Career path determination system
The route determination system enhances track information accuracy at railway junctions by analyzing spatial images and integrating turnout information, reducing derailment risks.
Patent Information
- Application Number
- JP2024192854
- Authority / Receiving Office
- JP · JP
- Patent Type
- Applications
- Current Assignee / Owner
- Filing Date
- 2024-11-01
- Publication Date
- 2026-05-18
AI Technical Summary
Existing track information determination methods are prone to misjudgment due to varying photographing conditions, leading to potential derailments at railway junctions.
A route determination system that acquires spatial images of railway junctions, analyzes the position of movable rails, and determines track information using turnout information stored in a database, enhancing accuracy by combining image analysis with stored data.
Accurately determines the track information of vehicles passing through junctions, reducing the risk of derailments by improving the precision of track determination.
Smart Images

Figure 2026080821000001_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to a technique for determining track information on which a vehicle travels.
Background Art
[0002] In a track where it is possible to switch the position of the rail on which the vehicle travels by a switch, if there is a mismatch between the track on which the vehicle is scheduled to travel and the position of the rail switched by the switch, it may lead to an accident such as derailment. Therefore, there has been a technique for photographing a track where a switch is installed and determining track information on which the vehicle travels based on the photographed image (for example, Patent Document 1).
Prior Art Documents
Patent Documents
[0003]
Patent Document 1
Summary of the Invention
Problems to be Solved by the Invention
[0004] In the case of Patent Document 1, since the track information was determined using only the track image photographed when the vehicle travels, there is a possibility of misjudgment depending on the photographing conditions and the photographing state, and an improvement in the determination accuracy has been demanded.
[0005] The present invention has been made in view of the above problems, and is a technique for acquiring a spatial image including a branch point, analyzing the spatial image, determining the position of the moving rail, and determining the track information on which the vehicle passing through the branch point travels based on the determination result and the stored switch information.
Means for Solving the Problems
[0006] The present invention provides a route determination system for determining the route information of a vehicle passing through a junction in a railway facility comprising: a plurality of tracks composed of a combination of fixed rails or movable rails whose rail position moves; and a turnout provided at a junction of the tracks, comprising the movable rail and a switch that moves the movable rail, the route determination system comprising: a storage means for storing turnout information relating to the turnout; an image acquisition means for acquiring a spatial image which is a moving image or a plurality of still images of a target space including the junction; a movable rail position determination means for analyzing the spatial image acquired by the image acquisition means and determining the position of the movable rail in the target space; and a route determination means for determining the route information of the vehicle passing through the junction based on the determination result of the movable rail position determination means and the turnout information. [Effects of the Invention]
[0007] According to the present invention, the track information of a vehicle passing through a junction can be determined with high accuracy by analyzing a spatial image of the target space including the junction, determining the position of the moving rail, and determining the track information of a stored turnout. [Brief explanation of the drawing]
[0008] [Figure 1] This figure shows an example of the view from the driver's seat of an inspection vehicle. [Figure 2] (a) is a view of the inspection vehicle from the outside front, and (b) is a view of the interior of the inspection vehicle. [Figure 3] This figure shows an example of the content displayed on the display screen of the operating terminal. [Figure 4] This is a diagram showing an example of railway track facilities. [Figure 5] This is a diagram showing an example of railway track facilities. [Figure 6] This figure shows an example of the hardware configuration of a course determination system. [Figure 7]This figure shows an example of the software configuration for a career path determination system. [Figure 8] This figure shows an example of a database of turnout information stored in the memory unit. [Figure 9] This figure shows an example of the main flowchart for a career path determination system. [Figure 10] This figure shows an example of the process for determining whether a rail is moving. [Figure 11] This figure shows an example of a flowchart for the moving rail detection process and the track detection process. [Figure 12] (a) is a diagram showing an example of notification when the determined track information is reported, and (b) is a diagram showing an example of notification when the determined track information satisfies predetermined conditions. [Modes for carrying out the invention]
[0009] The following describes embodiments of the present invention (hereinafter sometimes referred to as "these embodiments"). Note that the embodiments listed below are illustrative, and the present invention is not limited to the configurations of the embodiments described below.
[0010] First, an overview of the path determination system 100 (hereinafter sometimes referred to as "this system") according to this embodiment will be described. Figure 1 shows an example of the view from the driver's seat of inspection vehicle 10. From the driver's seat of the inspection vehicle 10, the track equipment, including the track 50 on which the inspection vehicle 10 is running, can be seen through the front glass 12 located in front of the instruments 13 and other equipment necessary for driving. During track 50 inspection, a worker (also called a driver) boards an inspection vehicle 10 and drives the vehicle along track 50. When approaching a track junction as shown in Figure 1, the worker stops or slows down the inspection vehicle 10, visually checks the condition of the turnout 55, and confirms that track 50 is passable for the inspection vehicle 10. Previously, the condition of the turnout 55 was only checked visually, which could lead to errors in checking and the vehicle traveling on track 50 that was not passable, resulting in serious accidents. Therefore, the route determination system 100 uses an imaging device 300 to capture images of the target space, which is the track equipment, while the inspection vehicle 10 is traveling, and analyzes the captured images to determine the condition of the turnout 55. As a result, in addition to visual inspection, the worker can use the determination results of the route determination system 100 to confirm which track 50 the inspection vehicle 10 is traveling on, thereby improving safety during travel.
[0011] The inspection vehicle 10 travels forward or backward at a predetermined speed (in this embodiment, 5 to 45 km / h, preferably 25 km / h or less). An imaging device 300 mounted on the inspection vehicle 10 traveling on the track 50 captures images of the target space, including the track 50 on which the inspection vehicle 10 is traveling. In this embodiment, the imaging device 300 captures an image of a predetermined area in front of the inspection vehicle 10 (in this embodiment, 5 to 100 m in front of the inspection vehicle 10, preferably 20 to 50 m) as the target space. The target space includes multiple tracks 50 and switches 55. As will be described in detail later, in this embodiment, when determining the position of the movable rail (also called "tongue rail") 52, the distance between the fixed rail (also called "basic rail") 51 and the movable rail 52, as well as the type of switch (also called "point") 56, are used for the determination, so the image is captured so that these are included. Incidentally, as described above, since the inspection vehicle 10 can also travel backward, it is preferable to make the position of the lens (not shown) of the imaging device 300 movable so that the imaging device 300 can image a predetermined range in the rear, or to provide a plurality of lenses in the imaging device 300. Alternatively, an imaging device (not shown) for imaging a predetermined range in the rear may be provided separately from the imaging device 300. In this case, for example, it may be provided behind the inspection vehicle 10 so that the imaging device for imaging the predetermined range in the rear can image the predetermined range behind the inspection vehicle 10. Thus, with the inspection vehicle 10 as the object space, a predetermined range in the direction in which the inspection vehicle 10 travels is imaged by the imaging device 300.
