Obstacle detection device

Abstract

To mitigate a load of calibration work on attitude information of an obstacle detection device mounted onto a platform or a platform door.SOLUTION: A light projection part 121 emits light. A light receiving part 122 detects reflected light which is the light emitted by the light projection part 121 having hit an object and reflected. A scanning part 120 rotates the light projection part 121 and the light receiving part 122 to scan a space. An inertial measurement part 111 measures an inertial motion of an obstacle detection device 1. An arithmetic logical unit 112 calculates attitude information of an own device according to inertial measurement information. A range finding control part 113 calculates a distance to the object having reflected the light emitted by the light projection part 121 according to reciprocation time of the light, and generates a range finding image. The arithmetic logical unit 112 extracts a floor face-corresponding region from the range finding image generated by the range finding control part 113, and calibrates the attitude information until it indicates a horizontal level.SELECTED DRAWING: Figure 2

Description

Technical Field

[0001] The present invention relates to a technology for calibrating attitude information of an obstacle detection device.

Background Art

[0002] Platform screen doors (including platform fences) are devices installed on railway platforms and the like to prevent passengers from falling onto tracks or coming into contact with trains running on the tracks. These platform screen doors are provided with passage blocking members such as doors, fences, and ropes for blocking the passage of people. If there is an obstacle on the track of these passage blocking members, contact may cause an accident. Accordingly, obstacle detection devices incorporating distance measurement sensors for detecting obstacles are used in platform screen doors.

[0003] This obstacle detection device monitors a detection space near the platform screen door using the distance measurement sensor to detect the presence or absence of an obstacle, thereby performing abnormality determination. The distance measurement sensor used in the obstacle detection device is, for example, a 3D-TOF (Time of Flight) camera that irradiates light and three-dimensionally measures the distance to an object in space based on the flight time of the light.

[0004] Incidentally, the coordinates of the space measured by the distance measurement sensor are determined according to the attitude of the sensor itself, but this attitude may change unexpectedly due to sudden disturbance, accumulation of external force over a long period of time, or the like. Therefore, the attitude information of the distance measurement sensor needs to be calibrated.

[0005] Patent Document 1 discloses a camera device in which, when a position and orientation calibration routine is executed by an image processor, one or more planes in a range image captured by a 3D-TOF camera are detected, a reference plane is selected from the one or more detected planes, and position and orientation parameters of the 3D-TOF camera relative to the reference plane, for example, the height above the reference plane and / or the camera roll angle and / or the camera pitch angle are calculated.

Prior Art Literature

Patent Literature

[0006] [Patent Document 1] Special Publication No. 2011-530706 [Overview of the Initiative] [Problems that the invention aims to solve]

[0007] However, the technology described in Patent Document 1 requires the operator to roughly input the orientation of the installed camera and to select a reference surface from one or more detected surfaces, thus requiring the operator to be highly skilled.

[0008] One of the objectives of the present invention is to reduce the burden of calibration work for the attitude information of an obstacle detection device attached to a platform or platform door. [Means for solving the problem]

[0009] The present invention is attached to a platform or platform doors installed on said platform, generates a distance measurement image based on the round-trip time of light scanned toward the space including the platform floor, and based on the attitude information of the device obtained by inertial measurement. The distance measurement image is provisionally corrected, and the provisionally corrected The first embodiment of the present invention provides an obstacle detection device that extracts a region corresponding to the floor surface from the distance measurement image, calibrates the orientation information so that the extracted region is horizontal, and detects an obstacle in a detection space determined based on the calibrated orientation information and the floor surface indicated by the region.

[0010] According to the first embodiment of the obstruction detection device, the burden of calibration work for the attitude information of the obstruction detection device attached to the platform door can be reduced. In the first embodiment of the obstacle detection device, a second embodiment may be adopted in which, in the calibration of the attitude information, at least one of the following is performed: the area corresponding to the floor surface is divided into a near-range area and a far-range area, and the area is divided into a left-side area and a right-side area, and the attitude information is calibrated so that the height difference between the divided areas becomes small. In the obstacle detection device of the first embodiment, the information acquired by the inertial measurement is acceleration, and a third embodiment may be adopted in which the orientation and inclination of the device are calculated based on the acceleration and the distance measurement image is provisionally corrected.

