JP2019191763A - Vehicle detection device and vehicle detection method - Google Patents

Abstract

To improve accuracy of detecting the presence/absence of a vehicle along a platform at a railroad station.SOLUTION: An image processing device 1 according to an embodiment comprises: a straight line detection unit 112 for detecting a straight line included in a photographic image captured by an imaging unit which images an area including a railroad track on which a vehicle runs; a vehicle determination unit 113 for determining that the photographic image does not include a vehicle if the photographic image includes a straight line having an angle within a predetermined range; and an output unit 114 for outputting information which indicates whether or not the photographic image includes a vehicle on the basis of the determination made by the vehicle determination unit 113.SELECTED DRAWING: Figure 2

Description

  The present invention relates to a vehicle detection device and a vehicle detection method for detecting a vehicle in a captured image.

  At a railway station, in order to detect a fallen object on a railway track and to control opening and closing of a platform door, it is determined whether or not there is a railway vehicle stopped on the platform. As a monitoring device that determines the presence or absence of a vehicle, it follows the platform by taking the difference in edge (contour) between the template image when the vehicle is not operating and the current image when the vehicle is operating There is known an apparatus for determining the presence or absence of a vehicle that enters or advances (see, for example, Patent Document 1).

JP 2001-341642 A

  When the brightness of the environment is low, such as at night, the edges in the image are difficult to see. Since the apparatus described in Patent Document 1 makes a determination based on the difference between the template image and the current image, there is a problem in that the presence or absence of the vehicle cannot be accurately determined if the difference is small in a state where the brightness of the environment is low. there were.

  Therefore, the present invention has been made in view of these points, and an object thereof is to provide a vehicle detection device and a vehicle detection method capable of improving the detection accuracy of the presence or absence of a vehicle along a platform of a railway station. To do.

  The vehicle detection device according to the first aspect of the present invention includes a straight line detection unit that detects a straight line included in a captured image captured by an imaging unit that captures an area including a track on which the vehicle travels, and the captured image is stored in advance. A determination unit that determines that the captured image does not include the vehicle when the straight line having an angle within a set range is included, and whether the captured image includes the vehicle based on the determination by the determination unit An output unit that outputs information indicating whether or not.

  The determination unit may determine that the captured image does not include the vehicle when the captured image includes the straight line in the same direction as the traveling direction of the vehicle.

  The straight line detection unit may set a part of the captured image as a detection region and detect the straight line included in the detection region.

  The straight line detection unit detects the straight line included in each of a plurality of divided images obtained by dividing the captured image along the traveling direction of the vehicle, and the determination unit is configured such that the straight line detection unit includes the plurality of divided images. The direction in which the vehicle enters or advances along the track may be specified based on the straight line detected in each of the above.

  The straight line detection unit may detect the straight line by performing Hough transform after binarizing the captured image and extracting a contour.

  In the vehicle detection method according to the second aspect of the present invention, a processor detects a straight line included in a captured image captured by an imaging unit that captures an area including a track on which the vehicle travels, and the captured image includes When the image includes the straight line having an angle within a preset range, the captured image includes the vehicle based on the determination that the captured image does not include the vehicle and the determination in the determination step. Outputting information indicating whether or not.

  According to the present invention, it is possible to improve the detection accuracy of the presence or absence of a vehicle along the platform of a railway station.

1 is a schematic diagram of an image processing system according to a first embodiment. 1 is a block diagram of an image processing system according to a first embodiment. It is a figure which shows the captured image which the imaging device imaged. It is a figure which shows the image of the coordinate range of a detection area. It is a figure which shows the straight line which the straight line detection part which concerns on 1st Embodiment detected in the image. It is a figure which shows the flowchart of the image processing method which concerns on 1st Embodiment. It is a figure which shows the captured image which the imaging device imaged. It is a figure which shows the division | segmentation image which divided | segmented the image along the driving direction of a vehicle. It is a schematic diagram which shows the conditions for which the vehicle determination part which concerns on 2nd Embodiment specifies the direction of approach or advance of a vehicle.

