, the building limit area 405 on the adjacent track), a process of calculating the continuity of a straight line area obtained based on the edge information like the adjacent track area 404 is effective. As a result, for example, when a rupture in the linear region is found, it can be determined (detected) that there is an obstacle that blocks the adjacent trajectory 402. For example, the obstacle 400 shown in FIG. 9 has broken the continuity of the adjacent trajectory 402.

  However, when an actual train traveling environment is taken into consideration, there is a possibility that an object that is not an obstacle on the adjacent track 402 (for example, an operation facility) is reflected in the camera image 403 of the side monitoring camera 110. For example, a utility pole 410 shown in FIG. 9 is a service facility installed between a first track 401 and a second track (adjacent track) 402, and is not an obstacle on the adjacent track 402. However, in the case of FIG. 9, similarly to the obstacle 400, the utility pole 410 is also photographed in the building limit area 405 on the adjacent track, and the continuity of the adjacent track 402 is broken. Therefore, erroneous detection may be caused only by the calculation of the continuity of the linear region described above.

  In order to avoid such erroneous detection, for example, a motion stereo method that calculates a corner feature point in the building limit area 405 on the adjacent track and uses the movement amount of the corner feature point between frames is effective. . FIG. 9 shows a feature point 411 and a feature point 412 as corner feature points of the telephone pole 410 and the obstacle 400. The motion stereo method calculates the three-dimensional position information of the corner feature points in the real space by the principle of triangulation by using the traveling speed of the on-orbit running train 100 and the movement amount of the corner feature points between frames. It is a technique to do. The KLT (Kanade-Lucas-Tomasi) method is widely known for calculating the amount of movement of a corner feature point between frames.

  A method for avoiding erroneous detection using the motion stereo method will be specifically described with reference to FIG. FIG. 10 is a diagram illustrating an example of a change with time of the image illustrated in FIG. 9.

  In FIG. 10, a corner feature point of the utility pole 410 at time t is indicated by a feature point 411, and a corner feature point after a lapse of time Δt (time t + Δt) is indicated by a feature point 421. The feature point 411 and the feature point 421 indicate the same position on the utility pole 410. Similarly, for the obstacle 400, a corner feature point at time t is indicated by a feature point 412, and a corner feature point after a lapse of time Δt (time t + Δt) is indicated by a feature point 422.

Here, when the motion stereo method is used, an arbitrary feature point P in the camera image 403 is converted from P _A (u_a, v_a) to P _B (from the time t to the time t + Δt in the image coordinate system in the image). u_b, v_b), when the angle between the optical axes of the side monitoring cameras 110 before and after the movement is θ, and the traveling speed of the on-track traveling train 100 is V, the depth information Z of the feature point P is Can be calculated by the following equation (1). The height information H of the feature point P can be calculated by the following equation (2).

  In the above equations (1) and (2), f means the focal length of the side monitoring camera 110, which is a known value that can be obtained by camera calibration. The angle θ formed by the optical axis of the side monitoring camera 110 before and after the movement is obtained, for example, by the travel route information of the on-track traveling train 100 acquired by the train track route acquisition unit 121 and the train on-rail position calculation unit 122. It may be obtained by calculating the attitude change of the on-track traveling train 100 based on the on-track position information of the on-track traveling train 100. However, considering that a sharp turn does not occur in the running environment of the train, the calculation may be performed with θ = 0 when Δt is minute.

  By using the motion stereo method in this way, the adjacent on-orbit obstacle detection unit 124 obtains the three-dimensional position information (depth information Z) on the real space for the feature point P in the adjacent orbit building limit area 405. The height information H can be calculated, and an obstacle existing on the adjacent trajectory 402 can be detected based on the calculation result. That is, if the position specified by the depth information Z and the height information H is within the building limit area 405 on the adjacent track, the feature point P can be determined to be an obstacle existing on the adjacent track 402. If it is not within the building limit area 405 on the adjacent track (for example, the obstacle 400), it can be determined that the feature point P is not an obstacle existing on the adjacent track 402 (for example, the telephone pole 410). In addition, by combining the above-described continuity of the linear region and the motion stereo method, it becomes possible to close the trajectory and more accurately detect an object existing on the trajectory.

