JP2019098846A - Error calculation device and error calculation system - Google Patents

Abstract

To provide an error calculation device capable of highly accurately grasping an error between an actual stop position when a train stops at a station and a predefined target stop position.SOLUTION: An error calculation device 10 is provided with an acquisition unit 102, an identification unit 103 and a calculation unit 104. The acquisition unit acquires a captured image from an imaging device installed in a train. The identification unit identifies a marker set at a station where the train stops from the captured image. The calculation unit acquires a difference between the position of a prescribed part in the identified marker in the captured image and a reference position which is the position of a prescribed part in the captured image captured in advance when the train stops at a predefined target stop position, and calculates an error between the stop position and the target stop position of the train on the basis of the difference.SELECTED DRAWING: Figure 2

Description

  Embodiments of the present invention relate to an error calculation device and an error calculation system.

  Conventionally, there are technologies that control trains to operate automatically, such as ATO (Automatic Train Operation), and technologies that help a train stop at a defined target stop position, such as Train Automatic Stop-position Controller (TASC). Are known.

Patent No. 6045483 Patent No. 5634951 gazette Patent No. 6081549 Patent No. 6091294 Patent No. 5484280

  However, in the prior art, it has been difficult to accurately grasp an error between an actual stop position when a train stops at a station and a predetermined target stop position.

  The error calculation device of the embodiment includes an acquisition unit, a specification unit, and a calculation unit. The acquisition unit acquires a captured image from an imaging device installed in a train. The identification unit identifies, from the captured image, a marker installed at a station at which the train stops. The calculation unit includes a position in a captured image of a predetermined portion in the specified marker, and a reference position which is a position of the predetermined portion in the captured image captured in advance when the train is stopped at a predetermined target stop position. And calculate an error between the stop position of the train and the target stop position based on the difference.

FIG. 1 is a diagram showing an example of the entire configuration of the error calculation system according to the first embodiment. FIG. 2 is a block diagram showing an example of the function of the error calculation device according to the first embodiment. FIG. 3 is a view showing an example of a captured image obtained by capturing the marker according to the first embodiment. FIG. 4 is a flowchart showing an example of the flow of the error calculation process according to the first embodiment. FIG. 5 is a flowchart showing an example of the flow of the marker identification process according to the first embodiment. FIG. 6 is a diagram showing an example of a method of identifying a marker according to the second embodiment. FIG. 7 is a flowchart showing an example of the flow of the marker identification process according to the second embodiment. FIG. 8 is a view showing an example of the installation position of the marker according to the first modification. FIG. 9 is a view showing an example of the installation position of the marker according to the second modification. FIG. 10 is a view showing an example of the installation position of the marker according to the third modification. FIG. 11 is a view showing an example of the installation position of the marker according to the fourth modification. FIG. 12 is a diagram showing an example of the entire configuration of the error calculation system according to the third embodiment. FIG. 13 is a side view showing an example of the marker according to the third embodiment. FIG. 14 is a view showing an example of a captured image obtained by capturing a marker according to the third embodiment. FIG. 15 is a flowchart showing an example of the flow of the error calculation process according to the third embodiment.

First Embodiment
Generally, at a station where a train stops, a target stop position of the train is predetermined. However, in practice, the train stops before the target stop position or stops after passing the target stop position, so that the position at which the train stopped at the stop station (stop position) and the target stop position Errors may occur. In the error calculation system of the present embodiment, the error between the train stop position and the target stop position is calculated with high accuracy.

  FIG. 1 is a view showing an example of the entire configuration of an error calculation system S according to the present embodiment. The error calculation system S of the present embodiment includes the marker 2, the error calculation device 10, the cameras 3 a and 3 b, and the control device 11.

  The marker 2 is a mark used to calculate an error between a predetermined target stop position and an actual stop position when the train 1 stops at the station. As shown in FIG. 1, the marker 2 of the present embodiment is a plurality of circles arranged so that a plurality of columns and a plurality of rows intersect. The center in the width direction of the marker 2 is an example of a predetermined portion in the marker 2 in the present embodiment. The predetermined site is not limited to this.

  Moreover, in this embodiment, the marker 2 is installed on the floor of the platform 4 of the station so as to be parallel to the front-rear direction of the train 1. The marker 2 may be described on the floor of the home 4 with paint or the like, or a plate, a sheet or the like on which the marker 2 is displayed may be affixed to the floor of the home 4. The surface of the marker 2 may be processed to suppress the reflection of light by a known technique.

  The cameras 3 a and 3 b are imaging devices installed in the train 1. Hereinafter, when the camera 3a and the camera 3b are not particularly distinguished from one another, it is referred to as the camera 3. More specifically, the camera 3 is installed on the train 1 at an imaging angle at which the marker 2 installed on the station home 4 can be imaged. In the present embodiment, the camera 3 is installed on the upper part of the train 1's vehicle at an imaging angle that faces the home 4 and obliquely downward. Moreover, the camera 3 may be protruded from the vehicle of the train 1 and installed at a downward imaging angle. The camera 3 is, for example, a CCTV (Closed-Circuit Television) camera, but is not limited thereto. Further, the camera 3 may capture a moving image or may capture a still image.

  Also, in the example of FIG. 1, the cameras 3a and 3b are respectively installed on the left and right of the train 1, but it is possible to pick up the marker 2 even if the home 4 is located in either direction of the train 1 In order to When the positional relationship between the home 4 and the train 1 is the same at all the stations, any one of the cameras 3a and 3b may be used.

  The control device 11 is a device that controls the operation of the train 1 and is mounted on the train 1. The control device 11 is, for example, an automatic train operating device such as ATO. Further, the control device 11 has various types of information related to the operation of the train 1. The control device 11 and the error calculation device 10 are communicably connected to each other via a network or the like. The control device 11 transmits, for example, information indicating the next stop station of the train 1 to the error calculation device 10. Further, when the train 1 stops at the station, the control device 11 transmits to the error calculation device 10 a stop signal indicating that the train 1 has stopped.