(First embodiment)
FIG. 1 is a schematic diagram of an image processing system S according to the present embodiment. The image processing system S includes an image processing device 1 and an imaging device 2. The image processing system S may include devices such as other servers and terminals.

  The imaging device 2 is a camera installed at a predetermined angle from the upper part of the station platform PL toward the track RA on which the vehicle TR travels. The imaging device 2 includes an imaging unit that performs a bird's-eye image of a predetermined imaging range including the periphery of the platform PL, that is, the platform PL, the vehicle TR, and the track RA. The imaging device 2 transmits the captured image captured using the imaging unit to the image processing device 1. In the present embodiment, the imaging device 2 captures a grayscale captured image, but may capture a full-color captured image. An arbitrary number of imaging devices 2 are arranged so as to cover the entire area to be monitored around the platform PL.

  The image processing apparatus 1 is a computer that performs predetermined processing on a captured image captured by the imaging apparatus 2. The image processing device 1 functions as a vehicle detection device that detects a vehicle based on a captured image, and also functions as an object detection device that detects an object (foreign matter) based on the captured image. The image processing device 1 is connected to the imaging device 2 by wire or wirelessly. The image processing device 1 performs predetermined processing using the captured image received from the imaging device 2.

[Configuration of Image Processing System S]
FIG. 2 is a block diagram of the image processing system S according to the present embodiment. In FIG. 2, arrows indicate main data flows, and there may be data flows other than those shown in FIG. In FIG. 2, each block represents a functional unit configuration, not a hardware (device) unit configuration. Therefore, the blocks shown in FIG. 2 may be implemented in a single device, or may be separately implemented in a plurality of devices. Data exchange between the blocks may be performed via any means such as a data bus, a network, a portable storage medium, or the like.

  The image processing apparatus 1 includes a control unit 11, an interface 12, and a storage unit 13. The control unit 11 includes an image acquisition unit 111, a straight line detection unit 112, a vehicle determination unit 113, an output unit 114, and an object detection unit 115.

  The interface 12 is an interface for exchanging signals with the imaging device 2. The interface 12 performs predetermined processing on the signal received from the imaging device 2 to acquire data, and inputs the acquired data to the control unit 11. In addition, the interface 12 performs a predetermined process on the data input from the control unit 11 to generate a signal, and transmits the generated signal to the imaging device 2.

  The storage unit 13 is a storage medium including a ROM (Read Only Memory), a RAM (Random Access Memory), a hard disk drive, and the like. The storage unit 13 stores a program executed by the control unit 11 in advance. Further, the storage unit 13 stores a captured image captured by the imaging device 2 and area information indicating a determination target area. The storage unit 13 may be provided outside the image processing apparatus 1, and in that case, data may be exchanged with the control unit 11 via the interface 12.

  The control unit 11 is a processor such as a CPU (Central Processing Unit), for example, and by executing a program stored in the storage unit 13, the image acquisition unit 111, the straight line detection unit 112, the vehicle determination unit 113, and the output unit 114. And functions as the object detection unit 115. At least a part of the function of the control unit 11 may be executed by an electric circuit. Moreover, at least a part of the functions of the control unit 11 may be executed by a program executed via a network.

  The image processing system S according to the present embodiment is not limited to the specific configuration shown in FIG. For example, the image processing apparatus 1 is not limited to one apparatus, and may be configured by connecting two or more physically separated apparatuses in a wired or wireless manner.

[Description of vehicle detection method]
The image processing method according to the present embodiment includes a vehicle detection method and an object detection method. In the vehicle detection method, first, the imaging device 2 images a predetermined imaging range. The imaging device 2 may periodically capture images at predetermined time intervals, or may capture images according to instructions from the image processing device 1. The imaging device 2 transmits a captured image captured using the imaging unit to the image processing device 1.

  The image acquisition unit 111 of the image processing apparatus 1 acquires a captured image captured by the imaging apparatus 2 and stores the captured image in the storage unit 13. The image acquisition unit 111 may acquire a captured image every time the imaging device 2 captures an image, or may collectively acquire captured images captured by the imaging device 2 within a predetermined period.