  The detailed processing method of the obstacle detection in step S16 for detecting whether or not there is an obstacle that hinders the traveling of the train 300 running on the adjacent track while avoiding erroneous detection has been described above. After the processing in step S16, the detection result of the obstacle detection is input from the adjacent track obstacle detection unit 124 to the traveling permission determination unit 125, and the process proceeds to step S17.

  In step S17, the progress permission determination unit 125 confirms the presence or absence of an obstacle from the detection result of the obstacle detection in step S16. If there is an obstacle (YES in step S17), the process proceeds to step S18, and if there is no obstacle (NO in step S17), the process proceeds to step S19.

  In step S <b> 18, the progress permission determination unit 125 transmits a stop signal (or an abnormality occurrence signal indicating that an obstacle is present) of the train 300 on the adjacent track via the communication unit 130 to the train location position of the control system 200. The information is transmitted to the management unit 210. Then, based on the reception of the stop signal (or the abnormality occurrence signal), the control system 200 instructs the train control drive unit 310 of the adjacent on-track traveling train 300 to stop progressing, thereby causing the adjacent on-track traveling train 300 to stop. Can be stopped. In addition, the control system 200 may also perform processing such as recording the detection result of the obstacle detection. For example, by recording the presence of an obstacle together with the position information of the detected obstacle, it can be useful when giving an instruction to a worker to remove the obstacle.

  In step S18, besides the above-described processing method, for example, when an obstacle is detected in a state where the inter-vehicle distance between the on-track traveling train 100 and the adjacent on-track traveling train 300 is short, or as in the modified example described above. In the case of a configuration without the control system 200 or the like, the stop signal (described above) is transmitted from the traveling permission determination unit 125 of the on-track traveling train 100 to the train control driving unit 310 of the adjacent on-track traveling train 300 via the communication unit 130. An abnormal occurrence signal may be transmitted directly, and the train 300 running on the adjacent track may be immediately stopped.

  On the other hand, in step S19, there is no obstacle, so there is no need to stop the train 300 running on the adjacent track. Therefore, the travel permission determining unit 125 transmits a travel permission signal (or a signal indicating no abnormality indicating that no obstacle exists) to the train location control unit 210 of the control system 200 via the communication unit 130. By receiving the traveling permission signal (abnormality-free signal), the control system 200 can confirm that the adjacent track 402 has no abnormality. In step S19, since no abnormality has occurred with respect to the obstacle, no signal may be transmitted to the control system 200.

  After the processing in step S18 or step S19, the obstacle detection processing ends, but after a certain period has elapsed, the obstacle detection processing is performed again. As described above, by performing the obstacle detection processing illustrated in FIG. 8, it is possible to accurately detect an obstacle on the adjacent track 402 from the side of the on-track traveling train 100.

(1-5) Conclusion As described above, according to the obstacle detection system 1 (the obstacle detection device 120) according to the present embodiment, the adjacent trajectory 402 even in a double track environment where a plurality of trajectories are adjacent. The upper obstacle can be detected with high accuracy, and the safety of the traveling of the train 300 on the adjacent track can be enhanced.

  In particular, in the obstacle detection system 1 according to the present embodiment, information on an external sensor installed at an angle to the traveling axis of the vehicle (in this example, the camera image of the side monitoring camera 110 mounted on the side) 403) can be used to obtain the state of the adjacent orbit 402. As a result, it is possible to detect the presence or absence of an obstacle on the adjacent track (second track 402) in addition to the vehicle running track (first track 401). Obstacle detection in front of the upper running train 300) can be realized. That is, since the own vehicle (the on-track train 100) performs the obstacle detection in front of the adjacent on-track train 300, monitoring is performed while capturing the shooting target near the center of the angle of view from closer than when monitoring forward. And an obstacle can be accurately detected. In addition, since the vehicle is monitored on the adjacent track while moving, obstacles and blind spots due to the traveling environment such as structures near the track hardly occur, and obstacles on the adjacent track on the open track can be accurately detected.

