Track detection system and program

Abstract

This invention provides a track detection system and program that can accurately detect railway tracks. [Solution] The track detection system 10 includes an acquisition unit (arithmetic unit 12) that acquires an image of the area in front of the vehicle in the direction of travel, taken by an imaging unit 11 provided on the vehicle traveling on the track; a first detection unit (arithmetic unit 12) that performs a first detection process to detect a track in a first image region R1 near the vehicle from the image acquired by the acquisition unit; and a second detection unit (arithmetic unit 12) that performs a second detection process to detect a distant track that is continuous with the track detected by the first detection unit, with the first detection process by the first detection unit and the second detection process by the second detection unit being performed independently.

Description

Technical Field

[0001] The present invention relates to a track detection system and a program for detecting a track of a railway track.

Background Art

[0002] For the safe and smooth operation of railway vehicles, it is necessary to recognize obstacles in front of the vehicle traveling direction at an early stage and start deceleration and stop. In recent years, the development of automatic driving control has also been promoted in railway vehicles. In order to realize automatic driving, it is necessary to automatically detect obstacles in front of the vehicle and perform vehicle control such as deceleration and stop as needed. At this time, in order to selectively detect only obstacles that interfere with operation, it is essential to accurately determine whether the obstacle exists within the building limit set based on the track. In view of the above circumstances, it can be said that it is important to detect the track in front of the vehicle with high accuracy.

[0003] In relation to this, Patent Document 1 describes "a rail detection device that uses two types of cameras, a wide-angle camera and a telephoto camera, as a front monitoring device, narrows down the track pattern of the rail vanishing point, and eliminates false detection of the location to be imaged."

Prior Art Documents

Patent Documents

[0004]

Patent Document 1

Summary of the Invention

Problems to be Solved by the Invention

[0005] However, in the invention described in Patent Document 1, when there are a plurality of tracks in front of the host vehicle, there is a possibility that a track other than the track on which the host vehicle is traveling is detected as the route.

[0006] This invention was made against the background described above, and its purpose is to provide a track detection system and program that can detect tracks with high accuracy. [Means for solving the problem]

[0007] To solve the above problems, the track detection system described in claim 1 is: An acquisition unit that acquires an image of the area in front of the vehicle in the direction of travel, which is captured by an imaging unit installed on the vehicle as it travels on the railway tracks, A first detection unit performs a first detection process to detect a railway track in a first image region near the vehicle from the image acquired by the acquisition unit, A second detection unit performs a second detection process to detect a distant railway line that is continuous with the railway line detected by the first detection unit, Equipped with, The first detection process performed by the first detection unit and the second detection process performed by the second detection unit are performed independently.

[0008] The invention described in claim 2 is a track detection system according to claim 1, The first detection unit executes the first detection process at a higher frequency than the second detection process performed by the second detection unit. The first detection process by the first detection unit and the second detection process by the second detection unit are performed in parallel.

[0009] The invention described in claim 3 is a track detection system according to claim 1, The first detection unit sets the first image region based on the result of the first detection process in the image captured immediately before the image.

[0010] The invention described in claim 4 is a track detection system according to claim 1, The second detection unit sets a second image region in the image based on the result of the first detection process and detects the railway line in the second image region.

[0011] The invention according to claim 5 is the line detection system according to claim 1, wherein the second detection unit sets a second image area in the image based on the result of the second detection process in the image captured immediately before the image, and detects a line in the second image area.

[0012] The program according to claim 6 is to cause a computer to function as an acquisition unit that acquires an image in front of the traveling direction of the vehicle captured by an imaging unit provided in the vehicle traveling on a line, a first detection unit that executes a first detection process for detecting a line in a first image area near the vehicle among the images acquired by the acquisition unit, a second detection unit that executes a second detection process for detecting a distant line continuous with the line detected by the first detection unit, and to perform the first detection process by the first detection unit and the second detection process by the second detection unit independently.

Advantages of the Invention

[0013] According to the present invention, a line can be detected with high accuracy.