  In the present embodiment, in the imaging device 2, the horizontal direction of the captured image corresponds to the length direction of the platform PL of the station (that is, the traveling direction of the vehicle), and the vertical direction is the width direction of the platform PL or the line RA ( That is, it is arranged so as to correspond to a direction perpendicular to the traveling direction of the vehicle. In the following, in each captured image, the horizontal direction is the X direction (first direction), and the vertical direction is the Y direction (second direction). Further, the direction from the left side to the right side in the horizontal direction toward the track RA is set to the right side, and the direction from the right side to the left side in the horizontal direction is set to the left side.

  The straight line detection unit 112 detects a straight line included in the captured image captured by the imaging device 2. Specifically, first, the straight line detection unit 112 acquires a captured image captured by the imaging device 2 from the storage unit 13. 3A and 3B are diagrams illustrating a captured image P1 captured by the imaging device 2. FIG. In the captured image P <b> 1 in FIG. 3A, the track RA is shown because there is no vehicle TR along the platform PL. In the captured image P1 of FIG. 3B, since the vehicle TR is present along the platform PL, the vehicle TR is shown and the track RA is hidden by the vehicle TR.

  Next, the straight line detection unit 112 sets a part of the captured image P1 that may include the line RA as a detection area AR for detecting a straight line. By setting a part of the captured image P1 as the detection area AR, it is possible to reduce the processing load for determining the presence or absence of a straight line, which will be described later, and to suppress the detection of a straight line unrelated to the line RA.

  For example, the straight line detection unit 112 may accept the designation of the detection area AR by the user from a communication terminal (personal computer, tablet terminal, smartphone, etc.) used by the user. In this case, the straight line detection unit 112 displays the captured image P1 on the display unit of the user's communication terminal. The user designates an area (pixel range) in the captured image P1 using the input unit of the communication terminal. The straight line detection unit 112 sets a region designated by the user as the detection region AR. Thereby, the user can set a desired detection area AR.

  For example, the straight line detection unit 112 may set the detection area AR according to a predetermined rule stored in the image processing apparatus 1 in advance. In this case, the straight line detection unit 112 determines a region in the captured image P1 according to a rule and sets it as the detection region AR. The rule for setting the detection area AR is, for example, pixels in a predetermined range (α-th pixel to β-th pixel from the top) with reference to the upper side of the captured image P1. This eliminates the need for the user to manually specify the detection area AR on the captured image P1, thereby reducing the user's effort.

  Since the imaging device 2 captures an image from the platform PL toward the line RA, one of the two lines RA on the platform PL side is easily hidden by a person or the like on the platform PL. Therefore, the presence or absence of a straight line corresponding to the line RA is determined by setting the detection area AR so as to include one of the lines RA far from the platform PL (that is, the upper side of the captured image P1) of the two lines RA. Can be determined with high accuracy.

  For example, the straight line detection unit 112 may set the detection region AR based on the distance from the imaging device 2 to each pixel of the captured image P1. In this case, the imaging device 2 is configured as a stereo camera having a plurality of imaging units. The imaging device 2 that is a stereo camera simultaneously captures an imaging range using two imaging units, associates two captured images that are simultaneously captured, and transmits the images to the image processing device 1.

  The image acquisition unit 111 of the image processing device 1 uses a known method based on two captured images simultaneously captured by the image capturing device 2, and parallax between the two image capturing units of the image capturing device 2 in the two captured images. Is specified for each pixel of the captured image. The closer the distance from the imaging device 2 to the subject, the larger the parallax, and the farther the distance from the imaging device 2 to the subject, the smaller the parallax. Therefore, the image acquisition unit 111 stores the parallax value specified for each pixel of the captured image in the storage unit 13 in association with the captured image as distance information indicating the distance from the imaging device 2.

  Further, since the imaging device 2 captures the subject from above, the distance from the subject to the imaging device 2 corresponds to the height of the subject from a predetermined reference plane (for example, the surface on which the ground or the track RA is installed). . That is, the higher the subject height, the closer the distance to the imaging device, and the lower the subject height, the farther the distance to the imaging device. Therefore, the image acquisition unit 111 may use the height from a predetermined reference plane as distance information instead of the distance from the imaging device 2.