[0011] In the first embodiment of the obstacle detection device, a jig having a reference height is placed on the floor surface, and after the calibration of the posture information, the reference height from the floor surface is stored using the detection result of the jig. 4 This may be adopted as one of the following forms.

[0012] According to 4 aspect of the obstacle detection device, after the posture information of the device itself is calibrated based on the area corresponding to the floor surface, the reference height from the floor surface can be stored. In the fourth embodiment of the obstacle detection device, the reference height consists of the installation height from the floor surface and a margin amount provided above the installation height, and when it is determined that the distance around the area corresponding to the floor surface has changed due to the placement of the jig, the height information to the floor surface is stored as the installation height, and the difference between the height of the floor surface and the height of the jig is stored as the margin amount, which may be adopted as a fifth embodiment.

[0013] In the first or 4 aspect of the obstacle detection device, a configuration may be adopted as the 6 aspect, in which a region corresponding to the floor surface is extracted from a distance measurement image generated after calibration of the posture information, and a warning is issued when the extracted region does not indicate that it is horizontal.

[0014] According to 6 aspect of the obstacle detection device, a user can know that an error occurring after calibration has exceeded a certain level.

[0015] In the first or 4 aspect of the obstacle detection device, a configuration may be adopted as the 7 aspect, in which a region corresponding to the floor surface is extracted from a distance measurement image generated after calibration of the posture information, and the posture information is recalibrated when the extracted region does not indicate that it is horizontal.

[0016] According to 7 aspect of the obstacle detection device, a user does not need to undertake a calibration work when an error occurring after calibration exceeds a certain level. BRIEF DESCRIPTION OF THE DRAWINGS

[0017] [Figure 1] FIG. 1 is a diagram showing an example of the external appearance of an obstacle detection device 1 according to an embodiment of the present invention. [Figure 2] FIG. 2 is a diagram showing an example of the overall configuration of an obstacle detection system 9. [Figure 3] FIG. 3 is a diagram showing the obstacle detection device 1 when the y-axis coincides with the direction of gravity. [Figure 4] FIG. 4 is a diagram showing the obstacle detection device 1 when the y-axis is deviated from the direction of gravity. [Figure 5]Flow chart illustrating an example of the operation flow of the obstacle detection apparatus 1. [Figure 6] Flow chart illustrating an example of the operation flow of posture calibration. [Figure 7] Figure illustrating an example of how a floor-corresponding region is adjusted by being divided into a near side and a far side. [Figure 8] Figure illustrating an example of how a floor-corresponding region is adjusted by being divided into a right side and a left side. [Figure 9] Flow chart illustrating an example of the operation flow for storing a margin amount. [Figure 10] Figure for explaining the margin amount. DETAILED DESCRIPTION OF THE INVENTION

[0018] Embodiment In several drawings described below, the space is represented as an xyz right-handed coordinate space. Among the coordinate symbols shown in the drawings, the symbol with a point inside a circle represents an arrow pointing from the back side to the front side of the drawing sheet. A direction along the x-axis in the space is referred to as the x-axis direction. Further, among the x-axis directions, the direction in which the x-component increases is referred to as the +x direction, and the direction in which the x-component decreases is referred to as the -x direction. For the y and z components, the y-axis direction, +y direction, -y direction, z-axis direction, +z direction, and -z direction are defined in accordance with the above definitions.

[0019] Figure 1 is a diagram illustrating an example of the external appearance of the obstacle detection apparatus 1 according to an embodiment of the present invention. Further, Figure 2 is a diagram illustrating an example of the overall configuration of an obstacle detection system 9. The z-axis in Figure 1 corresponds to the observation direction of the obstacle detection apparatus 1, and the y-axis and x-axis respectively indicate the axes for rotating the direction of projected light during main scanning and sub-scanning.