Brief Description of the Drawings

[0014] [Figure 1] It is a system configuration diagram of the line detection system according to the present embodiment. [Figure 2] It is a flowchart showing an example of the procedure of the first detection process. [Figure 3] It is a flowchart showing an example of the procedure of the second detection process. [Figure 4] It is a diagram showing an example of the first image area set in the frame image. [Figure 5] It is a diagram showing an example of the second image area set in the frame image. [Figure 6] It is a diagram showing an example of independently and parallelly executing the first detection process and the second detection process. [Figure 7]It is a diagram showing an example of a second image area when the line is in a straight shape. [Figure 8] It is a diagram showing an example of a second image area when the line is in a curved shape. [Figure 9] It is a flowchart showing an example of the procedure of an obstacle detection process. [Figure 10] It is a diagram showing an example of a building limit in a railway track.

Embodiments for Carrying out the Invention

[0015] <Configuration of the Track Detection System> Hereinafter, an embodiment of a track detection system according to the present invention will be described with reference to the drawings. FIG. 1 is a block diagram showing a configuration example of a track detection system mounted on a railway vehicle (train). As shown in FIG. 1, a track detection system 10 according to the present embodiment includes an imaging unit 11 installed at the front end side of the leading vehicle so that the lens faces forward and captures an image of the front in the traveling direction (progressing direction), an arithmetic unit 12 that performs processing of the image captured by the imaging unit 11, a tachogenerator 13 that detects the number of rotations of the wheels to calculate the traveling speed and the self-position, an on-vehicle unit 14 that receives position correction information from a transponder ground unit, a communication device 15 that receives information transmitted from an operation management system and a security system via digital radio or a track circuit, a storage device 16 that stores information used when the arithmetic unit 12 processes the captured image, and the like. Furthermore, although not shown, the track detection system 10 may include a GPS device that receives a GPS signal from a GPS satellite and a monitor that displays an image captured by the imaging unit 11.

[0016] The storage device 16 stores information on roadside facilities related to the track of the route on which the vehicle travels, information on building limits, and the like. Also, the storage device 16 may store the ID (identification number) or train number of the vehicle.

[0017] The arithmetic unit 12 consists of a microprocessor (MPU), a non-volatile memory device such as ROM that stores the program executed by the MPU, and a read / write memory device such as RAM. The arithmetic unit 12 has image processing capabilities, as well as a function to calculate its own position (kilometers) based on signals from the speed generator 13 (and, if equipped with a GPS device, a function to convert latitude and longitude information obtained from GPS signals into kilometers). However, the kilometers calculated based on signals from the speed generator 13 and latitude and longitude information obtained from GPS signals contain errors. Therefore, the arithmetic unit 12 has a function to correct its own position using position correction information received from the transponder ground unit by the on-board unit 14.

[0018] Along railway lines, safety systems are installed that have the function of transmitting signals from ATC ground equipment to trains and controlling signals and points to determine the position of trains running based on signals from track circuits, etc., and to prevent them from getting too close to preceding trains. In addition, existing railway systems also have operation management systems that have the function of determining the position and route of trains, monitoring whether they are running according to the timetable, and providing route information to the safety system. The track detection system 10 of this embodiment is not particularly limited, but it is configured to obtain information about the position of the vehicle from a safety system or operation management system via a communication device 15 using a vehicle ID or train number, thereby determining the route on which the vehicle is traveling and the direction of travel, and to be able to pre-read information about facilities along the track ahead in the direction of travel from a storage device 16.

[0019] <Functions of the track detection system> Next, the functions of the track detection system 10 of this embodiment will be described. Figure 2 shows a flowchart of the procedure for the first detection process performed by the track detection system 10. Figure 3 shows a flowchart of the procedure for the second detection process performed by the track detection system 10. The track detection system 10 executes the first detection process and the second detection process independently and in parallel. The first detection process in Figure 2 and the second detection process in Figure 3 are initiated when the power to the track detection system 10 is turned on. When the power to the track detection system 10 is turned on, power is supplied to the imaging unit 11, and imaging begins. Imaging by the imaging unit 11 may be initiated in conjunction with the movement of the train. The imaging unit 11 captures multiple frame images at a predetermined frame rate (for example, 10 fps) and sequentially outputs the captured frame images to the processing unit 12.