  In this manner, the obstacle detection system 1 according to the present embodiment is capable of providing the adjacent track 402 under the double track environment without being affected by the performance of the external sensor, the traveling section, the traveling environment, and the like described above as the problems of the related art. The upper obstacle can be detected accurately, and the traveling safety of the traveling train 300 on the adjacent track can be enhanced.

  The above-mentioned Patent Document 1 describes a function of detecting the presence or absence of an obstacle on an adjacent track when an adjacent track is included in an image captured by a forward monitoring camera installed on the front of the vehicle. I have. FIG. 7 of Patent Literature 1 shows an example in which a traveling trajectory of a company is taken as a center and a part of an adjacent trajectory reflected in the field of view is detected.

  However, in the case of the on-orbit obstacle detection system disclosed in Patent Literature 1, obstacles on adjacent orbits photographed in the lateral direction can be accurately detected since the original shooting purpose of the monitoring camera is to detect a forward obstacle. It is not easy to detect, and when the train is traveling on a steep curve or slope, or when the position of the adjacent track is far from the running track of the vehicle, the adjacent track may not necessarily be detected depending on the angle of view of the monitoring camera. Cannot be taken. That is, the on-orbit obstacle detection system disclosed in Patent Literature 1 does not disclose the idea of sufficiently photographing the vicinity of an adjacent orbit, and it is assumed that it is not possible to detect the presence or absence of an obstacle on an adjacent orbit.

  Also, in the case of irregular small obstacles such as pebbles that can cause derailment, the number of pixels required for detection is not reflected in the image, making it difficult to detect, even when shooting with a surveillance camera installed in front of the vehicle. Could be. On the other hand, for example, if the focal length of the lens mounted on the surveillance camera is increased in order to project small obstacles, the angle of view of the camera will be narrowed accordingly, and depending on the traveling section, the periphery of the adjacent track may not be sufficiently projected Improve the nature.

  As described above, in the related art including the on-orbit obstacle detection system disclosed in Patent Document 1, it must be said that it is difficult to accurately detect the presence or absence of an on-orbit obstacle in an open orbit. .

  Further, in the obstacle detection system 1 according to the present embodiment, an object that is not included in the building limit area 405 on the adjacent orbit is reflected in the information of the external sensor (camera image 403 by the side monitoring camera 110). However, as described in step S16 in FIG. 8, by applying a known method such as the motion stereo method, for example, only obstacles that hinder the traveling of the on-track train 300 while avoiding erroneous detection are avoided. Can be accurately detected.

  Further, according to the obstacle detection system 1 according to the present embodiment, when the obstacle detection device 120 detects an obstacle on the adjacent track 402, the on-orbit traveling train equipped with the obstacle detection device 120 Since it is possible to stop the traveling of the traveling train 300 on the adjacent track from 100 directly or via the control system 200, it is possible to realize quick safety control.

  The obstacle detection system 1 (obstacle detection device 120) according to the present embodiment can be used not only as a system for supporting the surroundings of a driver in manned driving, but also for unmanned operation of a train on an open track. It can also be used as a forward monitoring system for trains running on adjacent tracks to realize the above.

(2) Second Embodiment A second embodiment of the present invention will be described. In the obstacle detection system 1 according to the first embodiment, one external sensor (side monitoring camera 110) is mounted on the on-track traveling train 100, whereas in the second embodiment, Is different in that a plurality of external sensors (side monitoring cameras 110A, 110B, 110C) are mounted on the on-track traveling train 100.

  FIG. 11 is a conceptual diagram of an obstacle detection device according to the second embodiment of the present invention. FIG. 12 is a block diagram illustrating a configuration example of an obstacle detection system according to the second embodiment of the present invention. Note that FIGS. 11 and 12 correspond to FIGS. 1 and 2 exemplified in the first embodiment, and the same components are denoted by the same reference numerals and description thereof will be omitted.