[0020] The obstacle detection system 9 illustrated in Figure 2 includes a platform screen door 2 and the obstacle detection apparatus 1 connected to the platform screen door 2. The obstacle detection apparatus 1 is an apparatus attached to a platform or to the platform screen door 2 installed on the platform, and includes a control system 11 and an optical system 12.

[0021] The optical system 12 includes a scanning unit 120, a light-emitting unit 121, and a light-receiving unit 122. The light-emitting unit 121 is a device that emits light, such as a laser diode or an LED (light-emitting diode). The light emitted by the light-emitting unit 121 is, for example, near-infrared light.

[0022] The light-receiving unit 122 has a light detection device such as a focusing lens or an avalanche photodiode (APD), or an image sensor such as a CCD (Charge Coupled Device) or CMOS (Complementary Metal Oxide Semiconductor), and detects the reflected light that is reflected when the light emitted by the light-emitting unit 121 hits an object.

[0023] The scanning unit 120 is a device that scans space by rotating the light-emitting unit 121 and the light-receiving unit 122 around a predetermined axis to change the direction of light emission and reception. The scanning unit 120 uses the aforementioned y-axis and x-axis for primary and sub-scanning to rotate (oscillate) the light-emitting unit 121 and the light-receiving unit 122. The space scanned by the scanning unit 120 includes the floor surface of the platform.

[0024] The control system 11 includes a power supply unit 110, an inertial measurement unit 111, a calculation unit 112, a distance measurement control unit 113, a scanning control unit 114, and a communication control unit 115. Of the control system 11, the inertial measurement unit 111, the calculation unit 112, the distance measurement control unit 113, the scanning control unit 114, and the communication control unit 115 may each be implemented by individual processors, controllers, etc. Alternatively, any of these may be implemented by a common processor. The processor used here is, for example, a CPU (Central Processing Unit). The control system 11 may also include RAM (Random Access Memory), ROM (Read Only Memory), a solid-state drive, a hard disk drive, etc.

[0025] The power supply unit 110 is a device that supplies power to each component of the obstruction detection device 1. The power supply unit 110 shown in Figure 2 receives power from the platform door 2 via electrical wiring.

[0026] The inertial measurement unit 111 is an inertial measurement device (IMU) that measures the inertial motion of the obstacle detection device 1. The inertial measurement unit 111 shown in Figure 2 has, for example, a gyro sensor and an acceleration sensor, which measure three-dimensional angular velocity and acceleration. Hereinafter, the information including the angular velocity or acceleration of the obstacle detection device 1 measured by the inertial measurement unit 111 will also be referred to as inertial measurement information.

[0027] The calculation unit 112 acquires inertial measurement information measured by the inertial measurement unit 111 and calculates attitude information, which represents the orientation and tilt of the device, based on this inertial measurement information. In other words, the calculation unit 112 performs a calculation to associate the direction of gravity identified from the attitude information with the coordinate axes set on the device. The calculation unit 112 supplies the calculation result to the distance measurement control unit 113.

[0028] Figure 3 shows the obstacle detection device 1 when the y-axis coincides with the direction of gravity. In Figure 3, az is the acceleration in the z-axis direction, and ay is the acceleration in the y-axis direction. In Figure 3, the direction of gravity lies on the yz-plane. In the example shown in Figure 3, if the acceleration due to gravity is g, then ay = g.

[0029] Figure 4 shows the obstacle detection device 1 when the y-axis is offset from the direction of gravity. The direction of gravity shown in Figure 4 lies on the yz-plane and has an inclination of θ with respect to the y-axis direction. In this case, the combined acceleration of az and ay is equal to the acceleration due to gravity. That is, az = g·sinθ and ay = g·cosθ.