[0020] First, let's explain the first detection process shown in Figure 2. When the first detection process begins, the computing unit 12 reads the data of the frame image (current frame image) captured at the present moment by the imaging unit 11 (step A1). In other words, the computing unit 12 acquires an image of the area in front of the vehicle in the direction of travel, which is captured by the imaging unit 11 installed on the vehicle traveling on the railway tracks. The computing unit 12 functions as an acquisition unit.

[0021] Next, the computing unit 12 obtains from the storage device 16 the result of the first image region line detection process, described later, in the frame image (previous frame image) captured by the imaging unit 11 immediately before the current frame image (step A2). If the memory device 16 does not contain the result of the line detection process within the first image region in the immediately preceding frame image, the arithmetic unit 12 omits step A2. Next, the arithmetic unit 12 sets the first image region within the current frame image based on the result of the line detection process within the first image region in the immediately preceding frame image acquired in step A2 (step A3).

[0022] Figure 4 shows an example of the first image region R1 set in the current frame image. The dashed lines shown in Figure 4 represent the results of the track detection process within the first image region in the immediately preceding frame image. In step A3, the arithmetic unit 12 sets a first image region R1 within the current frame image that includes the result of the track detection process within the first image region of the previous frame image and is located near the vehicle. The size of the first image region R1 is estimated to be the size within which the transmission lines included in the current frame image are contained. The computing unit 12 can set the size of the first image region R1 to be smaller than the default value if the frame rate during imaging by the imaging unit 11 is high and the time difference between the time the current frame image was captured and the time the previous frame image was captured is less than a predetermined value. If step A2 is omitted, the arithmetic unit 12 sets the first image region R1 at a predetermined position within the current frame image. This predetermined position is, for example, the center in the left-right direction of the current frame image and also the lower end.

[0023] Next, the arithmetic unit 12 performs a first image region track detection process to detect the track within the first image region R1 set in step A3 (step A4). In other words, the arithmetic unit 12 performs a first detection process to detect the track in the first image region R1 near the vehicle, among the images taken by the imaging unit 11 in front of the vehicle in the direction of travel. The arithmetic unit 12 functions as a first detection unit.

[0024] Next, the arithmetic unit 12 associates the results of the first image region line detection process performed in step A4 with the current frame image and saves them to the storage device 16 (step A5). Next, the arithmetic unit 12 determines whether or not the power to the line detection system 10 has been turned off (step A6). If the power to the track detection system 10 is turned off (step A6; YES), the arithmetic unit 12 terminates the first detection process. On the other hand, if the power to the track detection system 10 is not turned off (step A6; NO), the arithmetic unit 12 moves the first detection process to step A1 and reads the next frame image captured by the imaging unit 11.

[0025] In the first detection process described above, the track detection system 10 sets a first image region within the current frame image based on the result of the track detection process within the first image region in the immediately preceding frame image, and detects the track within the first image region. This suppresses the misidentification of structures and other objects located near the track as part of the track. Furthermore, since the range in which the track detection system 10 performs the track detection process can be limited to the first image region, the processing time for detecting the track can be shortened.

[0026] The track detection system 10 executes steps A1 to A6 of the first detection process at predetermined first intervals. The predetermined first interval is, for example, X [msec].

[0027] Next, we will explain the second detection process shown in Figure 3. When the second detection process is started, the arithmetic unit 12 obtains the results of the first image region line detection process that has been executed at that time (step B1). Next, the arithmetic unit 12 reads the frame image to which the results of the line detection process within the first image region acquired in step B1 are associated (step B2). Next, the arithmetic unit 12 sets a second image region within the frame image read in step B2 (step B3).

