JP2007113240A - Self-location detection method and facility of robot by camera sensor and distance sensor - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide a simple method and a facility capable of precisely making detection of a self-location of a robot autonomously traveling inside of a pipe conduit such as a sewer or the like in real time. <P>SOLUTION: A camera sensor and a distance sensor are mounted on the autonomous traveling type robot making detection of defects in the pipe conduit by autonomously traveling inside of the sewer or the like, description points including a manhole, a T-shaped junction, a Y-shaped junction, a differential diameter junction and a joint section, etc. are identified by the distance sensor, and an operational calculation of the self-location of the autonomous traveling robot is made as a distance from the description points by contrasting a number of edge pixels of an interested domain obtained from image data acquired by the camera sensor with a number of pixels in a table of a number of pixels corresponding to the distance from the description points formed in advance. <P>COPYRIGHT: (C)2007,JPO&INPIT

Description

  The present invention relates to a method for detecting the self-position of a robot that autonomously travels inside a pipe tub such as a sewer pipe and detects defects in the pipe tub in real time, and an apparatus therefor.

  It is important to detect the presence and level of defects inside the sewer pipes and to take necessary measures accurately in order to secure and maintain the safety of the infrastructure. Thus, a robot that autonomously travels inside a sewer pipe or the like and detects defects inside the pipe is used. It is important to accurately obtain the position of the defect inside the sewer pipe such as the sewer pipe captured by the autonomous traveling robot, that is, the position information inside the sewer pipe such as the autonomous traveling robot.

  The self-position of a traveling body such as an autonomous traveling robot that autonomously travels inside a sewer pipe, such as a sewer pipe, and the traveling information of the traveling body using encoder information that detects the rotational speed of a driving wheel or a separately disposed rotating wheel In general, detection is performed by calculating a distance. However, in a robot or the like that travels in a pipe, for example, slip often occurs between the inner wall surface of the pipe and a driving wheel or a separately disposed rotating wheel, resulting in an error in calculating a traveling distance based on the number of rotations of the wheel, It is difficult to determine the self-position of the autonomously traveling robot using only the encoder information.

  On the other hand, it is conceivable to correct the self-position of the traveling body such as the autonomous traveling robot when the traveling body such as the autonomous traveling robot arrives at a known feature point, for example, the position of the manhole in the traveling process. No means has yet been proposed for simply detecting feature points such as manholes, T-shaped joint points, Y-shaped joint points, different-diameter joint points, and joint portions in the equal pipe rod for each type of each feature point.

On the other hand, a self-position detection system such as a vehicle using GPS is known (for example, see Patent Document 1). However, a self-position detection system for a traveling body such as a vehicle using GPS cannot be applied as a self-position detecting means for a traveling body such as an autonomous traveling robot that travels inside a pipe or the like buried underground such as a sewer pipe. .
Japanese Patent Laid-Open No. 08-082526

There is also known a means for processing the image data acquired by a CCD camera or the like to know the distance from the camera installation position to the target (for example, see Patent Document 2). However, this prior art only provides a means for detecting the distance from the fixed camera installation point to the target, and determines the self-position of an autonomous traveling robot or the like that autonomously travels inside a pipe such as a sewer pipe. It does not teach means for real-time detection.
Japanese Patent Laid-Open No. 2002-042142

  As mentioned above, since there is no system for accurately detecting the self-position of an autonomous traveling robot etc. that autonomously travels inside a sewer pipe etc., the autonomous running robot etc. It was difficult to determine the acquisition position of the image data indicating the defect portion in the pipe acquired inside and to determine the self-position of the autonomous traveling robot or the like when the autonomous traveling robot or the like breaks down. An object of this invention is to provide the simple method and apparatus which detect the self-position of the autonomous traveling type robot which autonomously travels inside pipes, such as a sewer pipe, correctly and in real time.