[0012] Next, the installation positions of the imaging device 300 and the operation terminal 500 mounted on the inspection vehicle 10 will be described with reference to FIG. 2. FIG. 2(a) is a view of the inspection vehicle 10 seen from the outer front. The imaging device 300 is provided above the driver's seat, for example, above the inner side of the windshield 12 provided in front of the inspection vehicle 10. FIG. 2(b) is a view showing the vicinity of the driver's seat of the inspection vehicle 10. The imaging device 300 is provided so that the lens (not shown) of the imaging device 300 can image the track facilities in front of the inspection vehicle 10 through the windshield 12. By providing the imaging device 300 at such a position, a range similar to the field of view of the operator from the driver's seat is imaged. Also, the operation terminal 500 is placed in the space beside the instrument 13 necessary for driving the inspection vehicle 10. The operation terminal 500 displays images captured by the imaging device 300, determination results, and the like. Note that the installation positions of the imaging device 300 and the operation terminal 500 are not limited to the example of FIG. 2. As described above, when it is possible to image a predetermined range in the traveling direction of the inspection vehicle 10, such as behind the inspection vehicle 10, it may be appropriately determined so that the position where the imaging device 300 is installed becomes the optimal position with respect to the inspection vehicle 10.
[0013] Fig. 3 shows an example of the display screen of the operation terminal 500. On the display screen of the operation terminal 500, an image captured by the imaging device 300, a determination result, information related to the running of the inspection vehicle 10, information related to the inspection work, etc. are displayed. It is preferable that the information necessary for the operator to perform the inspection work is displayed so that it is easy to grasp.
[0014] (1) Captured Image Display Area Displays the captured image by the imaging device 300. The determination result may be superimposed on the displayed image. By visually checking the captured image display area, the operator can grasp the state of the line in progress. (2) Alert Display Area Based on the determination result (especially the determination result in case of determination abnormality), the running speed of the inspection vehicle 10, the running position of the inspection vehicle 10, etc., matters that should be alerted to the driver during running are displayed. For example, by displaying a speed excess alert before the installation position of the switch 55, the operator can prevent forgetting to stop temporarily or decelerate at the installation position of the switch 55. (3) Speed Information Display Area Displays the calculated running speed of the inspection vehicle 10. (4) Determination Result Display Area Displays the determination result in characters, symbols, etc. (5) Switch Information Display Area Displays switch information such as the type of the switch 55 installed in a predetermined section (for example, from station XX to station XX) and the position information where the switch 55 is installed. It can be used as a reference when the operator visually checks the switch 55. (6) Collation Result Display Area Displays the collation result between the determined line information and the information in (5). In particular, when the switching mode by the installed switch 55 is stored in advance (when the line to proceed is stored in advance), by displaying the collation result between the stored switching mode of the switch 55 and the determined line information, the safety during the running of the inspection vehicle 10 can be enhanced. (7) System Control Display Area Displays information related to the control of this system. The content and display area shown on the display screen are examples only and are not limited to these. Information related to the operation of the inspection vehicle 10, information related to inspection work, and information that should be made known to the workers will be displayed in a way that is easy for the workers to understand.
[0015] Next, the track equipment used by the route determination system 100 to determine the track condition will be explained using Figures 4 and 5. The route determination system 100 determines the track information of a vehicle passing through a junction in a track facility that includes a plurality of tracks 50 composed of either fixed rails 51 or movable rails 52 whose rail position moves, and a turnout 55 provided at a junction of the tracks 50, consisting of a movable rail 52 and a switch 56 that moves the movable rail 52.
[0016] "Fixed rail 51" is a rail that is fixed to sleepers (not shown) or the like by fastening devices (not shown), and is a rail that does not move due to a drive source such as a switch 56. Of the fixed rails 51, the fixed rail 51 that the movable rail 52 is in close contact with at a junction where a turnout 55 is installed will be referred to as fixed rail 51a for convenience of explanation. The "movable rail 52" is a rail that has the function of moving a vehicle such as an inspection vehicle 10 from one track 50 to another track 50 at a junction where a turnout 55 is installed, and its position is moved by a switch 56. "Track 50" refers to the vehicle track formed by rails, and the type and shape of the rails are not limited as long as they are formed by rails, and the rails can be fixed rails, movable rails, straight rails, curved rails, etc. "Multiple tracks 50 formed by combinations" can be any combination of rails as described above, and may include combinations of rails of the same type, combinations of rails of different types, combinations of rails of the same shape, combinations of rails of different shapes, etc. Specifically, this includes tracks formed by fixed rails 51 and fixed rails 51, tracks formed by fixed rails 51 and movable rails 52, etc. A "vehicle" is a vehicle with wheels, and in this embodiment, it is a vehicle that travels on a track formed by rails, such as a railway freight car or passenger car. The location where the rails are installed is not limited to the location where railway company freight cars and passenger cars run on tracks, but can be any location such as amusement facilities or factory facilities. A "junction point" is a location on the track where the track 50 on which the inspection vehicle 10 is traveling can be changed to another track 50, and is within the range in which the tip of the movable rail 52 can move. In this embodiment, it refers not only to a specific point on the track, but also to the range necessary to change the track 50 on which the inspection vehicle 10 is traveling, and the area necessary to change the track 50 on which the inspection vehicle 10 is traveling. The "turnout 55" is the entire device for branching a vehicle such as an inspection vehicle 10 from one track 50 to another track 50. In this embodiment, it consists of a movable rail 52, a switch 56 which is a device installed at the branching point that switches the rails and also serves as a drive source, and a turnout rod 57 which connects the movable rail 52 and the switch 56. "Track facilities" refers to the location where the track 50 is installed, and where at least fixed rails 51 are installed. At junctions, fixed rails 51, movable rails 52, and turnouts 55 are installed. Guard rails 54 are also installed near junctions. The "guard rails 54" are rails that have functions such as preventing unintended entry onto the track when a vehicle travels over a gap in the gauge of the track that occurs in the area where the turnouts 55 are installed.
[0017] Figure 6 is a conceptual diagram showing an example of the hardware configuration of the path determination system 100. The route determination system 100 consists of a processing unit 210, an imaging device 300, and an operation terminal 500. A positioning device 400 used to determine the position of the inspection vehicle 10 may also be provided. The arithmetic processing unit 210 is a so-called computer and includes a CPU (Central Processing Unit) 220, memory 230, input / output interface (I / F) 240, communication unit 250, etc. CPU220 refers to a so-called processor, and in addition to general CPUs, it may also include application-specific integrated circuits (ASICs), DSPs (Digital Signal Processors), GPUs (Graphics Processing Units), etc. Memory 230 consists of RAM (Random Access Memory), ROM (Read Only Memory), and auxiliary storage devices (such as hard disks). The input / output interface 240 can be connected to user interface devices such as a display device 260 and an input device 270. The display device 260 is a device that displays a screen corresponding to drawing data processed by the CPU 220, such as an LCD (Liquid Crystal Display) or CRT (Cathode Ray Tube) display. The input device 270 is a device that accepts user input such as a keyboard or mouse. The display device 260 and the input device 270 may be integrated and implemented as a touch panel. The communication unit 250 communicates with other computers via the communication network 600 and exchanges signals with other devices such as printers. Portable recording media may also be connected to the communication unit 250. In this embodiment, the communication unit 250 is connected to the operation terminal 500, imaging device 300, positioning device 400, etc., via the communication network 600. The hardware configuration of the path determination system 100 is not limited to the example in Figure 6. The arithmetic processing unit 210 may include other hardware elements not shown. The number of each hardware element is also not limited to the example in Figure 6. For example, the arithmetic processing unit 210 may have multiple CPUs. The arithmetic processing unit 210 may also be implemented by multiple computers consisting of multiple enclosures. Furthermore, the functions and performance of the arithmetic processing unit 210 described above may be provided to the operation terminal 500.