  Further, the control device 11 stores the error between the stop position of the train 1 and the target stop position transmitted from the error calculation device 10 described later. The control device 11 adjusts the value of the control parameter of the automatic operation control so that the error between the stop position of the train 1 and the target stop position becomes smaller by performing analysis using a method such as machine learning.

  The error calculation device 10 is mounted on the train 1 and calculates an error between the stop position of the train 1 and the target stop position based on the position of the image of the marker 2 in the captured image captured by the camera 3. Details of the function of the error calculation device 10 will be described later.

  The control device 11 and the error calculation device 10 according to the present embodiment include a control device such as a central processing unit (CPU), a storage device such as a read only memory (ROM) or a random access memory (RAM), and a hard disk drive (HDD). Etc., and a communication interface etc. which can be connected to a network etc., and has a hardware configuration using a normal computer. Specifically, the control device 11 and the error calculation device 10 are a PC (Personal Computer) or a server.

  Next, the function of the error calculation device 10 of the present embodiment will be described in detail. FIG. 2 is a block diagram showing an example of the function of the error calculation device 10 according to the present embodiment. As shown in FIG. 2, the error calculation device 10 of the present embodiment includes a reception unit 101, an acquisition unit 102, a specification unit 103, a calculation unit 104, a transmission unit 105, and a storage unit 150.

  The storage unit 150 of the present embodiment stores home position information, a marker length, the X coordinate of the reference position, and the maximum value of the computable errors. The storage unit 150 is, for example, a storage device such as an HDD.

  The home position information is information in which the stop station of the train 1 is associated with which direction the home 4 at the stop station is located to the left or right of the train 1.

  The marker length is the length in the width direction (horizontal direction, left-right direction) of the marker 2 in the real space. More specifically, the marker length in the present embodiment is the distance between the center of the leftmost circle and the center of the rightmost circle among the plurality of circles of the marker 2.

  The X coordinate of the reference position is a coordinate of the center (predetermined portion) in the width direction of the marker 2 in the captured image captured in advance when the train 1 stops at the target stop position. In the present embodiment, the positional relationship between the marker 2 and the camera 3 is set in advance so that the X coordinate of the reference position is at the center in the width direction of the captured image. The X coordinate of the reference position is not limited to this.

  Moreover, the maximum value of the errors that can be calculated is the maximum value of the errors between the stop position of the train 1 and the target stop position that can be calculated from the captured image by the error calculation device 10 of the present embodiment. In the present embodiment, when the entire marker 2 is not included in the imaging range of the camera 3, the error between the stop position of the train 1 and the target stop position can not be determined from the captured image. For this reason, the upper limit of the error in which the whole of the marker 2 is included in the imaging range of the camera 3 is set as the maximum value of the computable errors. In this embodiment, the maximum value of errors that can be calculated is “X cm”.

  The receiving unit 101 acquires information on the operation of the train 1 from the control device 11. More specifically, the receiving unit 101 receives, from the control device 11, information indicating the next stop station of the train 1. The receiving unit 101 also receives a stop signal indicating that the train 1 has stopped from the control device 11.

  The acquisition unit 102 acquires a captured image from the camera 3. More specifically, based on the information indicating the next stopping station received by the receiving unit 101 from the control device 11, the acquiring unit 102 determines the direction of the home 4 of the next stopping station from the home position information of the storage unit 150. Search for. Then, the acquisition unit 102 acquires a captured image of the camera 3 on the home 4 side of the next stop station. The acquiring unit 102 repeatedly acquires, for example, a captured image from the camera 3 at predetermined time intervals.

  The identifying unit 103 identifies the marker 2 from the captured image acquired by the acquiring unit 102 when the receiving unit 101 receives the stop signal.

  More specifically, the specifying unit 103 extracts a plurality of circles forming the marker 2 from the captured image 90. As a method of extracting a plurality of circles, a known image processing method such as Hough transform may be adopted. FIG. 3 is a view showing an example of a captured image 90 obtained by capturing the marker 2 according to the present embodiment. Twelve circles 21a to 21l constituting the marker 2 of the present embodiment are arranged in a grid of four columns vertically and three rows horizontally. Further, the position in the vertical direction (vertical direction) of the captured image 90 is indicated by Y coordinates, and the position in the horizontal direction (width direction, horizontal direction) is indicated by X coordinates. Hereinafter, when the circles 21a to 21l are not particularly distinguished, they are referred to as a circle 21.

  In addition, when the identifying unit 103 extracts a plurality of circles 21, circular stains, patterns, and the like that are unrelated to the marker 2 may be extracted. For example, in FIG. 3, the circle P is a circular pattern unrelated to the marker 2. The identifying unit 103 groups the extracted plurality of circles 21 into columns and rows, and excludes a circle P that is not included in any column or row. More specifically, the identifying unit 103 determines that the circles 21 whose difference in X coordinates in the captured image 90 is less than or equal to the first threshold are included in the same column, and the difference in Y coordinates is less than or equal to the second threshold. It is determined that the circle 21 is included in the same line. The identifying unit 103 excludes the circle P, and if the number of the circles 21 included in each column and each row matches the number “12” of the circles 21 of the marker 2, the circles included in each column and each row Identify 21 as marker 2. The identifying unit 103 sends the X coordinate of the circle 21 identified as the marker 2 to the calculating unit 104.

  The first threshold and the second threshold are predetermined. In the present embodiment, the first threshold and the second threshold are the same value, but may be different values. Although the X coordinate of the circle 21 included in the same column and / or the Y coordinate of the circle 21 included in the same row have substantially the same value, differences may occur depending on imaging conditions and the like. In order to allow such difference, the identifying unit 103 uses a first threshold and a second threshold. Further, the number of circles 21 of the marker 2 is an example of a predetermined predetermined number in the present embodiment.