  In order to solve the above-mentioned problem, the invention according to claim 1 is a method for detecting the self-position of a robot that autonomously travels inside a sewer pipe or the like and detects a defect or the like inside a pipe tub. A camera sensor and a distance sensor are mounted on an autonomous traveling robot, and feature points such as manholes, T-shaped joints, Y-shaped joints, different-diameter joints, joints, etc. in sewer pipes are identified by the distance sensor. By comparing the number of edge pixels of the image data acquired by the camera sensor with the number of edge pixels corresponding to the distance from each feature point created in advance, the self-position of the autonomous mobile robot is determined from the feature points. This is a self-position detection method for a robot using a camera sensor and a distance sensor that calculate and calculate the distance.

  The invention according to claim 2 is an apparatus for detecting the self-position of a robot that autonomously travels inside a sewer pipe and detects a defect or the like inside the pipe, and the autonomous traveling robot and the autonomous robot Measuring the radial distance of the sewer pipe etc. from the camera sensor mounted on the running robot and the autonomous running robot, and based on the result, manholes in the sewer pipe etc., T-shaped junction, Y-shaped junction, Distance sensor for identifying feature points such as joints with different diameters and joints, image processing means for performing edge processing on images acquired by the camera sensor, and edge pixels corresponding to the distance from the feature points, which are created in advance By comparing the number of edge pixels in the table of numbers with the number of edge pixels obtained by edge processing of the image acquired by the camera sensor, A self-position detecting apparatus for a robot by the camera sensor and a distance sensor and a processing unit for calculating calculated as the distance from the feature points.

  According to the present invention, feature points such as manholes, T-shaped joints, Y-shaped joints, different-diameter joints, and joints can be accurately detected for each type by the distance sensor, so each feature point is autonomous. The self-position of the traveling robot can be corrected, and the self-position of the autonomous traveling robot can be detected with high accuracy. Further, the self-position of the autonomously traveling robot can be calculated and calculated as the distance from each feature point based on the image data acquired by the camera sensor.

  By combining information from the distance sensor and camera sensor, the navigation map (pipe and feature point arrangement / position information) of the autonomous traveling robot can be reproduced. Furthermore, according to the present invention, since the quantified sensor information is used, the self-position of the autonomous traveling robot can be accurately determined. Further, in the present invention, when calculating the self-position of the autonomous traveling robot, only a part of the image data such as a region of interest such as a manhole existing portion of the image data acquired by the camera sensor is subjected to edge processing and the number of pixels. Since the distance from each feature point to the autonomous robot is calculated by comparing the number of edge pixels corresponding to the distance from each feature point created in advance. The amount of image data is greatly reduced, and the self-position of the autonomous traveling robot can be detected in real time.

  In the present invention, as an autonomous traveling robot (in-pipe traveling device) equipped with a camera sensor and a distance sensor, the autonomous traveling robot (shown in FIG. 12 and FIG. 13) proposed by the applicant in Japanese Patent Application No. 2005-017384 ( An in-pipe traveling apparatus) can be suitably used. As shown in FIGS. 12 and 13, the in-pipe travel device is supported by pivots 18 to 21 provided on the frame 2 and the front right side, the front left side, the rear right side, and the rear left side of the frame 2. First to fourth arms 3 to 6 that are rotatable in a horizontal plane, and first to fourth drives that are pivotally supported by one ends of the first to fourth arms 3 to 6 and rotated in the horizontal plane. It becomes the structure provided with the 1st and 2nd springs 15 and 16 which apply the restoring force around the rotating shafts 18 to 21 to the wheels 11 to 14 and the first to fourth arms 3 to 6.

  The autonomous traveling robot (in-pipe travel device) configured as described above can freely travel on a curved pipe, a pipe with a changing inner diameter, a pipe with a branch, etc., without using a complicated steering mechanism or external control mechanism. Therefore, it is useful as an inspection device for inspecting the inside of various pipes, especially water and sewage pipes buried in the ground, gas pipes and the like.

  In the present invention, the camera sensor mounted on the autonomous traveling robot (in-pipe traveling device) is preferably composed of one camera equipped with a fisheye lens. An autonomously traveling robot (in-pipe travel device) is a camera that is equipped with a fisheye lens-equipped camera (imaging device) and a distance sensor, autonomously moves inside a pipe such as a sewer pipe, and has a fisheye lens in the tube. The image data obtained by imaging in (1) is stored in a storage device such as a hard disk (not shown) provided in the body of the autonomous traveling robot (in-pipe traveling device) or transmitted to an external computer.