[0030] Note that in the examples shown in Figures 3 and 4, the direction of gravitational acceleration lies on the yz plane, but it does not have to be on the yz plane. In this case, for example, the angle θ between the direction of gravity projected orthogonally onto the yz plane and the y-axis direction is determined, and furthermore, the angle φ between the direction of gravity and the yz plane is determined. The accelerations in the x-axis direction ax (not shown), ay, and az are calculated using the aforementioned θ, φ, and g, respectively, such that the combined acceleration is equal to the acceleration due to gravity.

[0031] The scanning control unit 114 shown in Figure 2 controls the scanning unit 120 described above. The communication control unit 115 is connected to the platform door 2 via a communication line and exchanges information with the platform door 2 through this communication line.

[0032] The distance measuring control unit 113 instructs the light emitting unit 121 to emit light. The distance measuring control unit 113 also acquires a detection signal for reflected light from the light receiving unit 122. The distance measuring control unit 113 then calculates the round-trip time of the light scanned by the optical system 12 toward the space including the platform floor. This round-trip time is the time from when the light emitting unit 121 emits light until the light receiving unit 122 detects the reflected light. Based on the round-trip time of the light described above, the distance measuring control unit 113 calculates the distance to the object that reflected the light emitted by the light emitting unit 121.

[0033] Furthermore, the distance measurement control unit 113 exchanges information with the scanning control unit 114 and the calculation unit 112 to identify the direction from which the light described above was irradiated, or the direction from which the reflected light described above was detected. The distance measurement control unit 113 then generates an image (distance measurement image) by reflecting the calculated distance to the object in the pixel value corresponding to the identified direction. In other words, the obstacle detection device 1 having this distance measurement control unit 113 is an example of an obstacle detection device that generates a distance measurement image based on the round-trip time of light scanned toward the space including the platform floor.

[0034] The calculation unit 112 extracts an area corresponding to the floor surface (hereinafter also referred to as the floor surface equivalent area) from the distance measurement image generated by the distance measurement control unit 113, and calibrates the attitude information until this floor surface equivalent area shows horizontal. Then, after calibrating the attitude information, the calculation unit 112 identifies the detection space based on the floor surface equivalent area, monitors the detection space in the sequentially generated distance measurement images, and determines whether or not there are any obstructions.

[0035] <Operation of the obstruction detection device> Figure 5 is a flowchart showing an example of the operation flow of the obstacle detection device 1. The control system 11 of the obstacle detection device 1 generates a distance measurement image (step S101). The control system 11 also acquires inertial measurement information, including acceleration measured by the inertial measurement unit 111 (step S102), and calculates attitude information indicating the attitude of the device (i.e., direction and tilt) based on this (step S103).

[0036] Then, the control system 11 performs a preliminary correction on the generated distance measurement image based on the calculated attitude information (step S104). Here, "preliminary correction" means roughly correcting the coordinates rather than correcting them in detail. For example, if the obstacle detection device 1 is installed upside down from its normal orientation, the control system 11 identifies this rough orientation and performs a preliminary correction, such as reversing the top and bottom of the distance measurement image accordingly.

[0037] After provisionally correcting the distance measurement image, the control system 11 extracts the area corresponding to the floor surface from the provisionally corrected distance measurement image (step S105). In other words, the obstacle detection device 1 having this control system 11 is an example of an obstacle detection device that extracts the area corresponding to the floor surface from a distance measurement image generated based on the attitude information of the device obtained by inertial measurement.

[0038] The control system 11 then calibrates the orientation information of the device based on the distance to the floor indicated by each pixel in the floor-equivalent area (step S200). The orientation information of the obstacle detection device 1 is calibrated so that the floor-equivalent area indicates horizontal. In other words, the obstacle detection device 1 having this control system 11 is an example of an obstacle detection device that calibrates the orientation information of the device so that the extracted floor-equivalent area indicates horizontal. Hereafter, the process of calibrating orientation information will also be referred to as orientation calibration.