  Then, the straight line detection unit 112 identifies the distance from the imaging device 2 of each pixel of the captured image P1 based on the distance information stored in the storage unit 13, and the identified distance in the captured image P1 within a predetermined range. An area is determined and set as a detection area AR. The range of the distance is a predetermined range including the distance from the imaging device 2 to the line RA. This eliminates the need for the user to manually specify the detection area AR on the captured image P1, thereby reducing the user's effort.

  Next, the straight line detection unit 112 extracts an image P2 in the coordinate range of the detection area AR from the captured image P1. 4A and 4B are diagrams illustrating an image P2 in the coordinate range of the detection area AR. In the image P2 of FIG. 4A corresponding to the captured image P1 of FIG. 3A, since there is no vehicle TR along the platform PL, a part of the track RA is shown. An image P2 in FIG. 4B corresponding to the captured image P1 in FIG. 3B is in a state where the vehicle TR is present along the platform PL, and thus a part of the vehicle TR is shown.

  The straight line detection unit 112 binarizes the image P2, and extracts an edge (contour) included in the image P2 by a known method such as a gradient method or a Laplacian method. Thereby, in the image P2, the edge portion becomes white and the other portions become black. The straight line detection unit 112 detects a straight line by performing a Hough transform on the image P2 from which the edge is extracted.

  Specifically, the straight line detection unit 112 projects each coordinate (x, y) on the edge in the image P2 into the ρ−θ space by using an equation of ρ = x · cos θ + y · sin θ. The straight line detection unit 112 identifies a point (ρ, θ) where a number of projected curves equal to or greater than a predetermined threshold overlap in the ρ-θ space as a parameter representing a straight line, and an image corresponding to the identified parameter. A straight line on P2 (xy space) is detected.

  The straight line detection unit 112 may detect straight lines corresponding to all points (ρ, θ) in which the number of overlapping curves in the ρ-θ space is equal to or greater than a predetermined threshold. A predetermined number of straight lines may be detected in descending order of the number of overlapping.

  The straight line detection unit 112 is not limited to the specific method described above, and a known method capable of detecting a straight line included in the captured image can be used. The straight line detection unit 112 may detect a straight line by extracting an edge from an image in which the image P2 is binarized instead of binarized, or may detect a straight line by extracting an edge from a grayscale image. By binarizing the image P2, it is possible to reduce noise in the image P2, and to detect a straight line while reducing the calculation load.

  Next, the vehicle determination unit 113 specifies whether each straight line detected by the straight line detection unit 112 is a straight line in the same direction as the traveling direction of the vehicle TR (that is, the extending direction of the track RA). Specifically, the vehicle determination unit 113 has a preset range including the traveling direction of the vehicle TR (the horizontal direction of the captured image P1) (for example, a range of −10 degrees to +10 degrees with respect to the traveling direction of the vehicle TR). A straight line having an inner angle is specified as a straight line in the same direction as the traveling direction of the vehicle TR.

  5A and 5B are diagrams illustrating the straight lines L1 and L2 detected by the straight line detection unit 112 in the image P2. An image P2 in FIG. 5A corresponding to the captured image P1 in FIG. 3A includes a straight line L1 that the vehicle determination unit 113 specifies as the same direction as the traveling direction of the vehicle TR. That is, the image P2 in FIG. 5A shows a state in which the vehicle TR is not present along the platform PL. Therefore, the straight line L1 corresponding to the line RA extending along the same direction as the traveling direction of the vehicle TR by the Hough transformation is illustrated. Including.

  The image P2 in FIG. 5B corresponding to the captured image P1 in FIG. 3B does not include the straight line L1 that the vehicle determination unit 113 specifies as the same direction as the traveling direction of the vehicle TR, and the vehicle determination unit 113 includes a straight line L2 that is identified as not being in the same direction as the traveling direction of the vehicle TR. That is, the image P2 in FIG. 5B does not include the straight line L1 corresponding to the track RA along the traveling direction of the vehicle TR due to the Hough transformation because the vehicle TR is present along the platform PL. Instead, the image P2 in FIG. 5B includes a straight line L2 corresponding to the structure of the vehicle TR (for example, a connecting portion of the vehicle TR) extending in a direction different from the traveling direction of the vehicle TR.