[0028] Figure 5 shows an example of the second image region R2 set in the frame image. In step B3, the arithmetic unit 12 sets the second image region R2 above the first image region R1 within the frame image. In other words, the arithmetic unit 12 sets the second image region R2 at a position within the frame image that is estimated to include a distant line that is continuous with the line detected by the line detection process within the first image region.

[0029] Next, the arithmetic unit 12 performs a second image region line detection process to detect lines within the second image region R2 set in step B3 (step B4). In other words, the arithmetic unit 12 performs a second detection process to detect distant lines that are continuous with the lines detected by the first detection unit. The arithmetic unit 12 functions as a second detection unit.

[0030] The arithmetic unit 12 may also divide the second image region R2 into multiple image regions R21 to R2n, as shown in Figure 5. In this case, in steps B3 and B4, the arithmetic unit 12 sets image region R21 in the frame image at a position above the first image region R1 and estimated to include a line continuous with the line detected by the line detection process within the first image region. Next, the arithmetic unit 12 detects a line within image region R21. Next, the arithmetic unit 12 sets image region R22 in the frame image, above image region R21, at a position where it is estimated to include a transmission line that is continuous with the transmission line detected within image region R21. Next, the arithmetic unit 12 detects a transmission line within image region R22. Subsequently, the arithmetic unit 12 performs the same processing for image regions R23 to R2n.

[0031] Next, the arithmetic unit 12 associates the results of the second image region line detection process performed in step B4 with the frame image and saves them to the storage device 16 (step B5). Next, the arithmetic unit 12 determines whether or not the power to the line detection system 10 has been turned off (step B6). If the power to the track detection system 10 is turned off (step B6; YES), the arithmetic unit 12 terminates the second detection process. On the other hand, if the power to the track detection system 10 is not turned off (step B6; NO), the arithmetic unit 12 proceeds to step B1 for the second detection process.

[0032] The track detection system 10 executes steps B1 to B6 of the second detection process at predetermined second intervals. The predetermined second interval is, for example, Y [msec]. Here, X <Yとする。

[0033] Figure 6 shows an example in which the first detection process and the second detection process are executed independently and in parallel. In the example shown in Figure 6, the track detection system 10 performs a first detection process on each frame image. Meanwhile, the track detection system 10 performs a second detection process on frame image 1 based on the results of the first image region track detection process corresponding to frame image 1. After the second detection process on frame image 1 is completed, the track detection system 10 performs a second detection process on frame image 5, for example, based on the results of the first image region track detection process corresponding to frame image 5. After the second detection process on frame image 5 is completed, the track detection system 10 performs a second detection process on frame image 9 based on the results of the first image region track detection process corresponding to frame image 9.

[0034] The frequency of the first detection process is determined based on the time required for the first detection process by the track detection system 10, and the frequency of the second detection process is determined based on the time required for the second detection process by the track detection system 10. The track detection system 10 needs to perform the first detection process more frequently than the second detection process. Therefore, the size of the first image region is set to be smaller than the size of the second image region. By performing the first detection process at a high frequency, the probability that the railway tracks are included within the first image region R1 shown in Figure 4 increases, thereby suppressing the failure to detect the tracks. As a result, it becomes possible to detect railway tracks located far from the vehicle. Therefore, even when distant tracks are unclear in frame images taken under low-light conditions such as backlighting or at night, when the tracks curve sharply ahead, or when there are multiple tracks ahead of the vehicle, the tracks can be reliably detected, improving the tracking performance in track detection.

[0035] <Variation> Next, a modified example of the track detection system 10 according to this embodiment will be described.