[0018] Figure 7 is a conceptual diagram showing an example of the software configuration of the course determination system 100. As shown in Figure 7, the arithmetic processing unit 210 includes an image acquisition unit 221, a moving rail position determination unit 222, a track determination unit 223, and a storage unit 231. It is also preferable to have a turnout identification unit 224, a turnout information extraction unit 225, a determined track determination unit 226, and an output processing unit (not shown) that outputs information to the operation terminal 500. Each of these processing modules is realized, for example, by the CPU 220 executing a computer program stored in the memory 230. This computer program may be installed via the input / output I / F 240 or communication unit 250 from a portable recording medium such as a DVD (Digital Versatile Disc) or memory card, or from another computer on a network, and stored in the memory 230.
[0019] The route determination system 100 includes a storage means for storing turnout information related to the turnout 55. In this embodiment, the turnout information is stored in the memory 230. The storage means corresponds to the storage unit 231 provided in the memory 230. "Turnout information" refers to information about turnout 55, and includes at least location information or an image of where turnout 55 is installed. Location information includes latitude and longitude, a certain number of kilometers, or information about the installation location such as "just before XX station," and the information may include either text data or image data. It may also include information about turnout 55 itself, such as its shape and type (model number, type, type, image showing its shape, etc.), or information that identifies turnout 55 (ID, serial number). Figure 8 shows an example of a database of turnout information to be stored in the memory unit 231. As shown in Figure 8, turnout information such as turnout ID, latitude and longitude, kilopoint, type, and image is stored in association with each turnout 55 where it is installed. Note that not all turnout information is stored for all turnouts 55; for example, turnout information such as latitude and longitude may be stored for all turnouts 55, but turnout type information may be stored for only some of the turnouts 55.
[0020] The route determination system 100 includes image acquisition means for acquiring spatial images, which are either moving images or multiple still images, of a target space including branching points. As will be described in detail later, in this embodiment the image acquisition means corresponds to the image acquisition unit 221 of the processing unit 210, and the image acquisition unit 221 is implemented by the CPU 220. The "target space" is the range captured by the imaging device 300, and in this embodiment, it is the range that includes the branching point. The branching point may be located at any position relative to the target space. A "spatial image" is an image captured by the imaging device 300, which may be a moving image or multiple still images. The imaging device 300 may be a general RGB camera or a monochrome camera, and there are no restrictions on the performance or specifications of the imaging device 300. "Acquiring" means taking spatial images (spatial image data) into the processing unit 210. The source and method of acquisition of spatial image data are not limited; for example, the spatial image data may be acquired into the processing unit 210 via a medium on which the spatial image data is recorded, or spatial image data captured by the imaging device 300 may be acquired into the processing unit 210 via the communication network 600 or wirelessly. In this embodiment, in order to sequentially determine the track condition using spatial image data captured while the inspection vehicle 10 is in motion, the spatial image data is acquired into the processing unit 210 via the communication network 600 or wirelessly so that it can be acquired sequentially. The acquired spatial images are stored in the memory 230.
[0021] The path determination system 100 includes a moving rail position determination means that analyzes a spatial image acquired by an image acquisition means and determines the position of the moving rail 52 in the target space. As will be described in detail later, in this embodiment the moving rail position determination means corresponds to the moving rail position determination unit 222 of the arithmetic processing unit 210, and the moving rail position determination unit 222 is implemented by the CPU 220. "The position of the moving rail in the target space" refers to the position of the moving rail 52 relative to a predetermined reference in the target space, where the predetermined reference can be something fixed (e.g., a fixed rail 51, a switch 56) or something absolute regarding position (e.g., latitude and longitude). "Determining the position" means determining the relative positional relationship with the predetermined reference, for example, determining the distance of the moving rail 52 to the predetermined reference, the area between the predetermined reference and the moving rail 52, the shape, etc. If the predetermined reference is absolute, it means determining the absolute position of the moving rail 52, for example, determining the latitude and longitude of the moving rail 52.
[0022] The route determination system 100 includes a track determination means that determines the track information of the vehicle passing through the junction based on the determination result of the moving rail position determination means and the turnout information. As will be described in detail later, in this embodiment the track determination means corresponds to the track determination unit 223 of the arithmetic processing unit 210, and the track determination unit 223 is implemented by the CPU 220. "Based on the determination result of the moving rail position determination means and the turnout information" means not only using both the determination result of the moving rail position determination means and the turnout information of the turnout 55 determined by the identified moving rail position determination means, but also using the determination result of the moving rail position determination means and the turnout information of multiple turnouts 55 stored in the storage unit 231 (sometimes called the "turnout information database"), and also changing the weighting of the other information depending on the content of one piece of information (for example, changing the weighting of the determination result and the turnout information depending on the content of the determination result), and changing the weighting of the determination result and the turnout information when making a determination, and using the determination result of the moving rail position determination means and the turnout information in a certain relationship (certain conditions) to determine the track information on which a vehicle passing through a junction will proceed. Furthermore, this includes not only using the determination results of the moving rail position determination means and the turnout information as they are, but also using the determination results of the moving rail position determination means and / or the turnout information to identify other information (details will be described later, but for example, a track-learned model) used to determine the track information to which a vehicle passing through a turnout point will proceed. "Track information" refers to information about the tracks 50 in the target space, and in particular, it includes information indicating the state of the tracks 50, information indicating the state of progress on the tracks 50, information indicating whether the track conditions allow vehicles such as inspection vehicles 10 to proceed, information indicating which tracks 50 are passable, information indicating the position of the movable rail 52 on the track, information indicating the state of the movable rail 52 on the track 50, and, if the tracks 50 to be traveled on are predetermined, whether the track conditions allow for travel on the predetermined tracks 50.
[0023] Figure 9 shows an example of the main flowchart of the route determination system. The determination of track information by the route determination system 100 is performed by one or more computers, such as the arithmetic processing unit 210 described above. The determination of track information by the route determination system 100 includes steps (S100) to (S300).
[0024] In step (S100), the processing unit 210 acquires spatial images (spatial image data) captured by the imaging device 300. The spatial images are captured at a frame rate of 5 to 30 fps, preferably 10 to 15 fps. If the frame rate is too low, there is a possibility of missing images, and if the frame rate is too high, the processing load becomes too high. Therefore, it is preferable to determine the frame rate considering the travel speed of the inspection vehicle 10 and the processing performance of the path determination system 100. The acquired spatial images are stored in the memory 230.