  As illustrated in FIGS. 11 and 12, the obstacle detection system 2 according to the second embodiment includes side monitoring cameras 110A and 110B as an example of a plurality of external sensors mounted on the on-track traveling train 100. , 110C. These are cameras that monitor on an adjacent track similarly to the side monitoring camera 110 in the first embodiment. For example, the side monitoring camera 110A is attached to the first vehicle and the side monitoring camera is attached to the second vehicle. As the monitoring camera 110B, a side monitoring camera 110C is mounted on the third vehicle. The sensor data (camera image) acquired by each of the side monitoring cameras 110A to 110C is input to the building limit area calculation unit 123 on the adjacent track of the obstacle detection device 120.

  FIG. 13 is a diagram for explaining an example of the arrangement of the side monitoring cameras according to the second embodiment. As shown in FIG. 13, the on-track traveling train 100 traveling on the first track 401 in the upward direction in the figure has side monitoring cameras 110A to 110C in order from the first to third cars in the traveling direction. Is installed. The side monitoring cameras 110A to 110C are cameras for photographing the second trajectory (adjacent trajectory) 402, and the photographing range is indicated by camera photographing areas 111A to 111C. Each camera is installed at an angle with respect to the traveling axis of the on-track traveling train 100. In FIG. 13, as an example, all the cameras are installed perpendicularly to the traveling axis of the train and at an angle of depression, but the present invention is not limited to this, and the position and orientation may be set for each camera. In addition, the angle of view and the shooting cycle may be changed for each camera.

  In this example, a plurality of side monitoring cameras 110A to 110C are shown as an example of the plurality of external sensors. However, the type of the specific external sensor is not limited. For example, a combination of a camera and LiDAR may be used. Good. FIG. 13 shows an obstacle 400 as an example of an irregular small obstacle on the adjacent orbit 402.

  As shown in FIG. 13, in the second embodiment, a plurality of side monitoring cameras 110 </ b> A to 110 </ b> C can capture a wider range than a single camera (see FIG. 3). Also, it is possible to capture an overlapping area. For this reason, by processing images from these cameras, it is possible to detect obstacles with higher accuracy.

  Next, an obstacle detection process according to the second embodiment will be described focusing on differences from the first embodiment.

  FIG. 14 is a flowchart illustrating an example of a processing procedure of an obstacle detection process according to the second embodiment. Note that the processing illustrated in FIG. 14 has many in common with the obstacle detection processing illustrated in FIG. 8 in the first embodiment, and a detailed description thereof will be omitted. The difference from the first embodiment (see FIG. 8) is that since a plurality of external sensors are mounted, it is necessary to execute the obstacle detection processing on adjacent tracks for the number of external sensors. The obstacle detection processing according to the second embodiment is realized by executing the processing procedure illustrated in FIG. 14 at regular intervals.

  According to FIG. 14, when the obstacle detection process is started, first, the building limit area calculation unit 123 on the adjacent track acquires a camera image captured by any of the side monitoring cameras 110A to 110C (step S11). ). For example, as an initial step S11, a camera image of the side monitoring camera 110A is obtained.

  Then, in the next step S12, the adjacent on-track building limit area calculation unit 123 calculates an adjacent track area in the acquired camera image. As the method of calculating the adjacent trajectory region, the method described in the first embodiment can be adopted.

  In the next step S21, in the same manner as in step S13 of FIG. 8, the adjacent orbital building limit area calculation unit 123 determines that the adjacent orbital area exists in the camera image of the side monitoring camera 110A acquired in step S11. It is determined whether or not to perform. If it is determined that there is an adjacent trajectory region, the process proceeds to step S14. On the other hand, if it is determined that there is no adjacent orbit area, it is not immediately determined that there is no obstacle as shown in FIG. 8, but the camera images by the other side monitoring cameras 110B and 110C are checked in step S22. Proceed to.