[0036] (Variation 1) In the first image region line detection process (step A4) of the modified example 1, the arithmetic unit 12 detects lines within the first image region R1 and determines whether the detected lines are straight or curved. If the detected railway line in the first image region railway line detection process is determined to be a straight line, in step B3 of the second detection process, the arithmetic unit 12 sets a second linear image region R2 within the frame image that is presumed to contain a straight railway line (see Figure 7). On the other hand, if the track detected in the first image region track detection process is determined to be curved, in step B3 of the second detection process, the arithmetic unit 12 sets a second image region R2 within the frame image that corresponds to the curve radius of the track and is estimated to contain a curved track (see Figure 8). As a result, if the detected railway line in the first image region is determined to be straight, the second image region R2 can be limited to the straight-line area, thereby reducing the burden on the railway line detection process in the second image region. Furthermore, the arithmetic unit 12, acting as the second detection unit, can also set a second image region in the current frame image based on the result of the second detection process in the frame image (previous frame image) captured immediately before the current frame image by the imaging unit 11. In the second image region, changes in the track shape within the second image region can be observed from a distance. Due to the characteristic that the track shape does not change abruptly, the difference between the track shape in the previous frame image and the track shape in the current frame image becomes small. Therefore, by setting the second image region in the current frame image based on the results of the second detection process in the previous frame image, the burden on the track detection process within the second image region can be reduced.

[0037] (Modification 2) In step B3 of the second detection process in Modification 2, the arithmetic unit 12 can set the second image region R2 based on the track information stored in the storage device 16. This allows the second image region R2 to be set appropriately for image regions with a high probability of containing tracks, even in situations where the track curves sharply ahead or in special track configurations such as level crossings and junctions. Compared to setting the second image region R2 over a wide area, this reduces the burden on the track detection process within the second image region in step B4.

[0038] <Other> This section describes how to apply the track detection results in this embodiment to the obstacle detection system in front of the train. Figure 9 shows a flowchart of the obstacle detection process performed by the track detection system 10.

[0039] When the obstacle detection process is started, the computing unit 12 obtains the results of the first image region line detection process and the second image region line detection process from the storage device 16 as line detection results (step C1). Next, the arithmetic unit 12 sets the building clearance based on the track detection results obtained in step C1 (step C2). In railways, to ensure the operation of trains and the safety of passengers and staff, building clearances are set by railway operators based on ministerial ordinances, which are spatial restrictions on the installation of buildings, track equipment such as signals, temporary construction structures, trees, etc. The building clearance CG, as shown in Figure 10, is based on the center of the track (railway) and allows for a slight margin in the vehicle clearance, which is the standard for determining the cross-section of a train. It is set within a plane perpendicular to the track. Therefore, the actual building clearance takes the form of a three-dimensional tube shape along the direction of track extension.

[0040] Next, the arithmetic unit 12 detects an obstacle within the building clearance set in step C2 (step C3), and terminates the obstacle detection process. As described above, by using the high-precision track detection results of the present invention (results of track detection processing within the first image area and results of track detection processing within the second image area) in the obstacle detection process, it is possible to accurately determine whether or not the obstacle is located within the building clearance set based on the track, and to detect obstacles only within the building clearance. This makes it possible to suppress the detection of lookouts and other objects on the track that are not the target of detection.

[0041] <Effects> As described above, the track detection system 10 according to this embodiment includes: an acquisition unit (arithmetic unit 12) that acquires an image of the area in front of the vehicle in the direction of travel, captured by an imaging unit 11 provided on the vehicle traveling on the track; a first detection unit (arithmetic unit 12) that performs a first detection process to detect a track in a first image region R1 near the vehicle from the image acquired by the acquisition unit; and a second detection unit (arithmetic unit 12) that performs a second detection process to detect a distant track that is continuous with the track detected by the first detection unit. The first detection process by the first detection unit and the second detection process by the second detection unit are performed independently. Therefore, the tracking performance of track detection in the first image region R1 near the vehicle can be improved, preventing the track from being undetected and ensuring reliable track detection. This allows for reliable detection of railway tracks even in low-light conditions such as backlighting or nighttime when distant tracks are unclear in frame images, when the tracks curve sharply ahead, or when multiple tracks are present in front of the vehicle. Therefore, railway tracks can be detected with high accuracy.