[0025] In step (S200), the processing unit 210 analyzes the spatial image acquired by the image acquisition means and determines the position of the moving rail 52 in the target space. The moving rail position determination process is performed for each frame of the acquired spatial image. For example, the following preprocessing steps 1) to 4) are performed on the acquired spatial image to identify the fixed rail 51 and the moving rail 52. 1) Sharpening process (a process that detects and enhances the contours of objects (e.g., fixed rail 51, movable rail 52) in a spatial image) 2) Edge processing (a filter that emphasizes areas where the brightness of an image changes abruptly) 3) Line segment detection process (process to detect line segments) 4) Line segment filtering process (Among the detected line segments, a process is performed to identify line segments that are continuous in the vertical direction in the spatial image as fixed rails 51 or movable rails 52.) Then, the position of the movable rail 52 is determined from the identified fixed rail 51 and movable rail 52. Furthermore, the processing is not limited to the above if the positions of the fixed rail 51 and the movable rail 52 can be identified by analyzing the acquired spatial image.
[0026] In this embodiment, the movable rail position determination means determines the position of the movable rail 52 based on the distance between the movable rail 52 and the fixed rail 51 that is created as the movable rail 52 moves. The method for determining the position of the movable rail 52 will be explained with reference to Figures 4 and 5.
[0027] As shown in Figures 4 and 5, at the track junction, the movable rail 52 consists of a left movable rail 52 (corresponding to the "left movable rail") and a right movable rail 52 (corresponding to the "right movable rail"). Figure 4 shows that inspection vehicle 10, entering from position A, proceeds in the direction of B after passing the junction, and Figure 5 shows that inspection vehicle 10, entering from position A, proceeds in the direction of C after passing the junction. The wheels of inspection vehicle 10 move while each is in contact with the inside of the rails. In Figure 4, the switch 56 moves the left movable rail 52 so that it contacts the left fixed rail 51, and the right movable rail 52 so that it moves away from the right fixed rail 51. Therefore, at position A, the left wheel 11a and the right wheel 11b are in contact with the inside of the left and right fixed rails 51, respectively. However, after passing the lower end of the movable rail 52, the left wheel 11a will contact the inside of the left movable rail 52, and the right wheel 11b will contact the inside of the right fixed rail 51, and so after passing the branching point, the train will proceed in the direction of B. In Figure 5, the switch 56 moves the left movable rail 52 away from the left fixed rail 51, and the right movable rail 52 into contact with the right fixed rail 51. Therefore, at position A, the left wheel 11a and the right wheel 11b are in contact with the inside of the left and right fixed rails 51, respectively. However, after passing the lower end of the movable rail 52, the left wheel 11a will be in contact with the inside of the left fixed rail 51, and the right wheel 11b will be in contact with the inside of the right movable rail 52, and so after passing the branching point, the train will proceed in the direction of C. Therefore, the moving rail position determination means determines the position of the moving rail 52 based on the distance between the moving rail 52 and the fixed rail 51 that occurs when the moving rail 52 moves from the identified fixed rail 51 and moving rail 52. In addition, the moving rail position determination means determines the position of the moving rail 52 in the target space for each of the left moving rail 52 and the right moving rail 52.
[0028] In this embodiment, the distance between the fixed rail 51 and the movable rail 52 in the left region near the lower end of the movable rail 52 (referred to as the region (a) enclosed by a dotted line for explanatory purposes) and the distance between the fixed rail 51 and the movable rail 52 in the right region near the lower end of the movable rail 52 (referred to as the region (b) enclosed by a dotted line for explanatory purposes) are determined, and the positions of the left and right movable rails 52 are determined accordingly. In this embodiment, if the distance between the fixed rail 51 and the movable rail 52 is greater than or equal to a predetermined value, it is determined to be in an "open state," and if it is less than the predetermined value, it is determined to be in a "closed state." An example of the movable rail determination process is shown in Figure 10.
[0029] As shown in Figure 10, if the gap between the left fixed rail 51 and the left movable rail 52 is determined to be "closed," and the gap between the right fixed rail 51 and the right movable rail 52 is determined to be "open," then it is determined that the turnout 55 is controlling the movable rail 52 to proceed in direction B (Case 1). Also, if the gap between the left fixed rail 51 and the left movable rail 52 is determined to be "open," and the gap between the right fixed rail 51 and the right movable rail 52 is determined to be "closed," then it is determined that the turnout 55 is controlling the movable rail 52 to proceed in direction C (Case 2). Furthermore, if the gap between the left fixed rail 51 and the left movable rail 52 is determined to be "open," and the gap between the right fixed rail 51 and the right movable rail 52 is determined to be "open," then it is determined to be a judgment error (Case 3). This indicates that this is an impossible combination for the turnout 55 to control the movable rail 52. Similarly, if the gap between the left fixed rail 51 and the left movable rail 52 is determined to be "closed," and the gap between the right fixed rail 51 and the right movable rail 52 is also determined to be "closed," this is an impossible combination for the control of the movable rail 52 by the turnout 55, and is therefore determined to be a judgment error (Case 4).
[0030] In step (S300), the arithmetic processing unit 210 determines the track information on which the inspection vehicle 10 passing through the junction will proceed, based on the determination result of the moving rail position determination means and the turnout information. In this embodiment, in step (S200), in the cases of Case 1 and Case 2 shown in Figure 10, the turnout information of the turnout 55 determined by the moving rail position determination means is extracted from the turnout information (turnout information database) stored in the storage unit 231. As described above, the turnout information includes information about the location where the turnout 55 is installed, information about the shape and type of the turnout 55, and information that identifies the turnout 55, and is not limited to text data but may also be image data. For example, the turnout information corresponding to the turnout 55 whose position on the moving rail 52 has been determined is extracted from the turnout information (turnout information database) stored in the storage unit 231 as follows. For example, a positioning device 400 is mounted on the inspection vehicle 10, and the position of the inspection vehicle 10, determined using the positioning device 400, is used to extract turnout information corresponding to the turnout 55 whose position on the moving rail 52 was determined in process (S200) from the turnout information (turnout information database) stored in the memory unit 231. This extraction method is the same as the turnout information extraction method described later. If point information (XX kilometer point) can be identified from the spatial image used to determine the position of the moving rail 52 in process (S200), then the point information is used to extract turnout information corresponding to the turnout 55 that determined the position of the moving rail 52 from the turnout information (turnout information database) stored in the memory unit 231. In process (S200), using the spatial image in which the position of the moving rail 52 was determined, the turnout information corresponding to the turnout 55 that determined the position of the moving rail 52 is extracted from the turnout information (image data of the turnout information database) stored in the memory unit 231. The above-described method for extracting turnout information is merely an example, and turnout information corresponding to the turnout 55 whose position on the movable rail 52 has been determined may be extracted using other information. Furthermore, in this embodiment, since the turnout information is used in process (S300), it is sufficient to extract the turnout information before using it. As will be described later, if the turnout information is also used in process (S200), it is sufficient to extract the turnout information before that. In addition, if the turnout information corresponding to the turnout 55 whose position on the movable rail 52 has been determined cannot be extracted from the turnout information (turnout information database) in the storage unit 231, that fact should be indicated in the turnout information (extraction result). The case in which the corresponding turnout information cannot be extracted from the turnout information (turnout information database) in the storage unit 231 is when the turnout information for the corresponding position is not stored in the storage unit 231, when the turnout information for the corresponding point information is not stored in the storage unit 231, or when the turnout information for the corresponding image data is not stored in the storage unit 231.