  When the track RA is shown in the image P2, there is no vehicle TR along the platform PL, and when the track RA is not shown in the image P2, there is a vehicle TR along the platform PL. Therefore, the vehicle determination unit 113 determines that at least one of the straight lines detected by the straight line detection unit 112 in the image P2 is a straight line in the same direction as the traveling direction of the vehicle TR (that is, the detection area AR is the traveling direction of the vehicle TR). The captured image P1 (image P2) is determined not to include the vehicle TR.

  On the other hand, when all the straight lines detected by the straight line detection unit 112 in the image P2 are not in the same direction as the traveling direction of the vehicle TR (that is, the detection area AR is the same as the traveling direction of the vehicle TR). It is determined that the captured image P1 (image P2) includes the vehicle TR when the straight line of the direction is not included.

  Based on the determination by the vehicle determination unit 113, the output unit 114 outputs information indicating whether or not the captured image P1 includes the vehicle TR. The output unit 114 may store information in the storage unit 13 or other storage device. The output unit 114 may display information using a display device such as a liquid crystal display or a lamp. The output unit 114 may output sound indicating information using a speaker. The output unit 114 may print information using a printer.

  As described above, the image processing apparatus 1 binarizes and extracts an edge of the image P2 of the detection area AR that may include the line RA in the captured image P1, and detects a straight line corresponding to the line RA by Hough transform. Thus, the presence / absence of the vehicle TR along the platform PL is determined. Since the extraction of the edge from the captured image P1 is strong against changes in the amount of light, the image processing apparatus 1 can improve the detection accuracy of the presence / absence of the vehicle even when the brightness of the environment is low.

[Description of object detection method]
In the vehicle detection method, when the vehicle determination unit 113 determines that the captured image P1 (image P2) does not include the vehicle TR, the object detection unit 115 detects an object in the monitoring region of the captured image P1. The object detection unit 115 may set the entire captured image P1 (that is, the imaging range of the imaging device 2) as a monitoring region, or may set a part of the captured image P1 as a monitoring region.

  Specifically, the object detection unit 115 uses a captured image at any one of the past (first time) stored in the storage unit 13 as a reference image, and the monitoring area of the reference image and the latest (second time). The monitored area of the captured image is compared. The object detection unit 115 may use a captured image stored immediately before in time series as a reference image, or may use a captured image at a specific time (for example, when the imaging device 2 is installed) as a reference image. When the difference between the monitoring area of the reference image and the monitoring area of the latest captured image satisfies a predetermined condition (for example, the area of the difference is a predetermined value or more), the object detection unit 115 is in the monitoring area. A newly existing object is detected.

  The method by which the object detection unit 115 detects an object in the monitoring area is not limited to the specific method described above. For example, the object detection unit 115 may detect an object when a pixel having a distance within a predetermined range is generated in the monitoring region of the captured image P1 based on the above-described distance information. By appropriately setting the distance to be monitored by the user, the object detection unit 115 can be prevented from detecting a shadow formed on the ground as an object.

  When the object detection unit 115 detects an object existing in the monitoring area, the output unit 114 outputs information notifying that the object detection unit 115 has detected an object in the monitoring area of the captured image P1.

  On the other hand, in the vehicle detection method, when the vehicle determination unit 113 determines that the captured image P1 (image P2) includes the vehicle TR, the object detection unit 115 does not detect an object in the monitoring region of the captured image P1. . As described above, the object detection unit 115 detects the object in the captured image P1 only when the vehicle determination unit 113 determines that there is no vehicle TR along the platform PL. Therefore, the object detection unit 115 is on the platform PL. The detection of the vehicle TR as an object (foreign matter) can be suppressed, and the output of useless notification can be suppressed.