  Further, after excluding the circle P, when the number of extracted circles 21 does not match the number of circles 21 of the marker 2, the identifying unit 103 determines that the captured image 90 does not include the marker 2.

  Returning to FIG. 2, the calculating unit 104 calculates the difference between the center in the width direction of the marker 2 identified from the captured image 90 from the identifying unit 103 (the position in the captured image 90 of a predetermined part in the marker 2) and the reference position. Do. Specifically, the calculation unit 104 subtracts the X coordinate of the reference position stored in the storage unit 150 from the X coordinate of the center in the width direction of the marker 2 to obtain a difference. The difference is calculated, for example, by the number of pixels in the captured image 90. Then, the calculation unit 104 calculates an error between the stop position of the train 1 and the target stop position based on the difference.

  In addition, the calculation unit 104 divides the marker length stored in the storage unit 150 by the number of pixels in the width direction of the marker 2 in the captured image 90 to obtain the length in real space per pixel of the captured image 90. Calculate the Hereinafter, the length in real space per pixel of the captured image 90 is referred to as a conversion parameter.

  The calculation unit 104 calculates an error between the stop position of the train 1 and the target stop position by converting the calculated difference into a length in the real space. More specifically, the calculation unit 104 converts the calculated difference into a length in the real space by multiplying the calculated number of pixels of the difference by the conversion parameter. In addition, the method of converting the difference on the captured image 90 is not limited to this.

  The transmission unit 105 transmits the error between the stop position of the train 1 calculated by the calculation unit 104 and the target stop position to the control device 11. In addition, when the identifying unit 103 determines that the marker 2 is not included in the captured image 90, the transmitting unit 105 determines that the error between the stop position of the train 1 and the target stop position can be calculated as the maximum value “X cm” of the error. The larger value is transmitted to the control device 11 as the calculation result of the error.

  Next, an error calculation process performed by the error calculation device 10 configured as described above will be described. FIG. 4 is a flowchart showing an example of the flow of the error calculation process according to the present embodiment.

  The receiving unit 101 receives, from the control device 11, information indicating the next stop station of the train 1 (S1). Then, the acquisition unit 102 searches for the direction of the home 4 at the next stop station from the home position information of the storage unit 150 (S2).

  The acquisition unit 102 acquires a captured image 90 from the camera 3 on the home 4 side of the next stop station among the cameras 3 installed in the train 1 (S3).

  The receiving unit 101 determines whether a stop signal has been received from the control device 11 (S4). When the receiving unit 101 receives the stop signal, the receiving unit 101 notifies the identifying unit 103 that the stop signal has been received. When the receiving unit 101 receives the stop signal (S4 “Yes”), the identifying unit 103 starts the marker identifying process (S5). Details of the marker identification process of S5 will be described later with reference to FIG.

  When the reception unit 101 has not received the stop signal (S4 “No”), the processes of S3 and S4 are repeated.

  When the identifying unit 103 identifies the marker 2 from the captured image 90 in the marker identifying process of S5 (S6 “Yes”), the calculating unit 104 causes the identifying unit 103 to circle 21a to 21l in the captured image 90 illustrated in FIG. Get X coordinate of. The calculation unit 104 calculates the average value “x1” of the X coordinates of the circles 21a, 21e and 21i in the first row from the left of the marker 2 as the minimum X coordinates of the marker 2. In addition, the calculation unit 104 calculates the average value “x2” of the X coordinates of the circles 21 d, 21 h, 21 l in the fourth row from the left of the marker 2 as the maximum X coordinates of the marker 2. The calculation unit 104 calculates the X coordinate of the center in the width direction of the marker 2 by calculating the X coordinate of the middle point between the minimum X coordinate “x1” of the marker 2 and the maximum X coordinate and “x2”. .

  Further, in the example shown in FIG. 3, the X coordinate of the reference position is "x4". The calculation unit 104 calculates the difference D1 between the center in the width direction of the marker 2 in the captured image 90 and the reference position by subtracting the X coordinate "x3" from the X coordinate "x4" (S7). When the train 1 stops at the target stop position, the difference D1 is "0". Further, in the captured image 90 shown in FIG. 3, when the train 1 stops after passing the target stop position, the difference D1 becomes a positive number, and when the train 1 stops before the target stop position, the difference D1 Is a negative number.

  Then, the calculation unit 104 subtracts the minimum X coordinate of the marker 2 from the maximum X coordinate of the marker 2 in the captured image 90 to obtain the length in the width direction of the marker 2 in the captured image 90. The calculation unit 104 calculates a conversion parameter from the marker length stored in the storage unit 150 and the length in the width direction of the marker 2 in the captured image 90. The calculation unit 104 multiplies the calculated difference D1 by the conversion parameter to convert it into a length in the real space, and calculates an error between the stop position of the train 1 and the target stop position (S8).

  The transmission unit 105 transmits the error between the stop position of the train 1 calculated by the calculation unit 104 and the target stop position to the control device 11 (S9).

  Moreover, when the specification part 103 does not specify the marker 2 from the captured image 90 in the marker specification process of S5 (S6 "No"), the transmission part 105 is the difference between the stop position of the train 1 and the target stop position. It is transmitted to the control device 11 that it is larger than the maximum value “X cm” of the computable error (S10).

  Next, the details of the marker identification process of S5 will be described. FIG. 5 is a flow chart showing an example of the flow of the marker identification process according to the present embodiment.

  The identifying unit 103 extracts the circle 21 from the captured image 90 by a known image processing method (S101). At this time, a circle P other than the circle 21 may be extracted.

  The identifying unit 103 groups the extracted circle 21 and circle P into columns (S102). At the time of grouping, the identifying unit 103 determines that the circles 21 whose difference in X coordinates is equal to or less than the first threshold are included in the same column. In the example of the marker 2 shown in FIG. 3, the identifying unit 103 determines that the circles 21 a, 21 e, 21 i are in the first row, the circles 21 b, 21 f, 21 j are in the second row, the circles 21 c, 21 g, 21 k are in the third row, the circle 21 d , 21h and 21l are determined to be included in the fourth column.