[0031] The processes (S200) and (S300) described above are shown in Figure 11. Figure 11 shows an example of a flowchart for the moving rail detection process and the track detection process. In process (S200), the arithmetic processing unit 210 performs processes (S210) and (S220). Step (S210) is a preprocessing step for the acquired image data, and for example, the preprocessing steps 1) to 4) described above is performed. Step (S220) determines the position of the movable rail 52 from the fixed rail 51 and movable rail 52 identified in the preprocessing step, and performs the case classification shown in Figure 10. In process (S300), the arithmetic processing unit 210 performs processes (S310) to (S340). If process (S310) determines that it is case 1 or case 2, the process proceeds to process (S320); if it determines that it is case 3 or case 4, the process does not proceed to process (S320). Alternatively, this process may be omitted, and processes (S320) and (S330) may be performed even in the case of case 3 and case 4, which are determined to be errors in the moving rail determination process shown in Figure 9. Step (S320) involves extracting turnout information corresponding to the turnout 55 determined in step (S220) from the turnout information (turnout information database) stored in the memory unit 231. Step (S330) performs a process to determine track information based on the determination result of step (S220) and the turnout information extracted in step (S320). If step (S220) determines that it is either case 1 or case 2, and the turnout 55 whose position of the movable rail 52 was determined in step (S220) is consistent with the turnout information extracted in step (S320) (for example, if the position of the turnout 55 including the movable rail 52 identified in step (S220) is consistent with the position of the turnout information extracted in step (S320)), the determination result of the movable rail determination means is set to true and the track information is determined. On the other hand, if the process (S220) determines that it is either Case 1 or Case 2, but the turnout 55 that determined the position of the moving rail 52 in the process (S220) does not match the turnout information extracted in the process (S320) (for example, if the position of the turnout 55 including the moving rail 52 identified in the process (S220) does not match the position of the turnout information extracted in the process (S320), or if it cannot be extracted from the turnout information (turnout information database) of the storage unit 231), then it is determined as "undetermined". In this embodiment, the track information of the vehicle passing through the junction is determined using both the determination result of the moving rail position determination means and the turnout information. Step (S340) is a process that notifies the notification means that the determination result of the moving rail determination means in step (S220) was an error, as well as the track information determined in step (S330). The content of the notification will be described later. Although this step is not required, it is preferable to include it considering the work efficiency and safety of the workers.
[0032] Figure 12 shows an example of the display of determined track information on the display screen of the operation terminal 500, and is a diagram showing only the display area (1) of the display screen shown in Figure 3. Figure 12(a) is an example of a display in which an arrow 510 is attached to the track to be traveled based on the determined track information, and Figure 12(b) is an example of a notification (display example) indicating that there is a judgment error in the cases of Case 3 and Case 4 of Figure 10. The display screen of the operational terminal 500 corresponds to the notification means. Figure 12(a) shows that the track on which the vehicle is traveling is indicated by an arrow 510, as track information determined by the track determination means. By informing the operation terminal 500 of the determined track information, workers can understand the condition of the tracks, thus preventing them from entering the wrong track. In this manner, the track on which the vehicle is traveling, as determined by the track determination means, is notified by the notification means (corresponding to "arrow 510").
[0033] Furthermore, Figure 12(b) shows that if the judgment results for the left moving rail and the judgment results for the right moving rail meet predetermined conditions, a notification of an abnormal judgment is sent via the character 511. In this embodiment, cases 3 and 4 correspond to the "predetermined conditions". By notifying the operation terminal 500 that there is an abnormal judgment, workers can be encouraged to thoroughly visually check the condition of the track, especially the position of the moving rails 52, thereby improving the safety of the inspection vehicle 10 during its operation. Thus, if the track determination means determines that the determination result for the left moving rail and the determination result for the right moving rail meet predetermined conditions, the notification means will notify that there is a determination abnormality (corresponding to "character 511").
[0034] Furthermore, if the process described above (S330) results in "unable to determine," the notification means may either notify the operator that "unable to determine" is the case, or it may not notify the operator at all. By not notifying the operator at all, it may appear to the operator that the track determination means has not performed any process to determine the track information on which the vehicle is traveling.
[0035] As described above, the route determination system 100 has various methods for extracting turnout information, but here we will describe a turnout identification means that identifies a turnout 55 based on position information acquired by a positioning device 400 mounted on an inspection vehicle 10, and a turnout information extraction means that extracts turnout information for the identified turnout 55. The positioning device 400 (not shown) is a so-called GNSS (Global Navigation Satellite System) antenna and is mounted, for example, on the flat area between the instrument 13 and the front glass 12 as shown in Figure 2(b). To obtain accurate position information, it is preferable to mount it in the central part of the inspection vehicle 10 in the left-right direction. The data acquired by the positioning device 400 is input to the processing unit 210 via a communication network 600 or the like. The "turnout identification means" is a means for identifying the position of the inspection vehicle 10 based on position information acquired by the positioning device 400, and for identifying the turnout 55 that was photographed at that position from the identification result. The position information of the inspection vehicle 10 at the time the turnout 55 is identified (for example, latitude and longitude) is used as the position information of the turnout 55 that was photographed at that position. However, for example, the position information of the turnout 55 may be a value corrected by the distance to the target space photographed by the imaging device 300 mounted on the inspection vehicle 10. The turnout identification means corresponds to the turnout identification unit 224 provided in the arithmetic processing unit 210 shown in Figure 7. The "turnout information extraction means" is a means for extracting turnout information for a identified turnout 55 based on the location information of the identified turnout 55. Based on the location information of the turnout 55 identified by the turnout identification means, the turnout information for the identified turnout 55 is extracted from the turnout information (turnout information database) stored in the storage unit 231 shown in Figure 8. If extraction is not possible, that fact should be indicated in the turnout information (extraction result). The case in which the corresponding turnout information cannot be extracted from the turnout information (turnout information database) in the storage unit 231 is when the turnout information for the corresponding location information is not stored in the storage unit 231. Note that the turnout information extraction means corresponds to the turnout information extraction unit 225 provided in the arithmetic processing unit 210 shown in Figure 7. In this way, by using the turnout information extracted using the position information acquired by the positioning device 400 when the track determination means determines the track information, it is possible to improve the determination accuracy.