  In the present embodiment, the image processing apparatus 1 detects the object itself after determining the presence or absence of the vehicle TR in the captured image P1, but the external apparatus may detect the object. In this case, the image processing device 1 transmits information indicating the determination result of the presence or absence of the vehicle TR in the captured image P1 to the external device, and the external device detects an object based on the determination result received from the image processing device 1. Decide whether or not to do.

[Flowchart of image processing method]
FIG. 6 is a diagram illustrating a flowchart of the image processing method according to the present embodiment. The flowchart in FIG. 6 is started, for example, when the user performs a predetermined start operation on the image processing apparatus 1.

  The image acquisition unit 111 acquires the captured image P1 captured by the imaging device 2 and stores it in the storage unit 13 (S11). The straight line detection unit 112 sets a part of the captured image P1 acquired by the image acquisition unit 111 in step S11 as a detection area AR for detecting a straight line (S12).

  The straight line detection unit 112 extracts an image P2 in the coordinate range of the detection area AR set in step S12 from the captured image P1. Then, the straight line detection unit 112 binarizes and extracts an edge from the image P2, and performs a Hough transform to detect a straight line (S13). The vehicle determination unit 113 identifies whether each straight line detected by the straight line detection unit 112 in step S13 is a straight line in the same direction as the traveling direction of the vehicle TR (that is, the extending direction of the track RA) (S14).

  The vehicle determination unit 113 determines that the captured image P1 (image P2) does not include the vehicle TR when the image P2 (detection area AR) includes a straight line in the same direction as the traveling direction of the vehicle TR (YES in S15). (S16). The output unit 114 outputs information indicating that the captured image P1 does not include the vehicle TR. Then, the object detection unit 115 detects an object in the monitoring area of the captured image P1 acquired by the image acquisition unit 111 in step S11 (S17).

  When the object detection unit 115 detects an object existing in the monitoring region in step S17 (YES in S18), the output unit 114 indicates that the object detection unit 115 has detected an object in the monitoring region of the captured image P1. Information to be notified is output (S19). If the object detection unit 115 does not detect an object present in the monitoring area in step S17 (NO in S18), the image processing apparatus 1 proceeds to step S21.

  The vehicle determination unit 113 determines that the captured image P1 (image P2) includes the vehicle TR when the image P2 (detection area AR) does not include a straight line in the same direction as the traveling direction of the vehicle TR (NO in S15). (S20). The output unit 114 outputs information indicating that the captured image P1 includes the vehicle TR. In this case, the object detection unit 115 does not detect the object.

  When a predetermined end condition (for example, the user performing a predetermined end operation on the image processing apparatus 1) is not satisfied (NO in S21), the image processing apparatus 1 returns to step S11 and repeats the process. When the predetermined end condition is satisfied (YES in S21), the image processing apparatus 1 ends the process.

[Effect of the first embodiment]
The image processing apparatus 1 according to the present embodiment detects the straight line after binarizing and extracting the image P2 of the detection area AR that may include the line RA in the captured image P1, thereby detecting the straight line on the platform PL. The presence or absence of a vehicle TR along the road is determined. Since the extraction of the edge from the captured image P1 is strong against changes in the amount of light, the image processing apparatus 1 can improve the detection accuracy of the presence / absence of the vehicle TR even when the brightness of the environment is low.

  Furthermore, the object detection unit 115 detects an object in the captured image P1 only when the vehicle determination unit 113 determines that there is no vehicle TR along the platform PL. Thereby, the object detection part 115 can suppress detecting the vehicle TR in the platform PL as an object (foreign substance), and can suppress the output of useless notification.

(Second Embodiment)
The first embodiment determines the presence or absence of the vehicle TR along the platform PL, while the present embodiment determines the direction in which the vehicle TR enters or enters along the platform PL in addition to the presence or absence of the vehicle TR. judge.

  FIG. 7 is a diagram illustrating a captured image P1 captured by the imaging device 2. In the captured image P1 of FIG. 7, since the vehicle TR enters from the right side along the platform PL, the track RA is shown in a part of the left side of the captured image P1, and the right side of the captured image P1 is shown. The vehicle TR is shown in the part.