  Then, the identifying unit 103 determines the presence or absence of a circle P which is not included in any row (S103). The X coordinate of the circle P is not included in any of the four columns because the difference between the X coordinates of each other does not fall below the first threshold when compared with the X coordinate of the circle 21 included in any column .

  When there is a circle P which is not included in any column among the extracted circles 21 and circles P (S103 “Yes”), the identifying unit 103 determines from the extracted circle 21 and circle P in any column. The circle P not included is excluded (S104), and the process proceeds to the process of S105.

  In addition, when there is no circle P which is not included in any column among the extracted circle 21 and circle P (S103 “No”), the specifying unit 103 does not perform the exclusion process of S104, and the process of S105 is performed. Proceed to processing.

  Next, the identifying unit 103 groups the extracted circles 21 into rows (S105). At the time of grouping, the identifying unit 103 determines that the circles 21 whose difference in Y coordinate is equal to or less than the second threshold are included in the same row. In the example of the marker 2 shown in FIG. 3, the identifying unit 103 determines that the circles 21 a to 21 d are included in the first line, the circles 21 e to 21 h are included in the second line, and the circles 21 i to 21 l are included in the third line.

  Then, the identifying unit 103 determines the presence or absence of the circle 21 not included in any row (S106). When there is a circle 21 not included in any row among the extracted circles 21 (S106 “Yes”), the specifying unit 103 excludes the circle 21 not included in any row (S107) , And proceeds to the process of S108. As described above, since the specifying unit 103 performs the process of excluding the circle P using both the column and the row, the circle P can be excluded with higher accuracy.

  When there is no circle 21 not included in any row among the extracted circles 21 (S106 “No”), the identifying unit 103 proceeds to the process of S108 without performing the exclusion process of S107.

  Next, the specifying unit 103 determines whether the number of the circles 21 remaining after the circle P is excluded among the circles 21 and the circles P extracted from the captured image 90 is “12” (S108). . When the number of remaining circles 21 is “12” (S108 “Yes”), the identifying unit 103 identifies the remaining circles 21 as a marker 2 (S109).

  If the number of circles 21 remaining after excluding the circle P is not “12” (S108 “No”), the identifying unit 103 determines that the marker 2 is not included in the captured image 90 (S110). Here, the processing of this flowchart ends, and the processing returns to the processing of the flowchart of FIG. 4.

  As described above, the error calculation device 10 of the present embodiment identifies the marker 2 from the captured image 90, and the position in the captured image 90 of the predetermined part in the identified marker 2, the reference position on the captured image 90, The difference between the stop position of the train 1 and the target stop position is calculated on the basis of the difference D1. For this reason, according to the error calculation device 10 of the present embodiment, it is possible to grasp with high accuracy the error between the actual stop position when the train 1 stops at the station and the predetermined target stop position.

  Furthermore, the error calculation device 10 of the present embodiment converts the difference D1 between the position in the captured image 90 of the predetermined part and the reference position into the length in the real space, thereby stopping the train 1 and the target stop position. Calculate the error of For this reason, according to the error calculation device 10 of the present embodiment, the error between the stop position of the train 1 and the target stop position can be grasped with high precision with a full size numerical value.

  Furthermore, the error calculation device 10 according to the present embodiment extracts a plurality of circles 21 from the captured image 90, and each circle 21 is included from the difference in X coordinates and the difference in Y coordinates in the captured image 90 of the extracted circles 21. To determine which rows and columns are to be used, and when the number of circles 21 included in each column and each row matches a predetermined number determined in advance, the circles 21 included in each column and each row are marked as markers Identify as 2. For this reason, according to the error calculation device 10 of the present embodiment, the influence of dirt, patterns and the like on the station platform 4 is reduced to accurately grasp the error between the stop position of the train 1 and the target stop position. be able to.

  In addition, when the error calculation device 10 according to the present embodiment receives the stop signal from the control device 11, the marker 2 is identified from the captured image 90, so the stop position of the train 1 is detected only when the train 1 stops. A process of calculating an error from the target stop position is performed. Therefore, according to the error calculation device 10 of the present embodiment, the processing load can be reduced.

  In addition, the error calculation device 10 of the present embodiment transmits an error between the stop position of the train 1 and the target stop position to the control device 11 to provide feedback for the automatic operation control of the control device 11. Therefore, according to the error calculation device 10 of the present embodiment, it is possible to contribute to the improvement of the accuracy of the automatic driving control by the control device 11.

  In the present embodiment, the camera 3 is installed at the leading vehicle of the train 1, but it may be installed at a position where the marker 2 can be imaged, and the installation position is not limited to this. In addition, the installation position of the marker 2 is not limited to the vicinity of the target stop position of the train 1 and is installed behind the center 4 or near the center of the home 4 within the range of the length of the stopped train 1 Also good.

  Moreover, although the acquisition part 102 of the error calculation apparatus 10 of this embodiment is acquiring the captured image 90 from the camera 3 irrespective of whether the train 1 has stopped, the timing which acquires the captured image 90 is this. It is not limited to For example, the acquisition unit 102 may adopt a configuration for acquiring the captured image 90 from the camera 3 when the reception unit 101 receives the stop signal.

  Further, when it is determined that the marker 2 is not included in the captured image 90, the error calculation device 10 of the present embodiment transmits to the control device 11 that the error is larger than the maximum value “X cm” of the errors that can be calculated. However, the transmission content is not limited to this. For example, the error calculation device 10 may simply transmit that the marker 2 has not been identified.

  Further, the identifying unit 103 may send only the minimum X coordinate and the maximum X coordinate of the circle 21 to the calculating unit 104 instead of the X coordinates of all the circles 21.