[0036] As described above, examples have been explained in which the track on which the determined vehicle is traveling is notified by a notification means (Figure 12(a)), and in which a notification is given that there is a judgment abnormality when the judgment results of the left moving rail and the judgment results of the right moving rail meet predetermined conditions (Figure 12(b)). However, other information may also be notified. For example, if the information indicating the track information on which the inspection vehicle 10 is traveling, as determined by the track determination means, is information indicating the track on which the inspection vehicle 10 is traveling, a track determination means may be provided to perform the following processing. In inspection work, the track 50 on which the vehicle (inspection vehicle 10) is scheduled to travel is predetermined. The worker, after understanding the predetermined track 50 on which the vehicle (inspection vehicle 10) is scheduled to travel, stops or slows down the vehicle (inspection vehicle 10) before the turnout 55, visually checks the track condition, and confirms that the track 50 is in a condition suitable for the inspection vehicle 10 to travel on. The processing unit 210 acquires "work schedule information," which is information about the track on which the vehicle (inspection vehicle 10) is scheduled to travel, in advance and stores it in the memory 230. The processing unit 210 is then equipped with a track determination means that determines whether the track 50 on which the inspection vehicle 10 is scheduled to travel, as stored in the work schedule information, matches the track 50 on which the inspection vehicle 10 is scheduled to travel, as determined by the track determination means. If the determination result by the track determination means does not match, a notification means (for example, the display screen of the operation terminal 500) displays a message indicating that they do not match. Specifically, the (6) Matching Result Display Area on the display screen of the operation terminal 500 shown in Figure 3 will display a message indicating that the work schedule and the judgment result do not match. By displaying a message indicating that the work schedule and the judgment result do not match in this way, the safety of the inspection vehicle 10 while it is in motion can be enhanced. The line determination means corresponds to the line determination unit 226 provided in the arithmetic processing unit 210 shown in Figure 7. Furthermore, the determination process by the line determination means may be performed after step (S330) in the flowchart shown in Figure 11.
[0037] In the embodiment described above, the movable rail position determination means determines the position of the movable rail 52 by determining whether the distance between the identified fixed rail 51 and the movable rail 52, in particular the distance between the movable rail 52 and the fixed rail 51 near the lower end of the movable rail 52, is greater than or equal to a predetermined value. However, it is not limited to this. For example, the distance between the left movable rail 52 and the right fixed rail 51, and the distance between the left fixed rail 51 and the right movable rail 52 are specified for a predetermined section relative to the direction of travel of the inspection vehicle 10. If the distance between the left movable rail 52 and the right fixed rail 51 is within a certain range for the predetermined section, it is determined that the left movable rail 52 has not moved to a position where it is separated from the left fixed rail 51, and in Figure 4, it is determined that it is possible to proceed in direction B. Similarly, if the distance between the left fixed rail 51 and the right movable rail 52 is within a certain range for the predetermined section, it is determined that the right movable rail 52 has not moved to a position where it is separated from the right fixed rail 51, and in Figure 4, it is determined that it is possible to proceed in direction C. Furthermore, based on two determination results (for example, if the distance between the left movable rail 52 and the right fixed rail 51 is within a certain range for a predetermined section, and the distance between the left fixed rail 51 and the right movable rail 52 is not within a certain range for a predetermined section), it may be determined that the left movable rail 52 has not moved to a position where it is separated from the left fixed rail 51, and in the case of Figure 4, it may be determined that it is possible to proceed in direction B.
[0038] Furthermore, for example, the area between the left fixed rail 51 and the left movable rail 52 in a predetermined range near the lower end of the left movable rail 52 (for example, area (a) in Figure 4) (referred to as the rail-to-rail area of area (a)), and the area between the right fixed rail 51 and the right movable rail 52 in a predetermined range near the lower end of the right movable rail 52 (for example, area (b) in Figure 4) (referred to as the rail-to-rail area of area (b)), are identified. If the rail-to-rail area of area (a) is greater than or equal to the predetermined value, it is determined that the left movable rail 52 has moved to a position where it is separated from the left fixed rail 51, and in Figure 4, it is determined that it is possible to proceed in direction C. Also, if the rail-to-rail area of area (b) is greater than or equal to the predetermined value, it is determined that the right movable rail 52 has moved to a position where it is separated from the right fixed rail 51, and in Figure 4, it is determined that it is possible to proceed in direction B. Furthermore, based on the two determination results (for example, if the rail-to-rail area in region (a) is greater than or equal to a predetermined value, and the rail-to-rail area in region (b) is less than a predetermined value), it may be determined that the left movable rail 52 has moved to a position where it is separated from the left fixed rail 51, and in Figure 4, it may be determined that it is possible to proceed in direction C.
[0039] Furthermore, while the above-mentioned numerical values (whether or not they are greater than or equal to a predetermined value) such as the distance between the identified fixed rail 51 and the movable rail 52, or the area between rails within a predetermined range, were used as the judgment criteria, the shape of the distance between the fixed rail 51 and the movable rail 52, or the shape of the area between rails within a predetermined range, may also be used as the judgment criteria. In this case, an image showing the shape that serves as the judgment criterion may be stored, and the judgment may be made by comparing it with the image. Note that there may be only one type of image (one judgment image) used as the judgment criterion, but multiple types may also be provided. By doing so, it becomes possible to reliably determine the position of the movable rail 52 even when the imaging accuracy is poor.
[0040] Furthermore, the moving rail position determination means may determine the position of the moving rail 52 in the target space using a track-learned model obtained through machine learning. It is preferable to have multiple track-learned models for each turnout that satisfies the turnout conditions. An example model is shown below. a) Track-learned model of a turnout switch installed on a designated line. This is a track-trained model that has been trained using only images of turnouts 55 installed on a predetermined line. The turnout condition is "installed on a predetermined line". The images used for training are images of turnouts 55 installed on a predetermined line, and preferably images of the target space including the turnouts 55, taken by the inspection vehicle 10 while it is moving, not just of the turnouts 55. Furthermore, since inspection work is often carried out at night, it is preferable to include images taken at night, and also to include images taken under various weather conditions. b) Pre-programmed track models of a specified type of turnout. This is a track training model that has been trained using only images of a predetermined type of turnout 55. "Predetermined type" is the turnout condition. The images used for training are images of a predetermined type of turnout 55, and the location where the turnout 55 is installed can be anywhere. Since inspection work is often carried out at night, it is preferable to include images taken at night, and it is also preferable to include images taken under various weather conditions. c) Track learning model of a turnout installed at a predetermined location This is a track training model that has been trained using only images of turnouts 55 installed at predetermined locations. The turnout condition is "installed at predetermined locations". The images used for training are images of turnouts 55 installed at predetermined locations, and preferably images of the target space including the turnouts 55, as captured by the inspection vehicle 10, not just the turnouts 55 themselves. Furthermore, since inspection work is often carried out at night, it is preferable to include images taken at night, and also to include images taken under various weather conditions. Note that the track-trained models are not limited to those mentioned above. The moving rail position determination means then selects a track-learned model based on the turnout information and uses this track-learned model to determine the position of the moving rail 52 in the target space. As mentioned above, the turnout information can be extracted at any time as long as it is extracted before using the turnout information. In Figure 11, the turnout information is extracted before the track condition determination process (S330), but here, since the turnout information is used in the process of determining the position of the moving rail 52 (S220), it is necessary to extract the turnout information before that process. Then, the position of the moving rail in the target space is determined using the track-learned model selected based on the previously extracted turnout information. In this case, since the turnout information is extracted before the process of determining the position of the moving rail (S220) in the flowchart of Figure 11, process (S320) is not performed. In this way, even when switch information is used to identify other information (for example, a track-learned model) used to determine the track information to which a vehicle passing through a junction will proceed, and the determination result of the moving rail position determination means is calculated using this other information, the track information to which a vehicle passing through a junction will proceed is determined based on the determination result of the moving rail position determination means and the switch information, and since the position of the moving rail can be determined using appropriate information, the determination accuracy can be improved.