[0039] Figure 6 is a flowchart showing an example of the flow of posture calibration. The control system 11 divides the floor-equivalent area into a near-field area close to the device and a far-field area far from the device (step S201). Then, the control system 11 identifies representative height values ​​for each of the near-field and far-field areas and determines whether the difference between the representative height values ​​is greater than or equal to a threshold (step S202). Representative values ​​include, for example, the arithmetic mean, geometric mean, median, etc.

[0040] If the control system 11 determines that the difference in the representative height values ​​is greater than or equal to a threshold (step S202; YES), it calibrates the attitude information so that the difference becomes smaller (step S203).

[0041] Figure 7 shows an example of how the floor area is adjusted by distinguishing it into a front side and a back side. In Figure 7, the horizontal axis represents the depth direction, and the vertical axis represents the height direction. The depth direction is the direction away from the obstacle detection device 1, along the track. The height direction is the height direction relative to the platform floor, and is the direction directly opposite to the direction of gravity before calibration of step S203.

[0042] Here, since the floor-equivalent area is identified based on the attitude information of the obstacle detection device 1, the direction of the normal of the floor-equivalent area may not coincide with the height direction before calibration in step S203. The control system 11 divides the floor point cloud, which is a point cloud representing the floor-equivalent area, into a front side and a back side, and calculates representative values ​​(arithmetic mean, etc.) of the heights of each side as the front side height and the back side height. Then, the control system 11 calibrates the attitude information of the obstacle detection device 1 so that the difference between this front side height and the back side height is less than a threshold.

[0043] For example, if the front height is higher than the rear height by a threshold, the control system 11 rewrites the attitude information of the obstacle detection device 1 so that the front height is lower or the rear height is higher. At this time, the control system 11 rotates the height direction by a predetermined angle around the x-axis, recalculates the difference between the front height and the rear height, and continues rotating at this angle until this difference falls within the threshold.

[0044] On the other hand, if it is determined that the difference in the representative values ​​of height is not greater than or equal to a threshold (step S202; NO), the control system 11 proceeds to the next process without performing step S203.

[0045] Next, the control system 11 divides the floor-equivalent area into a left-side area to the left of the device and a right-side area to the right (step S204). Then, the control system 11 identifies a representative height value for each of the left-side and right-side areas and determines whether the difference between the representative height values ​​is greater than or equal to a threshold (step S205).

[0046] If it is determined that the difference in the representative height values ​​is greater than or equal to a threshold (step S205; YES), the control system 11 calibrates the attitude information so that the difference becomes smaller (step S206).

[0047] Figure 8 shows an example of how the floor area is adjusted by distinguishing it into a right and left side. In Figure 8, the horizontal axis represents the depth direction, and the vertical axis represents the home direction. The depth direction is the direction away from the obstacle detection device 1, along the track, as in Figure 7. The home direction is the direction from the track towards the home. The height direction, as mentioned above, is perpendicular to both the depth direction and the home direction in Figure 8. Here, the direction towards the foreground of the paper in Figure 8 is the height direction.

[0048] The control system 11 divides the floor point cloud, which represents the area equivalent to the floor surface, into left and right sides, and calculates representative values ​​(such as arithmetic mean values) of the heights of each side as the left height and right height. The control system 11 then calibrates the attitude information of the obstacle detection device 1 so that the difference between the left height and the right height is less than a threshold.

[0049] For example, if the left-side height is higher than the right-side height by a threshold, the control system 11 rewrites the attitude information of the obstacle detection device 1 so that the left-side height is lower or the right-side height is higher. At this time, the control system 11 rotates the home direction by a predetermined angle around the z-axis, recalculates the difference between the left-side height and the right-side height, and continues rotating at this angle until this difference falls within a threshold.

[0050] On the other hand, if it is determined that the difference in the representative height values ​​is not greater than or equal to a threshold (step S205; NO), the control system 11 terminates the process without performing step S206.

[0051] In this operation, the processes from step S201 to step S203 and the processes from step S204 to step S206 may be performed in the reverse order.