  Moreover, although the error calculation apparatus 10 of this embodiment employ | adopted the structure which receives a stop signal from the control apparatus 11, the method of judging the stop of the train 1 is not limited to this. For example, the error calculation device 10 may determine that the train 1 has stopped by detecting the speed of the train 1 from the output value of a tacho generator (not shown) interlocked with the rotation of the wheels of the train 1.

  Further, each piece of information stored in the storage unit 150 of the error calculation device 10 of the present embodiment may adopt a configuration in which it is transmitted from the control device 11 or another external system or the like.

  Further, although the control device 11 controls the operation of the train 1 in the present embodiment, the control device 11 may be a device that supports a part of the operation, such as the TASC. Moreover, the train 1 may employ | adopt the structure manually operated by the driver. Moreover, when the said structure is employ | adopted, the information which shows the next stop station of the train 1 may be manually input by the driver, and it is good also as what is transmitted to the error calculation apparatus 10 from another external system. .

Second Embodiment
In the first embodiment, the marker 2 is a plurality of circles arranged such that a plurality of columns and a plurality of rows intersect. In this embodiment, the marker 2 has a rectangular shape with different colors for each route of the train 1 so that the stop position of the train 1 can be obtained even when a plurality of routes with different target stop positions are installed in the same station. An error with the target stop position is calculated with high accuracy.

  The overall configuration of the error calculation system S of this embodiment is the same as that of the first embodiment. The hardware configuration of the control device 11 and the error calculation device 10 is the same as that of the first embodiment.

  The error calculation device 10 according to the present embodiment includes the receiving unit 101, the acquiring unit 102, the identifying unit 103, the calculating unit 104, the transmitting unit 105, and the storage unit 150, as in the first embodiment. . The receiving unit 101, the acquiring unit 102, and the transmitting unit 105 have the same functions as in the first embodiment.

  The storage unit 150 of the present embodiment stores home position information, a marker length, the X coordinate of the reference position, the maximum value of the computable errors, and the color to be extracted. The home position information, the X coordinate of the reference position, and the maximum value of errors that can be calculated are the same as in the first embodiment.

  The marker length in the present embodiment is the length from the left end to the right end of the rectangular marker 2.

  The color to be extracted is the color of the marker 2 used to calculate the error of the stop position on the route on which the train 1 on which the error calculation device 10 is mounted travels. The color to be extracted is an example of a specific color in the present embodiment.

  The specifying unit 103 according to the present embodiment includes the function according to the first embodiment, and extracts the image range of the color to be extracted from the captured image, whereby the own device (error calculation device 10) is mounted. The marker 2 of the route on which the train 1 travels is identified.

  FIG. 6 is a diagram showing an example of a method of identifying the marker 2 according to the present embodiment. As shown in FIG. 6, markers 2 a to 2 c are shown in the captured image 2090. The markers 2 a to 2 c are rectangular and have different colors for each train 1 route. The color of the markers 2a to 2c is not particularly limited, but black, gray, yellow (color of the Braille block) and the like which are likely to be present on the home 4 may not be used.

  In the example of FIG. 6, it is assumed that the color matching the color to be extracted stored in the storage unit 150 is the color of the marker 2b. The identifying unit 103 identifies the marker 2 b by extracting an image range of the color to be extracted from the captured image 2090 using an image processing method.

  More specifically, the specifying unit 103 extracts, from the captured image 2090, an image range that matches the color to be extracted. The identifying unit 103 may use an RGB color space or an HSV color space as a method of identifying colors used to extract an image range. Then, the specifying unit 103 generates an image 91 in which the image range other than the color to be extracted is converted to black. In the present embodiment, in order to specify the image range of the marker 2b more accurately, the specifying unit 103 further performs known image processing such as binarization and noise removal. Then, the specifying unit 103 specifies the rectangular marker 2b from the image 93 after image processing.

  The identifying unit 103 sends the minimum X coordinate and the maximum X coordinate of the identified marker 2 b to the calculator 104.

  The calculation unit 104 of the present embodiment has the same function as that of the first embodiment, and then, based on the minimum X coordinate "x1" and the maximum X coordinate "x2" of the marker 2b acquired from the identification unit 103, The X coordinate "x3" of the center in the width direction of 2b is determined. Then, the calculation unit 104 converts the difference D1 between the X coordinate “x4” of the reference position stored in the storage unit 150 and the X coordinate “x3” into the length in the real space, whereby the stop position of the train 1 The error between the and the target stop position is calculated.

  Next, processing executed by the error calculation device 10 of the present embodiment configured as described above will be described. The flow of the error calculation process of this embodiment is the same as the flow of the process of the first embodiment described with reference to FIG. 4, but the flow of the marker identification process is different from that of the first embodiment.

  FIG. 7 is a flowchart showing an example of the flow of the marker identification process according to the present embodiment.

  The identifying unit 103 extracts the image range of the extraction target color stored in the storage unit 150 from the captured image 2090 by a known image processing method (S201).

  The identifying unit 103 performs image processing for enhancing the extraction accuracy on the image 91 after the image processing of S201. For example, in the image 91 shown in FIG. 6, a part of the marker 2c which is not the color of the extraction target is mixed. The identifying unit 103 performs binarization processing on the image 91 using a known technique such as binarization with a specific threshold value or binarization of Otsu (discrimination analysis method) (S202). Then, the specifying unit 103 performs noise removal on the image 92 after the binarization processing using a known technique such as expansion processing and contraction processing (S203).

  The identifying unit 103 determines the presence or absence of a rectangular shape in the image 93 after image processing for noise removal (S204). More specifically, the specifying unit 103 detects, from the image 93, an image range in which white pixels (white pixels) are continuously present in the horizontal direction at a third threshold value or more. When the image 93 has an image range in which a predetermined number or more of white pixels are continuously present in the horizontal direction, the specifying unit 103 determines that the image 93 after the image processing has a rectangle (S204 “Yes”). In this case, the identifying unit 103 identifies the rectangle detected from the image 93 after image processing as the marker 2b (S205). The third threshold is assumed to be predetermined.