[0041] Alternatively, instead of selecting a track-learned model from the turnout information as described above, information (referred to as "learned model information") that specifies the track-learned model to be used by the moving rail position determination means for each turnout 55 may be stored in the memory 230 in advance, and the moving rail position determination means may select a track-learned model based on the learned model information and determine the position of the moving rail 52 in the target space. Thus, by having the moving rail position determination means determine the position of the moving rail 52 using a track-learned model obtained through machine learning, it is possible to determine the track information. Furthermore, by selecting the optimal track-learned model based on the turnout information of the turnout 55 to be determined and then determining the position of the moving rail 52, the determination accuracy can be improved.
[0042] Furthermore, when the moving rail position determination means uses a track-learned model obtained through machine learning to determine the position of the moving rail 52, the determination accuracy may be calculated using the track-learned model, and the determination accuracy of the determined track condition may be determined based on the calculated determination accuracy. In addition, the determined determination accuracy may be notified using a notification means, and for example, the notification method may be different depending on the determination accuracy. By notifying the determination accuracy in this way, workers can grasp the determination accuracy and improve work efficiency. Furthermore, if the determination accuracy is above a predetermined value, even if the turnout 55 whose position was determined in step (S220) does not match the turnout information extracted in step (S320), the determination result may be taken as true and used as the track condition determination result. In this case, it is preferable to notify the system that there was no match before using it as the track condition determination result. Thus, the way in which the other piece of information (turnout information) is used may be changed depending on the content of one piece of information (determination accuracy of the determination result of the moving rail position determination means).
[0043] <Variation> The above-described embodiment is merely an example and may be partially modified as appropriate.
[0044] In the above embodiment, the turnout 55 is a "single-opening turnout" as shown in Figures 4 and 5, but it is possible to similarly determine the track condition with other types of turnouts (for example, double-opening turnouts, split turnouts, inner turnouts, outer turnouts, overpass turnouts, diamond crossings, crossovers, sas crossings, single slip switches, double slip switches, three-rail turnouts, etc.). In any case of the type of turnout 55, the track condition determination process can be performed for each turnout 55 using the determination result of the position of the moving rail 52 and the turnout information. For example, when changing course to an adjacent track via multiple turnouts 55, such as in a sas crossing, the vehicle can temporarily stop or slow down at each turnout 55, determine the track condition, and then, based on the determination result, confirm the track 50 on which the inspection vehicle 10 is proceeding. This makes it possible to improve the safety of the inspection vehicle 10 during its operation.
[0045] In the above embodiment, the imaging device 300 continuously takes images during inspection work, and the processing unit 210 acquires all the captured data, but it is not limited to this. For example, the locations where the turnouts 55 are installed may be stored in the memory 230 of the processing unit 210 in advance, and the processing unit 210 may send a control signal to the imaging device 300 to take images only within a predetermined range including the locations where the turnouts 55 are installed, and the imaging device 300 may take images. Alternatively, an operator may manually take images using the imaging device 300. Furthermore, if a positioning device 400 is installed, the processing unit 210 may send a control signal to the imaging device 300 to take images only within a predetermined range including the locations where the turnouts 55 are installed, based on the data from the positioning device 400. By having the imaging device 300 take images only near the locations where the turnouts 55 are installed, it becomes possible to efficiently process the determination of track information.
[0046] In the above embodiment, as shown in the flowchart of Figure 11, the track information is determined by determining the position of the moving rail 52 and setting the result of the moving rail determination means as true when it matches the extracted turnout information (for example, when compared with the turnout information and it is the position where a turnout 55 is installed). However, it is not limited to this. For example, the position of the turnout 55 included in the spatial image to be analyzed may be identified from data from the positioning device 400 or information contained in the acquired image, and the position of the moving rail 52 may be determined by the moving rail determination means when it matches the identified position and the turnout information (for example, when compared with the turnout information and it is the position where a turnout 55 is installed). In this way, the position of the moving rail 52 is determined only when the position of the turnout 55 included in the spatial image to be analyzed is in the turnout information (turnout information database) stored in the storage unit 231, thus reducing the processing load.
[0047] In the above embodiment, if the track determination means determines that the track 50 on which the inspection vehicle 10 is scheduled to proceed, as stored in the work schedule information, does not match the track 50 on which the inspection vehicle 10 is scheduled to proceed, the (6) matching result display area on the display screen of the operation terminal 500 displays a message indicating that the work schedule and the determination result do not match. However, the embodiment is not limited to this. For example, in the (1) captured image display area on the display screen of the operation terminal 500 shown in Figure 3, the words "Matches Schedule" may be displayed along with a "○" for the track 50 to proceed and a "×" for the track 50 that should not be proceeded. Since the worker knows which track 50 to proceed on from the work schedule information, the display of "Matches Schedule" allows them to recognize that the work is proceeding according to schedule, thereby increasing the safety of the inspection vehicle 10 during its operation.
[0048] In the above embodiment, the notification means is the display screen of the operation terminal 500, but it is not limited to this. For example, the notification may be made by voice from a speaker provided in the operation terminal 500, or by a device other than the operation terminal 500. Any device and any notification method is acceptable as long as the worker can grasp the result of the track condition determination while the inspection vehicle 10 is running.
[0049] In the above embodiment, the determination result of the track information for each acquired spatial image (each frame) is reported to the notification means, but this is not limited to this. For example, if the same result is obtained for a predetermined number of consecutive frames, that result may be reported. By doing so, it is possible to prevent false reports due to errors in image analysis.
[0050] In the above embodiment, step (S330) was determined to be either case 1 or case 2 in step (S220), but was not determined to be "undeterminable" if it did not match the turnout information extracted in step (S320). However, this is not limited to this. For example, if step (S330) was determined to be "undeterminable," but the determination result by the track determination means performed after step (S330) matches, the determination result of the moving rail position determination means may be set to true. However, it is preferable to notify in a different manner than when step (S220) is determined to be either case 1 or case 2 and matches the turnout information extracted in step (S320). Doing so will lead to workers being able to perform visual confirmation more reliably, thereby improving the safety of the inspection vehicle 10 when it is in motion.
[0051] In the above embodiment, the track condition determination process in step (S330) was performed using information regarding the installation location of the turnout information to determine whether it was consistent with the extracted turnout information, but this is not limited to this. For example, multiple pieces of information included in the turnout information (for example, information regarding the installation location and information regarding the type of turnout 55) may be used, and the weight given to the determination result of the moving rail determination means may be varied depending on the type or number of consistent pieces of information. Also, when extracting turnout information from the turnout information (turnout information database) stored in the storage unit 231, multiple pieces of information may be used for extraction, not just one piece of information included in the turnout information.
[0052] In the above embodiment, the images used to train the track learning model were different depending on the location and type of turnout 55, but the system is not limited to this. For example, a track learning model (rainy weather track learning model) may be created using only images of turnouts 55 installed during rainy weather. The images used for training are preferably images of various types of turnouts 55 installed at various locations and on various lines. Also, since inspection work is often carried out at night, it is preferable to include images of rainy weather at night. For example, the learned model information described above may be stored to determine the position of the movable rail 52 based on the rainy weather track learning model in the case of rainy weather, and the learned model may be selected based on this information. In this way, it is possible to reliably determine the position of the movable rail 52 even in the case of rainy weather.