  The processing in steps S14 to S16 is the same as that in FIG. 8, and the adjacent track building limit area calculation unit 123 acquires the adjacent track vehicle information from the control system 200 (train on-rail position management unit 210), and The upper architectural limit area is calculated, and it is detected whether or not there is an obstacle in the adjacent railroad architectural limit area that hinders the traveling of the adjacent railroad traveling train 300 (obstacle detection).

  When the obstacle detection in step S16 is completed, it is confirmed whether or not the processing in steps S11 to S16 has been completed for the number of external sensors (step S22). If the processing has been completed for all the external sensors (side monitoring cameras 110A to 110C) (YES in step S22), the process proceeds to step S23. On the other hand, if at least one device has not been processed (NO in step S22), the process returns to step S11, and the process is performed on the sensor data of the unprocessed external sensor (for example, the side monitoring camera 110B).

  In step S23, the progress permission determination unit 125 integrates the obstacle detection results of the external sensors, and in step S24, confirms the presence or absence of an obstacle based on the integrated result. Specifically, for example, when obstacle detection is performed on the sensor data of the three side monitoring cameras 110A to 110C, and when an obstacle is detected in at least one of the data, the integration result is displayed. It is determined that there is an obstacle (YES in step S24), and the process proceeds to step S18. On the other hand, if no obstacle is detected in any of the three units, it is determined that there is no obstacle (NO in step S24), and the process proceeds to step S19.

  Thereafter, as in FIG. 8, an abnormality occurrence signal is transmitted in step S18, and an abnormality absence signal is transmitted in step S19, and the obstacle detection processing ends.

  As described above, by performing the obstacle detection process illustrated in FIG. 14, it is possible to accurately detect the obstacle on the adjacent track 402 based on the sensor data of the plurality of external sensors mounted on the on-track traveling train 100. it can.

  As described above, the obstacle detection system 2 according to the second embodiment is provided with a plurality of external sensors that capture images on adjacent tracks, thereby providing the obstacle detection system 1 according to the first embodiment. Can be regarded as an improvement.

  When the obstacle detection system 2 is compared with the obstacle detection system 1, the number of external sensors (side monitoring cameras 110) installed in the obstacle detection system 1 is one. When the vehicle runs at a high speed, there is a possibility that an obstacle may not be detected. Specifically, when the vehicle runs at high speed, the possibility that an image that is not suitable for obstacle detection is acquired due to blurring of the captured image or that the difference between the images (pixel movement amount) increases. Tracking obstacles can be difficult.

  On the other hand, in the obstacle detection system 2 according to the second embodiment, for example, an external sensor (side monitoring cameras 110A, 110B, 110C) is installed in each vehicle of the own vehicle, and the plurality of external sensors are used. By executing the obstacle detection using the acquired sensor data (camera image 403), a plurality of detection results can be compared with each other, and it can be expected that undetected obstacles are suppressed.

  Here, since many trains run in a plurality of rolling trains, for example, in the case of a five-car rolling train, the position where the external sensor can be installed is assumed to reach 100 m or more. Therefore, sensor data acquired at different timings (time) can be used for obstacle detection even if a plurality of external sensors are installed at equal intervals, for example, for a train length of 100 m.

  From the above, in addition to the effects obtained in the first embodiment, the obstacle detection system 2 according to the second embodiment suppresses the occurrence of undetected obstacles, particularly during high-speed running, Obstacle detection with higher accuracy than in the first embodiment can be realized.

  In addition, the obstacle detection system 2 according to the second embodiment uses a plurality of external sensors, so that, for example, even if one of the external sensors fails, the entire system does not immediately stop functioning. Since the obstacle detection can be continued, the effect of increasing the availability and durability of the system is achieved.

  Note that the present invention is not limited to the above embodiments, and includes various modifications. For example, the above-described embodiments have been described in detail for easy understanding of the present invention, and are not necessarily limited to those having all the configurations described above. For example, for a part of the configuration of each embodiment, it is possible to add, delete, or replace another configuration.