  In addition, when there is no image range in which white pixels are continuously present in the horizontal direction and the third threshold or more exists in the image 93 after image processing, the specifying unit 103 determines that the image 93 after image processing does not have a rectangle ( S204 "No"). In this case, the identifying unit 103 determines that the marker 2b is not included in the captured image 2090 (S206). Here, the processing of this flowchart ends and returns to the error calculation processing.

  As described above, the error calculation device 10 according to the present embodiment extracts the image range of the specific color from the captured image 2090 to specify the rectangular marker 2 b of the route along which the train 1 on which the own device is mounted travels. Do. For this reason, according to the error calculation device 10 of the present embodiment, even when a plurality of routes having different target stop positions get into the same station, the error between the stop position of the train 1 and the target stop position can be made with high accuracy. It can be calculated.

  Further, in the present embodiment, since the marker 2 has a rectangular shape, the error calculation device 10 can easily identify the marker 2 b from the image 93 after image processing.

  In the present embodiment, the identifying unit 103 of the error calculation device 10 detects the rectangle on the condition that the length of the white pixel is continuous in the horizontal direction, but the length of the continuous white pixel is the condition on the other You may add it. In addition, another shape may be adopted as the marker 2.

  Further, in the present embodiment, the storage unit 150 of the error calculation device 10 does not store only the extraction target color of the route of the train 1 on which the own device is mounted, but the route of the train 1 of each route You may employ | adopt the structure which matches and memorize | stores the color of the marker 2 used for calculation of the difference | error of a stop position. Further, the color to be extracted may be stored in the control device 11 and may be transmitted from the control device 11 to the error calculation device 10.

(Modification 1)
In the above-described first and second embodiments, the marker 2 is installed on the floor of the station home 4, but the installation position of the marker 2 is not limited to this.

  FIG. 8 is a view showing an example of the installation position of the marker 2 according to the present modification. As shown in FIG. 8, the marker 2 of the present modification is installed on the pillar 41 of the home 4 so as to face the track side (the train 1 side). The marker 2 may be described with a paint etc. in the pillar 41, and the board, sheet, etc. on which the marker 2 is displayed may be stuck on the pillar 41. The marker 2 may be installed on the wall of the home 4 so as to face the track side. In this modification, any marker 2 of the first embodiment and the second embodiment may be employed.

  Moreover, in this modification, the camera 3 is installed at an upper imaging angle of the sideways of the train 1 so as to face the marker 41 installed on the pillar 41 or the wall of the home 4 or the like.

  By installing the marker 2 as in the present modified example, it is possible to reduce that the marker 2 is stepped on by a passenger of the train 1 or that the marker 2 is deteriorated due to rain or the like.

(Modification 2)
In the station where the home door is installed, the marker 2 may be installed on the track side of the home door. FIG. 9 is a view showing an example of the installation position of the marker 2 according to the present modification. The marker 2 of this modification is installed on the side of the track side of the home door 42 installed at the end of the home 4 so as to face the train 1. In this modification, any marker 2 of the first embodiment and the second embodiment may be employed.

  Moreover, in this modification, the camera 3 is installed at the lower part of the vehicle of the train 1 at a lateral imaging angle so as to face the marker 2 installed on the home door 42.

  By installing the marker 2 as in the present modification, it is possible to reduce that the space between the marker 2 and the camera 3 is shielded by a passenger, an obstacle or the like, and the stop position and the target stop position of the train 1 The error of can be grasped with higher accuracy.

(Modification 3)
Further, the installation position of the marker 2 is not limited to the top of the home 4 and may be below the home 4. FIG. 10 is a view showing an example of the installation position of the marker 2 according to the present modification. The marker 2 of the present modification is installed on the side of the floor 4 below the floor 4 so as to face the train 1. In this modification, any marker 2 of the first embodiment and the second embodiment may be employed.

  Further, in this modification, the camera 3 is installed at a lateral imaging angle further below the installation position of the camera 3 of the modification 2 so as to face the marker 2 installed on the side under the floor of the home 4 Ru.

  By installing the marker 2 as in this modification, deterioration of the marker 2 due to passage of passengers, weather, etc. is reduced, and between the marker 2 and the camera 3 is shielded by passengers, obstacles, etc. It can be reduced. Furthermore, since it is difficult for anyone other than the administrator or the like to contact the installation position of the marker 2 of this modification, it is possible to prevent tampering with the marker 2 and the like.

(Modification 4)
Moreover, the marker 2 may be installed on a track instead of the home 4. FIG. 11 is a view showing an example of the installation position of the marker 2 according to the present modification. The marker 2 of this modification is installed between the rails 6a and 6b of the track. In this modification, any marker 2 of the first embodiment and the second embodiment may be employed.

  Moreover, one camera 3c is installed in the train 1 of this modification. The camera 3 c is installed at the bottom of the train 1 at a downward imaging angle so as to face the marker 2.

  By installing the marker 2 as in the present modification, it is possible to reduce the blocking between the marker 2 and the camera 3 by a passenger, an obstacle or the like. In addition, since it is difficult for anyone other than the administrator or the like to contact the installation position of the marker 2 of this modification, it is possible to prevent tampering with the marker 2 and the like.

Third Embodiment
In the first embodiment, one marker 2 was installed at the station home 4. In the present embodiment, a plurality of markers are installed at different angles.

  FIG. 12 is a diagram showing an example of the entire configuration of the error calculation system S according to the present embodiment. The error calculation system S of the present embodiment includes a plurality of markers 2 n and 2 m, an error calculation device 7010, a camera 3, and a control device 11.