[0053] In the above embodiment, the condition of the railway track was determined, but a similar process may be applied to inspecting the condition of other railway equipment, such as the condition of level crossing devices (e.g., damage).
[0054] <1> Multiple tracks consisting of a combination of fixed rails or movable rails whose position moves, A route determination system for determining the track information of a vehicle passing through a track junction, in a track facility having a turnout consisting of a movable rail and a switch for moving the movable rail, A storage means for storing information about the aforementioned turnout, Image acquisition means for acquiring a spatial image which is a moving image or multiple still images of the target space including the aforementioned branching point, A moving rail position determination means analyzes the spatial image acquired by the image acquisition means and determines the position of the moving rail in the target space, A route determination system comprising: a track determination means that determines track information on which the vehicle passing through the branch point will travel, based on the determination result of the moving rail position determination means and the turnout information. <2> The aforementioned turnout information includes at least location information or an image of where the turnout is installed. <1> The career path determination system described above. <3> An imaging device mounted on the vehicle for capturing the spatial image, A positioning device mounted on the aforementioned vehicle, A turnout identification means for identifying the turnout based on the position information acquired by the positioning device, The system further comprises a turnout information extraction means for extracting turnout information of a turnout based on the identified turnout. <1> or <2> The career path determination system described above. <3-1> When the turnout information extraction means extracts turnout information for the identified turnout, the track determination means determines the track information on which the vehicle passing through the turnout point will proceed, based on the determination result of the moving rail position determination means and the turnout information. <3> The career path determination system described above. <3-2> If the turnout information extraction means does not extract turnout information for the identified turnout, the track determination means does not perform a determination process to determine the track information on which the vehicle passing through the turnout point will proceed, based on the determination result of the moving rail position determination means and the turnout information. <3> The career path determination system described above. <4> The moving rail position determination means is characterized by determining the position of the moving rail based on the distance between the moving rail and the fixed rail that is created as a result of the moving rail's movement. <1> from <3> A route determination system described in one of the following: <4-1> The moving rail position determination means is characterized by determining the position of the moving rail based on the area between the moving rail and the fixed rail that is created when the moving rail moves within a predetermined area including the branching point. <1> A path determination system described in any one of the following <3-2>. <5> Equipped with a notification mechanism, The track on which the vehicle is traveling, as determined by the track determination means, is notified by the notification means. <1> from <4> A route determination system described in one of the following: <6> The information indicating the track information on which the vehicle is traveling, as determined by the track determination means, is information indicating the track on which the vehicle is traveling. The system includes a track determination means for determining whether the predetermined track on which the vehicle is scheduled to travel matches the track on which the vehicle is scheduled to travel, as determined by the track determination means. If the determination result by the determination line determination means does not match, the notification means will notify that it does not match. <5> The career path determination system described above. <7> At the aforementioned track junction, the movable rail consists of a left movable rail and a right movable rail. The moving rail position determination means determines the position of each of the left and right moving rails in the target space, The track determination means is characterized in that, if the determination result of the left moving rail and the determination result of the right moving rail meet predetermined conditions, it will notify that there is a determination abnormality using the notification means. <5> The career path determination system described above. <7-1> The predetermined condition is characterized in that the determination result for the left moving rail and the determination result for the right moving rail are the same. <7> The career path determination system described above. <8> The moving rail position determination means determines the position of the moving rail in the target space using a track-learned model obtained by machine learning. The aforementioned track learning model includes multiple models for each turnout that satisfies the turnout conditions. The moving rail position determination means is characterized by selecting a track learning model based on the turnout information and determining the position of the moving rail in the target space using the track learning model. <1> A path determination system described in any one of the following <7-1>. <8-1> The aforementioned turnout condition is characterized by being the location where the turnout is installed. <8> The career path determination system described above. [Explanation of symbols]
[0055] 10 Inspection Vehicles 11a wheels 11b Wheel 12 Front glass 13 Instruments 50 railroad tracks 51 Fixed rail 51a Fixed rail 52 Moving Rails 54 Guardrail 55 Branch switch 56 Switch 57 Switch pole 100 Career Path Determination System 210 Arithmetic Processing Unit 220 CPU 221 Image acquisition unit 222 Moving rail position determination unit 223 Track detection unit 224 Specific section of the turnout switch 225 Turnout Information Extraction Unit 226 Determination track determination unit 230 memory 231 Storage section 260 Display device 270 Input devices 300 Imaging devices 400 positioning devices 500 operational terminals 600 communication network
Claims
1. Multiple tracks consisting of a combination of fixed rails or movable rails whose position moves, A route determination system for determining the track information of a vehicle passing through a track junction, in a track facility having a turnout consisting of a movable rail and a switch for moving the movable rail, A storage means for storing information about the aforementioned turnout, Image acquisition means for acquiring a spatial image which is a moving image or multiple still images of the target space including the aforementioned branching point, A moving rail position determination means analyzes the spatial image acquired by the image acquisition means and determines the position of the moving rail in the target space, A route determination system comprising: a track determination means that determines track information on which the vehicle passing through the branch point will travel, based on the determination result of the moving rail position determination means and the turnout information.
2. The route determination system according to claim 1, wherein the turnout information includes at least location information or an image of the turnout being installed.
3. An imaging device mounted on the vehicle for capturing the spatial image, A positioning device mounted on the aforementioned vehicle, A turnout identification means for identifying the turnout based on the position information acquired by the positioning device, The route determination system according to claim 1 or 2, further comprising: a turnout information extraction means for extracting turnout information of a turnout based on the identified turnout.
4. The path determination system according to claim 1 or 2, characterized in that the moving rail position determination means determines the position of the moving rail based on the distance between the moving rail and the fixed rail that is created as a result of the moving rail moving.
5. Equipped with a notification mechanism, The route determination system according to claim 1 or 2, characterized in that the route on which the vehicle is traveling, as determined by the route determination means, is notified by the notification means.
6. The information indicating the track information on which the vehicle is traveling, as determined by the track determination means, is information indicating the track on which the vehicle is traveling. The system includes a track determination means for determining whether the predetermined track on which the vehicle is scheduled to travel matches the track on which the vehicle is scheduled to travel, as determined by the track determination means. The route determination system according to claim 5, characterized in that if the determination results from the determination means do not match, the notification means notifies that the results do not match.
7. At the aforementioned track junction, the movable rail consists of a left movable rail and a right movable rail. The moving rail position determination means determines the position of each of the left and right moving rails in the target space, The route determination system according to claim 5, characterized in that the track determination means notifies that there is a determination abnormality when the determination result of the left moving rail and the determination result of the right moving rail meet predetermined conditions.
8. The moving rail position determination means determines the position of the moving rail in the target space using a track-learned model obtained by machine learning. The aforementioned track learning model includes multiple models for each turnout that satisfies the turnout conditions. The route determination system according to claim 1 or 2, characterized in that the moving rail position determination means selects a track learning model based on the turnout information and determines the position of the moving rail in the target space using the track learning model.