  The straight line detection unit 112 according to the present embodiment extracts the image P2 in the coordinate range of the detection area AR from the captured image P1 acquired by the image acquisition unit 111, and further extracts the image P2 at the division position D in the traveling direction of the vehicle TR. Are divided into a plurality of divided images. In the example of FIG. 7, the straight line detection unit 112 divides the image P2 into four at three division positions D. However, the image P2 may be divided into two at at least one division position D. The straight line detection unit 112 performs binarization and edge extraction on each divided image, and detects a straight line by performing Hough transform.

  Fig.8 (a)-FIG.8 (c) are the figures which show the division | segmentation image P3 which divided | segmented the image P2 along the running direction of vehicle TR. The divided image P3 of FIG. 8A corresponds to the state of FIG. 3A where the vehicle TR is not present along the platform PL, and the divided image P3 of FIG. 8B is the vehicle TR viewed from the right side along the platform PL. Corresponding to the state of FIG. 7 entering, the divided image P3 of FIG. 8C corresponds to the state of FIG. 3B where the vehicle TR is along the platform PL. In each divided image P3, a straight line detected by the straight line detection unit 112 is represented.

  Then, the vehicle determination unit 113 specifies whether each straight line detected by the straight line detection unit 112 is a straight line in the same direction as the traveling direction of the vehicle TR, and a divided image including a straight line in the same direction as the traveling direction of the vehicle TR. It is determined that P3 includes the vehicle TR. 8A to 8C, when the vehicle determination unit 113 determines that the vehicle TR is included under each divided image P3 (not including the straight line L1 in the same direction as the traveling direction of the vehicle TR). ), “X” is shown when the vehicle determination unit 113 does not determine that the vehicle TR is included (when the straight line L1 in the same direction as the traveling direction of the vehicle TR is included).

  In the first embodiment, as shown in FIG. 8B, when the vehicle TR is part of the range of the captured image P1, the straight line detection unit 112 detects the straight line L1 corresponding to the track RA. The unit 113 determines that the captured image P1 does not include the vehicle TR. On the other hand, when at least one of the divided images P3 does not include the straight line L1 in the same direction as the traveling direction of the vehicle TR (that is, when the track RA is hidden in the vehicle TR), the vehicle determination unit 113 according to the present embodiment. ), It is determined that the captured image P1 includes the vehicle TR. Therefore, the image processing apparatus 1 has the vehicle TR even when the vehicle TR enters or advances along the platform PL or when the vehicle TR stops in the middle of the range of the captured image P1. Can be determined.

  Further, the vehicle determination unit 113 determines the presence or absence of the vehicle TR in each divided image P3 at a plurality of timings (here, the first time point and the second time point), so that the vehicle TR enters or advances along the platform PL. Is identified. The determination timing interval is set in advance according to the speed of the vehicle TR.

  FIG. 9A to FIG. 9D are schematic diagrams illustrating conditions for the vehicle determination unit 113 according to the present embodiment to specify the direction of entry or advancement of the vehicle TR. 9A to 9D show the determination results of the vehicle determination unit 113 for the four divided images P3 at the first time point and the second time point after the first time point (when the vehicle TR is included, “ “O”, “x” when the vehicle TR is not included, is displayed in the order of the divided images P3.

  In the case of FIG. 9A, the region including the vehicle TR on the right side increases from the first time point to the second time point. In such a case, the vehicle determination unit 113 determines that the vehicle TR has entered the platform PL leftward. In the case of FIG. 9B, the region including the vehicle TR on the left side increases from the first time point to the second time point. In such a case, the vehicle determination unit 113 determines that the vehicle TR has entered the platform PL in the right direction.

  In the case of FIG. 9C, the region including the left vehicle TR decreases from the first time point to the second time point. In such a case, the vehicle determination unit 113 determines that the vehicle TR has advanced leftward from the platform PL. In the case of FIG. 9D, the region including the right vehicle TR decreases from the first time point to the second time point. In such a case, the vehicle determination unit 113 determines that the vehicle TR has advanced to the right from the platform PL.