[0052] As shown in the flowchart in Figure 5, after completing the process in step S200 described above, the control system 11 identifies the detection space based on the calibrated attitude information of the device itself and the floor surface indicated by the floor surface equivalent area (step S106). This detection space is the space that the obstacle detection device 1 targets for obstacle detection, and has defined upper and lower height limits relative to the floor surface. That is, the control system 11 identifies the height direction based on the calibrated attitude information of the device itself, and identifies the space that is above the lower height limit and below the upper height limit relative to the floor surface extracted from the distance measurement image as the detection space.

[0053] The control system 11, having identified the detection space, scans this detection space using the optical system 12 and detects obstacles within the detection space (step S107). In other words, the obstacle detection device 1 having this control system 11 is an example of an obstacle detection device that detects obstacles within a detection space determined based on the calibrated attitude information of the device itself and the floor surface indicated by the floor surface equivalent area.

[0054] As explained above, in the obstacle detection system 9, the obstacle detection device 1 extracts the platform floor-equivalent area from the orientation information of its own device based on the inertial motion measured by the inertial measurement unit 111 and the distance measurement image generated by the distance measurement control unit 113. The obstacle detection device 1 then calibrates its own orientation information so that this floor-equivalent area is horizontal, and detects obstacles in the detection space identified based on the configured orientation information and floor-equivalent area. As a result, users do not need to perform the task of calibrating the orientation information of the obstacle detection device 1.

[0055] The configurations, shapes, sizes, and arrangements described in the above embodiments are merely schematic representations to the extent that the present invention can be understood and implemented. Therefore, the present invention is not limited to the described embodiments and can be modified in various forms as long as it does not deviate from the scope of the technical idea set forth in the claims.

[0056] <Variation> The above describes the embodiment, but the contents of this embodiment can be modified as follows. Furthermore, the following modifications may be combined.

[0057] <1> In the embodiment described above, the obstacle detection device 1 identified a detection space above the lower limit height from the floor surface. However, the boundary surface at the lower limit of the detection space (hereinafter also referred to as the lower boundary surface) may be identified by measuring the distance to a physical jig or the like placed on the floor surface.

[0058] Figure 9 is a flowchart illustrating an example of the operation flow for memorizing the margin amount. For example, platform door 2 has an input panel (not shown) that receives user input. The user places a jig (hereinafter also called a boundary jig) on ​​the platform floor to identify the lower boundary surface of the detection space and inputs that this boundary jig has been placed into the input panel. This boundary jig is a physical structure that has a reference height (referred to as the reference height).

[0059] The control system 11 of the obstacle detection device 1 monitors user input from the input panel and determines whether or not a boundary jig has been placed (step S301). While it is determined that no boundary jig has been placed (step S301; NO), the control system 11 continues this determination.

[0060] On the other hand, if it is determined that a boundary jig has been placed (step S301; YES), the control system 11 refers to the newly generated distance measurement image to determine whether or not the distance around the area corresponding to the floor surface has changed (step S302).

[0061] If it is determined that the distance around the area equivalent to the floor surface has not changed (step S302; NO), the control system 11 terminates abnormally (step S303).

[0062] On the other hand, if it is determined that the distance around the area equivalent to the floor surface has changed (step S302; YES), the control system 11 stores the height to the floor surface as the installation height (step S304), and stores the height distance (i.e., the height difference) between the floor surface and the upper surface of the boundary jig as the margin amount (step S305).

[0063] Figure 10 is a diagram illustrating the margin amount. The obstacle detection device 1, through the operation described above, stores the installation height of the floor point cloud representing the area equivalent to the floor surface, and the margin amount corresponding to the reference height of the boundary jig from this floor point cloud.

[0064] In other words, the obstacle detection device 1 having this control system 11 is an example of an obstacle detection device that places a jig having a reference height on the floor surface and stores the reference height from the floor surface using the detection result of this jig after calibrating the attitude information of the device itself.

[0065] This configuration allows the obstacle detection device 1 to shorten the calibration of attitude information and the identification of the detection space by using the stored installation height and margin amount when detecting an obstacle again in the future.