  Further, each of the above-described configurations, functions, processing units, processing means, and the like may be partially or entirely realized by hardware, for example, by designing an integrated circuit. In addition, the above-described configurations, functions, and the like may be implemented by software by a processor interpreting and executing a program that implements each function. Information such as a program, a table, and a file for realizing each function can be stored in a memory, a hard disk, a recording device such as an SSD (Solid State Drive), or a recording medium such as an IC card, an SD card, or a DVD.

  In addition, control lines and information lines indicate those which are considered necessary for explanation, and do not necessarily indicate all control lines and information lines on a product. In implementation, it may be considered that almost all components are interconnected.

1, 2 obstacle detection system 100 on-track running train 110 (110A, 110B, 110C) side monitoring camera 111 (111A, 111B, 111C) side monitoring camera shooting area 120 obstacle detection device 121 train track route acquisition unit 122 Train on-track position calculation unit 123 Adjacent track building limit area calculation unit 124 Adjacent on-track obstacle detection unit 125 Progress permission determination unit 130 Communication unit 200 Control system 210 Train on-track position management unit 300 Adjacent track running train 310 Train control drive unit 400 Obstacle 401 First trajectory 402 Second trajectory (adjacent trajectory)
403 Camera image 404 Adjacent track area 405 Adjacent track building limit area 410 Electric poles 411, 412, 421, 422 Feature points

Claims (16)