  As described above, the vehicle determination unit 113 can specify the direction in which the vehicle TR enters or advances along the platform PL based on the change over the plurality of timings of the number of the divided images P3 including the vehicle TR.

[Effects of Second Embodiment]
The image processing apparatus 1 according to the present embodiment determines the presence or absence of the vehicle TR for a plurality of divided images P3 obtained by dividing the captured image P1, so that the vehicle TR enters or advances along the platform PL, or Even if the vehicle TR is stopped in the middle of the range of the captured image P1, it can be determined that the vehicle TR is present along the platform PL.

  Furthermore, the image processing apparatus 1 can determine the direction in which the vehicle TR enters or advances along the platform PL by determining the presence or absence of the vehicle TR at a plurality of timings for the plurality of divided images P3. For example, the image processing apparatus 1 compares the direction in which the vehicle TR enters or advances with the direction of the incoming line previously associated with the platform PL, and can confirm that the determination that the vehicle TR is present is not a false detection if they match. effective. When the direction in which the vehicle TR enters or advances and the direction of the incoming line pre-associated with the platform PL do not match, the image processing apparatus 1 indicates that the determination that the vehicle TR is present may be erroneously detected. It may be output.

  As mentioned above, although this invention was demonstrated using embodiment, the technical scope of this invention is not limited to the range as described in the said embodiment, A various deformation | transformation and change are possible within the range of the summary. is there. For example, the specific embodiments of device distribution / integration are not limited to the above-described embodiments, and all or a part of them may be configured to be functionally or physically distributed / integrated in arbitrary units. Can do. In addition, new embodiments generated by any combination of a plurality of embodiments are also included in the embodiments of the present invention. The effect of the new embodiment produced by the combination has the effect of the original embodiment.

  The control unit 11 (processor) of the image processing apparatus 1 is a main component of each step (process) included in the method shown in FIG. That is, the control unit 11 reads out a program for executing the method shown in FIG. 6 from the storage unit 13, and executes the method shown in FIG. 6 by executing the program and controlling each unit of the image processing apparatus 1. To do.

  The image processing device 1 may be configured integrally with the imaging device 2. That is, the imaging device 2 may have at least a part of the functions of the image processing device 1 (that is, the image acquisition unit 111, the straight line detection unit 112, the vehicle determination unit 113, the output unit 114, and the object detection unit 115).

S Image processing system 1 Image processing device 2 Imaging device 11 Control unit 112 Straight line detection unit 113 Vehicle determination unit 114 Output unit

Claims (6)

A straight line detection unit that detects a straight line included in a captured image captured by an imaging unit that captures an area including a track on which the vehicle travels;
A determination unit that determines that the captured image does not include the vehicle when the captured image includes the straight line having an angle within a preset range;
An output unit that outputs information indicating whether the captured image includes the vehicle based on the determination by the determination unit;
A vehicle detection device.   The vehicle detection device according to claim 1, wherein the determination unit determines that the captured image does not include the vehicle when the captured image includes the straight line in the same direction as the traveling direction of the vehicle.   The vehicle detection device according to claim 1, wherein the straight line detection unit sets a part of the captured image as a detection region and detects the straight line included in the detection region. The straight line detection unit detects the straight line included in each of a plurality of divided images obtained by dividing the captured image along the traveling direction of the vehicle,
4. The method according to claim 1, wherein the determination unit specifies a direction in which the vehicle enters or advances along the track based on the straight lines detected by the straight line detection unit in each of the plurality of divided images. The vehicle detection device according to claim 1.   The vehicle detection device according to any one of claims 1 to 4, wherein the straight line detection unit detects the straight line by performing Hough transform after binarizing the captured image and extracting a contour. Processor
Detecting a straight line included in a captured image captured by an imaging unit that captures an area including a track on which the vehicle travels;
Determining that the captured image does not include the vehicle when the captured image includes the straight line having an angle within a preset range;
Outputting information indicating whether or not the captured image includes the vehicle based on the determination in the determination step;
Vehicle detection method for executing

Images