  The control device 11 has the same function as that of the first embodiment. Moreover, the camera 3 of this embodiment is installed in the lower part of the vehicle of the train 1 by the imaging angle of horizontal direction so that the markers 2 n and 2 m can be imaged. The camera 3 has the same function as that of the first embodiment.

  The markers 2n and 2m are installed at different angles on the station home 4 respectively. Hereinafter, when the markers 2 n and 2 m are not particularly distinguished, they are simply referred to as the marker 2. More specifically, the markers 2 are set to be imaged at different angles with respect to the camera 3. For example, the markers 2 are displayed on the plates 7a and 7b installed obliquely on the floor of the home 4 at different angles. The installation method is an example, and is not limited thereto. The number of markers 2 is not limited to two as long as it is plural.

  FIG. 13 is a side view showing an example of the marker 2 according to the present embodiment. As described above, by arranging the markers 2 n and 2 m at different angles, there is a high possibility that one of them can be identified from the captured image even if it is difficult to identify from the captured image due to light reflection or the like. Become. The installation angle of the marker 2 is not limited to the example shown in FIG.

  FIG. 14 is a view showing an example of a captured image 7090 obtained by capturing the marker 2 according to the present embodiment. The marker 2n of the present embodiment has a shape in which two straight lines cross each other. Moreover, the marker 2m is an elliptical shape. The shape of the marker 2 is not limited to this.

  The error calculation device 7010 according to the present embodiment includes the receiving unit 101, the acquiring unit 102, the identifying unit 103, the calculating unit 104, the transmitting unit 105, and the storage unit 150, as in the first embodiment. . The receiving unit 101, the acquiring unit 102, and the transmitting unit 105 have the same functions as in the first embodiment.

  The storage unit 150 of the present embodiment stores home position information, a marker length, the X coordinate of the reference position, the maximum value of the computable errors, and the offset parameter. The home position information and the maximum value of the computable errors are the same as in the first embodiment.

  The marker length in the present embodiment is the length in the width direction between the center in the width direction of the marker 2 n and the center in the width direction of the marker 2 m. The center in the width direction of the marker 2 n is a position where two straight lines intersect. The center in the width direction of the marker 2m is the center of the oval shape.

  The X coordinate of the reference position of the present embodiment is, as in the first embodiment, the coordinate of the center in the width direction of the marker 2 in the captured image captured in advance when the train 1 stops at the target stop position. The reference position in the present embodiment is a position that is the center in the width direction of all of the plurality of markers 2 in the captured image 7090 captured when the train 1 stops at the target stop position. The center in the width direction of the entire marker 2 in this embodiment is intermediate between the center in the width direction of the marker 2 n in the captured image 7090 and the center in the width direction of the marker 2 m. The center in the width direction of the entire marker 2 is an example of the predetermined portion in the present embodiment.

  The offset parameter is a parameter for correcting the center in the width direction of the marker 2 n on the captured image 7090 or the center in the width direction of the marker 2 m on the captured image 7090 to the center in the width direction of the entire marker 2. More specifically, the offset parameter is an estimate of the distance on the captured image 7090 between the center in the width direction of the marker 2 n or the center in the width direction of the marker 2 m and the center in the width direction of the entire marker 2 . The distance on the captured image 7090 is represented by, for example, the number of pixels.

  The identification unit 103 of the present embodiment identifies the marker 2 from the captured image 7090 using a known technique such as template matching after having the function of the first embodiment. More specifically, when at least one of the marker 2 n and the marker 2 m is specified, the specifying unit 103 determines that the marker 2 is specified from the captured image 7090. The identifying unit 103 sends the X coordinate of the center in the width direction of the identified marker 2 n or marker 2 m to the calculating unit 104.

  The calculation unit 104 of the present embodiment has the function of the first embodiment, and if the identification unit 103 identifies the marker 2, the position of the center of the identified one or two markers 2 in the width direction From this, the X coordinate of the center in the width direction of the entire marker 2 is determined.

  In addition, the calculating unit 104 compares the length in the width direction of the marker 2 on the captured image 7090 with the marker length stored in the storage unit 150, and sets the length on the captured image 7090 to the length in real space. Determine conversion parameters for conversion. Then, the calculation unit 104 converts the difference D7 between the X coordinate “x4” of the reference position stored in the storage unit 150 and the X coordinate “x73” of the center in the width direction of the entire marker 2 into a length in real space. By doing this, the error between the stop position of the train 1 and the target stop position is calculated.

  Next, an error calculation process performed by the error calculation device 7010 of the present embodiment configured as described above will be described. FIG. 15 is a flowchart showing an example of the flow of the error calculation process according to the present embodiment.

  The processes from the reception of the information indicating the stop station next to S1 to the process of reception of the stop signal of S4 are the same as the processes of the first embodiment.

  When the receiving unit 101 receives the stop signal (S4 “Yes”), the specifying unit 103 of the present embodiment specifies the marker 2 from the captured image 7090 by a known method (S71).

  When at least one of the marker 2 n and the marker 2 m is specified from the captured image 7090, the specifying unit 103 of the present embodiment determines that the marker 2 is specified from the captured image 7090 (S 6 “Yes”).

  When the number of markers 2 specified by the specifying unit 103 is “2” (S72 “Yes”), in other words, when the specifying unit 103 specifies both of the markers 2 n and 2 m, the calculating unit 104 is specified. The X coordinate "x73" of the center in the width direction between the X coordinate "x71" of the center in the width direction of the marker 2n and the X coordinate "x72" of the center in the width direction of the specified marker 2m is determined (S73).

  Further, when the number of markers 2 specified from the captured image 7090 by the specifying unit 103 is “1” (S72 “No”), the calculating unit 104 calculates the X coordinates of the center in the width direction of the markers 2n or 2m. The X coordinate "x 73" of the center in the width direction of the entire marker 2 is determined by adding or subtracting the offset parameter stored in the storage unit 150 (S74).