An obstacle detection system that detects an obstacle in a double track environment including an adjacent first trajectory and a second trajectory,
An external field sensor that is installed at an angle to the traveling axis of the first train on the first train traveling on the first track, and includes the second track on the sensor area,
Using the sensor data of the external world sensor, an obstacle detection device that detects the presence or absence of an obstacle on the second track,
A communication unit that transmits and receives data or signals between the obstacle detection device and the outside,
An obstacle detection system comprising: The obstacle detection device,
A train track route obtaining unit for obtaining a track route of the vehicle,
On-rail position calculating unit that calculates the on-rail position of the own vehicle based on the track route obtained by the train track route obtaining unit,
From the sensor data of the external world sensor, an adjacent on-track building limit area calculation unit that calculates a building limit area on the second track,
An adjacent track for detecting the presence or absence of an obstacle on the second track by determining whether an obstacle is present in the building limit area on the second track using sensor data of the external field sensor An upper obstacle detection unit,
The obstacle detection system according to claim 1, comprising: The adjacent on-orbit obstacle detection unit, among the sensor data of the external sensor, is reflected in the second orbit building limit area calculated by the adjacent on-track building limit area calculation unit, The obstacle detection system according to claim 2, wherein an object not existing on the second trajectory is removed from a detection target of the presence or absence of the obstacle. The obstacle detection device,
Advancing to determine whether to permit or stop the traveling of the second train traveling on the second track based on the detection result of the presence or absence of the obstacle on the second track by the adjacent track obstacle detecting unit. The obstacle detection system according to claim 2, further comprising: a permission determination unit. The traveling permission determining unit, when the adjacent on-track obstacle detecting unit detects that there is an obstacle on the second track, sends a stop signal for stopping the progress of the second train to the communication unit. The obstacle detection system according to claim 4, wherein the transmission is performed to the second train via the communication terminal. Further comprising a control center that manages the operation of a plurality of trains running in the double track environment,
The obstacle detection device mounted on the first train,
Calculating the on-board position of the vehicle and transmitting it to the control center, and transmitting a detection result of the presence or absence of an obstacle on the second track to the control center,
The control center,
Based on the on-rail position of the first train received from the obstacle detection device, while acquiring the vehicle information of the second train running on the second track, the received from the obstacle detection device The obstacle detection system according to any one of claims 1 to 4, wherein permission or stop of traveling of the second train is controlled based on a detection result of the presence / absence of an obstacle. A plurality of the outside world sensors are installed on the first train,
The said obstacle detection device detects the presence or absence of the obstacle on the said 2nd track using the sensor data of the said some external field sensor. The Claim 1 characterized by the above-mentioned. The obstacle detection system according to 1. Each of the first train running on the first track and the second train running on the second track is equipped with the external sensor and the obstacle detection device,
The obstacle detection device mounted on the first train detects the presence or absence of an obstacle on the second track using the sensor data of the external sensor mounted on the first train, The obstacle detection device mounted on the second train detects presence / absence of an obstacle on the first track using sensor data of the external sensor mounted on the second train. The obstacle detection system according to any one of claims 1 to 7, characterized in that: An obstacle detection method by an obstacle detection system that detects an obstacle in a double track environment including an adjacent first trajectory and a second trajectory,
The obstacle detection system includes:
An external field sensor that is installed at an angle to the traveling axis of the first train on the first train traveling on the first track, and includes the second track on the sensor area,
Using the sensor data of the external world sensor, an obstacle detection device that detects the presence or absence of an obstacle on the second track,
A communication unit that transmits and receives data or signals between the obstacle detection device and the outside,
An obstacle detection method comprising: A sensor data obtaining step in which the external sensor obtains the sensor area including the second orbit on the sensor data,
A train track route obtaining step in which the obstacle detection device obtains a track route of the vehicle,
The on-board position calculating step of calculating the on-board position of the vehicle based on the track route acquired in the train track route acquiring step,
The obstacle detection device, from the sensor data of the external sensor, the adjacent on-track building limit area calculation step of calculating the building limit area on the second track,
The obstacle detection device determines whether or not an obstacle is present in the building limit area on the second track using the sensor data of the external sensor, thereby obtaining an obstacle on the second track. Obstacle detection step on adjacent orbit to detect the presence or absence of
The obstacle detection method according to claim 9, further comprising: In the adjacent on-orbit obstacle detection step, the sensor data of the external sensor is reflected in the second orbit building limit area calculated in the adjacent orbit building limit area calculation step, The obstacle detection method according to claim 10, wherein an object that does not exist on the second trajectory is removed from a detection target of the presence or absence of the obstacle. The obstacle detection device, based on the detection result of the presence or absence of an obstacle on the second track by the adjacent track obstacle detection step, the progress of the second train traveling on the second track The obstacle detection method according to claim 10, further comprising: a progress permission determination step of determining permission or stop. In the proceeding permission determining step, when it is detected in the adjacent track obstacle detecting step that there is an obstacle on the second track, a stop signal for stopping the progress of the second train is transmitted to the communication unit. The obstacle detection method according to claim 12, wherein the signal is transmitted to the second train via a mobile station. The obstacle detection system further includes a control center that manages the operation of a plurality of trains running in the double track environment,
The obstacle detection device mounted on the first train calculates an on-board position of the vehicle and transmits the calculated position to the control center, and also detects the presence or absence of an obstacle on the second track by the control. Send to the center,
The control center acquires vehicle information of a second train traveling on the second track based on the on-rail position of the first train received from the obstacle detection device, and detects the obstacle. The control of permission or stop of traveling of the second train is controlled based on the detection result of the presence or absence of the obstacle received from the device. The method according to any one of claims 9 to 12, wherein Obstacle detection method. A plurality of the outside world sensors are installed on the first train,
The obstacle detection device detects presence or absence of an obstacle on the second track using sensor data of the plurality of external sensors. 15. The apparatus according to claim 9, wherein: Obstacle detection method according to the above. Each of the first train running on the first track and the second train running on the second track is equipped with the external sensor and the obstacle detection device,
The obstacle detection device mounted on the first train detects the presence or absence of an obstacle on the second track using the sensor data of the external sensor mounted on the first train, The obstacle detection device mounted on the second train detects the presence or absence of an obstacle on the first track using the sensor data of the external sensor mounted on the second train. The obstacle detection method according to any one of claims 9 to 15, characterized in that:

Images