[0042] In the line detection system 10 according to this embodiment, the first detection unit (arithmetic unit 12) executes the first detection process at a higher frequency than the second detection process performed by the second detection unit (arithmetic unit 12). The first detection process by the first detection unit (arithmetic unit 12) and the second detection process by the second detection unit (arithmetic unit 12) are performed in parallel. Therefore, by performing the first detection process at a high frequency, it is possible to prevent the failure to detect tracks and ensure reliable detection of tracks. This makes it possible to reliably detect tracks located far from the vehicle. Furthermore, since the first detection process and the second detection process can be executed simultaneously, the tracking performance of track detection can be improved even when the shape of the track included in the image captured by the imaging unit 11 changes over time.

[0043] In the track detection system 10 according to this embodiment, the first detection unit (arithmetic unit 12) sets the first image region R1 based on the result of the first detection process in the image taken immediately before the aforementioned image. Therefore, it is possible to suppress the misidentification of structures located near the tracks as part of the railway line. In addition, since the range in which the track detection process is performed can be limited to the first image region, the processing time for detecting the tracks can be shortened.

[0044] In the line detection system 10 according to this embodiment, the second detection unit (arithmetic unit 12) sets a second image region R2 in the image based on the result of the first detection process and detects the line in the second image region R2. Therefore, in the first detection process, it is possible to reliably detect tracks that are continuous with the track detected, and to reliably detect tracks even at a distance.

[0045] In the line detection system 10 according to this embodiment, the second detection unit (arithmetic unit 12) sets a second image region R2 in the image based on the result of the second detection process in the image taken immediately before the image, and detects the line in the second image region R2. Therefore, the burden on the line detection process within the second image region can be reduced.

[0046] Although embodiments of the present invention have been described above, the present invention is not limited to the above embodiments, and various modifications and changes are possible. For example, although the above embodiment states that the arithmetic unit 12 functions as an "acquisition unit," a "first detection unit," and a "second detection unit," it is not limited to this. External devices installed outside the vehicle and connected to the arithmetic unit 12 in a communicative manner may also function as an "acquisition unit," a "first detection unit," and a "second detection unit." [Explanation of symbols]

[0047] 10. Track detection system 11 Imaging Unit 12. Processing Unit (Acquisition Unit, First Detection Unit, Second Detection Unit) 13 Speed ​​Generator 14. Onboard unit 15. Communication equipment 16 Storage device

Claims

1. An acquisition unit that acquires an image of the area in front of the vehicle in the direction of travel, which is captured by an imaging unit installed on the vehicle as it travels on the railway tracks, A first detection unit performs a first detection process to detect a railway track in a first image region near the vehicle from the image acquired by the acquisition unit, A second detection unit performs a second detection process to detect a distant railway line that is continuous with the railway line detected by the first detection unit, Equipped with, A track detection system that independently performs the first detection process by the first detection unit and the second detection process by the second detection unit.

2. The first detection unit executes the first detection process at a higher frequency than the second detection process performed by the second detection unit. The track detection system according to claim 1, wherein the first detection process by the first detection unit and the second detection process by the second detection unit are performed in parallel.

3. The track detection system according to claim 1, wherein the first detection unit sets the first image region based on the result of the first detection process in the image taken immediately before the image.

4. The track detection system according to claim 1, wherein the second detection unit sets a second image region in the image based on the result of the first detection process and detects a track in the second image region.

5. The track detection system according to claim 1, wherein the second detection unit sets a second image region in the image based on the result of the second detection process in the image taken immediately before the image, and detects a track in the second image region.

6. Computers An acquisition unit that acquires an image of the area in front of the vehicle in the direction of travel, which is captured by an imaging unit installed on the vehicle as it travels on the railway tracks. A first detection unit performs a first detection process to detect a railway track in a first image region near the vehicle from the image acquired by the acquisition unit. A second detection unit performs a second detection process to detect a distant railway line that is continuous with the railway line detected by the first detection unit. To make it function as, A program that independently performs the first detection process by the first detection unit and the second detection process by the second detection unit.

Images