  Then, the calculation unit 104 calculates a difference D7 between the X coordinate "x73" of the center of the entire marker 2 in the width direction obtained in the process of S73 or S74 and the X coordinate "x4" of the reference position (S75).

  The conversion process of S8 and the transmission process of S9 are the same as in the first embodiment. Further, the process of S10 transmission when the specifying unit 103 does not specify the marker 2 from the captured image 7090 (S6 “No”) is also the same as that of the first embodiment.

  As described above, when one or more of the plurality of markers 2 installed at different angles are specified from the captured image 7090, the error calculation device 7010 according to the present embodiment determines the predetermined position from the center position of the specified marker 2. Find the position of the site. For this reason, according to the error calculation device 7010 of the present embodiment, even if it is difficult to specify the marker 2 from the captured image 7090 due to the illumination condition etc., if any one marker 2 can be specified from the captured image 7090, the train An error between the stop position of 1 and the target stop position can be calculated. For this reason, according to the error calculation device 7010 of the present embodiment, it is possible to reduce the influence of the illumination conditions and the like and grasp the error between the stop position of the train 1 and the target stop position with high accuracy.

  In addition, the shape of the markers 2n and 2m of this embodiment is an example, and you may apply the marker 2 of 1st Embodiment or 2nd Embodiment to this embodiment. Also, the marker 2 of the first embodiment and the marker 2 of the second embodiment may be combined and applied to the present embodiment.

  As described above, according to the first to third embodiments, the error between the stop position of the train 1 and the target stop position can be grasped with high accuracy.

  The error calculation program executed by the error calculation device 10 of the present embodiment is a file of an installable format or an executable format, such as a CD-ROM, a flexible disk (FD), a CD-R, a DVD (Digital Versatile Disk), etc. Is provided by being recorded on a computer readable recording medium.

  Further, the error calculation program executed by the error calculation apparatus 10 of the present embodiment may be stored on a computer connected to a network such as the Internet and provided by being downloaded via the network. Further, the error calculation program executed by the error calculation device 10 of the present embodiment may be provided or distributed via a network such as the Internet. In addition, the error calculation program of the present embodiment may be provided by being incorporated in advance in a ROM or the like.

  The error calculation program executed by the error calculation device 10 of the present embodiment has a module configuration including the above-described units (reception unit, acquisition unit, identification unit, calculation unit, transmission unit), and as an actual hardware The above units are loaded onto the main storage device by the CPU (processor) reading and executing the error calculation program from the storage medium, and the reception unit, acquisition unit, identification unit, calculation unit, and transmission unit are on the main storage device. Is supposed to be generated.

  While certain embodiments of the present invention have been described, these embodiments have been presented by way of example only, and are not intended to limit the scope of the invention. These novel embodiments can be implemented in various other forms, and various omissions, substitutions, and modifications can be made without departing from the scope of the invention. These embodiments and modifications thereof are included in the scope and the gist of the invention, and are included in the invention described in the claims and the equivalent scope thereof.

1 Train 2, 2a to 2c, 2n, 2m Marker 3, 3a, 3b Camera 4 Home 10, 7010 Error Calculation Device 11 Control Device 21, 21a to 21l Circle 90, 2090, 7090 Captured Image 91, 92, 93 Image 101 Received Part 102 Acquisition part 103 Identification part 104 Calculation part 105 Transmission part 150 Storage part D1, D7 Difference S error calculation system

Claims (8)

An acquisition unit configured to acquire a captured image from an imaging device installed on a train;
An identifying unit for identifying from the captured image a marker installed at a station at which the train stops;
A position in the captured image of the predetermined portion in the specified marker, and a reference position which is a position of the predetermined portion in the captured image captured in advance when the train is stopped at a predetermined target stop position; Calculating a difference between the stop position of the train and the target stop position based on the difference;
Error calculation device comprising: The predetermined site is the center in the width direction of the marker,
The calculation unit calculates an error between the stop position of the train and the target stop position by converting the difference between the position in the captured image of the predetermined part and the reference position into a length in real space. ,
The error calculation device according to claim 1. The specification unit extracts a plurality of circles from the captured image, and among the plurality of extracted circles, the circles having a difference in horizontal coordinates in the captured image equal to or less than a first threshold are in the same row It is determined that it is determined that the circle whose difference in vertical coordinate in the captured image is equal to or less than the second threshold is included in the same row, and the number of the plurality of circles included in the column and the row Identify the plurality of circles included in the column and the row as the marker, when the number of circles matches the predetermined number determined in advance
The error calculation device according to claim 1. The marker is rectangular and has a different color for each train route,
The specification unit specifies the marker of a route on which a train on which the error calculation device is mounted travels by extracting an image range of a specific color from the captured image.
The error calculation device according to claim 1. A plurality of the markers are installed at different angles, respectively.
The predetermined site is the center in the width direction of the plurality of markers,
The calculation unit obtains the position of the predetermined site from the position of the center of the specified marker when the specifying unit specifies one or more of the markers.
The error calculation device according to claim 1. The receiver further comprises: a reception unit that receives a stop signal indicating that the train has stopped from the controller of the train;
The identification unit identifies the marker from the captured image acquired by the acquisition unit when the reception unit receives the stop signal.
The error calculation device according to any one of claims 1 to 5. The transmission unit further transmits the error calculated by the calculation unit to the control device.
The error calculation device according to claim 6. An error calculation system comprising a marker and an error calculation device, comprising:
The marker is placed at the station where the train stops,
The error calculation device
An acquisition unit configured to acquire a captured image from an imaging device installed in the train;
An identification unit that identifies the marker from the captured image;
A position in the captured image of the predetermined portion in the specified marker, and a reference position which is a position of the predetermined portion in the captured image captured in advance when the train is stopped at a predetermined target stop position; Calculating a difference between the stop position of the train and the target stop position based on the difference;
An error calculation system comprising:

Images