[0066] <2> In the embodiment described above, after completing the attitude calibration process shown in step S200, the control system 11 identifies the detection space and detects obstacles within the detection space. However, if the floor-equivalent area extracted from the distance measurement image generated after attitude calibration does not show horizontal, the system may warn the user.

[0067] In other words, the obstacle detection device 1 having this control system 11 is an example of an obstacle detection device that extracts an area corresponding to the platform floor from a distance measurement image generated after calibration of the device's attitude information, and issues a warning if this extracted area does not indicate a horizontal position.

[0068] With this configuration, even if the user's device's posture changes due to aging or sudden external forces after the posture information has been calibrated, and the range monitored by the distance measurement image changes, the user can still be aware of that change.

[0069] <3> Furthermore, the control system 11 may recalibrate the attitude information of its own device if the floor-equivalent area extracted from the distance measurement image generated after attitude calibration does not indicate horizontality.

[0070] In other words, the obstacle detection device 1 having this control system 11 is an example of an obstacle detection device that extracts an area corresponding to the platform floor from a distance measurement image generated after calibration of the device's attitude information, and recalibrates the attitude information if this extracted area does not indicate horizontality.

[0071] With this configuration, even if the user's device's orientation changes due to aging or sudden external forces after initially calibrating its orientation information, and the range monitored by the distance measurement image changes, they can recalibrate their device's orientation information and re-identify the detection space.

[0072] <4> In the embodiment described above, the control system 11 had a CPU, but it may have other configurations. For example, the control system 11 may include a GPU (Graphics Processing Unit), an ASIC (Application Specific Integrated Circuit), an FPGA (Field Programmable Gate Array), a programmable logic device, etc. [Explanation of Symbols]

[0073] 1... Obstacle detection device, 11... Control system, 110... Power supply unit, 111... Inertial measurement unit, 112... Calculation unit, 113... Distance measurement control unit, 114... Scanning control unit, 115... Communication control unit, 12... Optical system, 120... Scanning unit, 121... Light projection unit, 122... Light receiving unit, 2... Platform door, 9... Obstacle detection system.

Claims

1. An obstacle detection device that is attached to a platform or platform doors installed on the platform, generates a distance measurement image based on the round-trip time of light scanned toward the space including the platform floor, provisionally corrects the distance measurement image based on the attitude information of the device obtained by inertial measurement, extracts an area corresponding to the floor from the provisionally corrected distance measurement image, divides the extracted area corresponding to the floor into a near-range area close to the device and a far-range area far from the device, and divides it into a left-side area to the left of the device and a right-side area to the right of the device, and rewrites the attitude information so that the difference in the representative height values ​​of each divided area is less than a threshold, thereby calibrating the attitude information so that the area corresponding to the floor shows horizontal, and detects an obstacle in a detection space determined based on the calibrated attitude information and the floor indicated by the area.

2. The obstacle detection device according to claim 1, wherein the information obtained by the inertial measurement is acceleration, and the orientation and tilt of the device are calculated based on the acceleration to provisionally correct the distance measurement image.

3. The obstacle detection device according to claim 1, wherein a jig having a reference height is placed on the floor surface, and the reference height from the floor surface is stored using the detection result of the jig after the calibration of the posture information.

4. The obstacle detection device according to claim 3, wherein the reference height consists of the installation height from the floor surface and a margin amount provided above the installation height, and when it is determined that the distance around the area corresponding to the floor surface has changed due to the placement of the jig, the height information to the floor surface is stored as the installation height, and the difference between the height of the floor surface and the height of the jig is stored as the margin amount.

5. An obstacle detection device according to claim 1 or 3, which extracts a region corresponding to the floor surface from a distance measurement image generated after calibration of the posture information, and issues a warning if the extracted region does not indicate horizontality.

6. An obstacle detection device according to claim 1 or 3, wherein the device extracts a region corresponding to the floor surface from a distance measurement image generated after calibration of the posture information, and recalibrates the posture information if the extracted region does not indicate horizontality.

Images