  Thus, the present invention is a system that detects the self-position of the autonomous traveling robot in real time by the fusion of the camera sensor and the distance sensor mounted on the autonomous traveling robot. A distance sensor measures the radial distance of a sewer pipe, etc. from an autonomously traveling robot, and features such as manholes, T-shaped joints, Y-shaped joints, different-diameter joints, and joints based on the results Is identified for each type. In addition, the image data in the tube is acquired by the camera sensor, and the edge processing of the region of interest such as the feature point is performed by the image processing means.

  On the other hand, a table (map) of the number of edge pixels corresponding to the distance from each feature point is created in advance. Thus, the number of pixels obtained by performing edge processing on the region of interest (ROI) of the image acquired by the camera sensor, and the number of edge pixels corresponding to the distance from each feature point in the table. In contrast, the self-position of the autonomously traveling robot is calculated and calculated as the distance from each feature point.

  FIG. 1 shows a self-position detecting device for a robot using a camera sensor and a distance sensor according to the present invention. As shown in FIG. 1, a distance sensor 34 including a camera sensor 32 and sensors 34-1 to 34-4 is mounted at the head of an autonomous traveling robot (in-pipe traveling apparatus) 31. The distance sensor 34 identifies feature points such as manholes, T-shaped joints, Y-shaped joints, different-diameter joints, and joints in the sewer piping network for each type, and image data and edges acquired by the camera sensor 32. By comparison with the pixel number table (map), the self-position of the autonomous traveling robot (in-pipe traveling device) 31 at that time is calculated and calculated by the computer as the distance from each feature point.

  In the present invention, a distance in the radial direction such as the distance from the autonomously traveling robot (in-pipe travel device) 31 to the inner wall surface of the sewer pipe or the like is measured by the distance sensor 34 including the sensors 34-1 to 34-4. Based on the results, feature points such as manholes, T-shaped joints, Y-shaped joints, different-diameter joints, and joints in the sewer piping network are identified for each type.

In this embodiment, the distance sensor 34 including the sensors 34-1 to 34-4 is model No. manufactured by Sharp Electric Co., Ltd. A GP2D120 infrared rangefinder is used. The present invention is not limited to this, and various distance sensors can of course be used.

  With reference to FIGS. 2 to 11, a process when the robot self-position detection method using the camera sensor and the distance sensor of the present invention is carried out will be described.

1) Image acquisition by camera sensor and calculation of self-position of autonomous traveling robot (in-pipe traveling device) 31 a. Calculation of the Number of Edge Pixels From the image data acquired by the camera sensor 32 mounted on the autonomous traveling robot (in-pipe traveling device) 31, for example, the image data acquired at the feature points such as the vicinity of the manhole shown in FIG. A region of interest ROI to be shown (manhole, T-shaped joint point, Y-shaped joint point, different-diameter joint point, joint portion, etc.) is extracted (see FIG. 2). Thereafter, edge processing is performed. Here, the edge processing is extraction of pixels emphasized in white in the image data, and an image as shown in FIG. 10 is obtained. In the stationary part of the main pipe where no feature point exists, the edge as shown in FIG. 10 does not appear and the whole is a black image. FIG. 11 shows an edge extracted from an image obtained at a position far from the manhole. In the obtained edge image, the number of edge pixels N _edge is calculated.

I. Creation of Distance Table Calculate and calculate the self-position of the autonomous traveling robot (in-pipe traveling device) 31 from the number of edge pixels N _edge of the region of interest (ROI) in the image data acquired by the camera sensor 32, as shown in FIG. Therefore, a distance table indicating the number of edge pixels corresponding to the distance from each feature point is created in advance. A distance table is created for each type of feature point. Generally, a sewer pipe or the like has the following feature points as shown in FIG.
a. The maximum distance between manholes M is 50 m.
b. The maximum length (interval of the joint J) of the main pipe P in a sewer pipe or the like is 5 m.
c. The T-shaped joint point and the Y-shaped joint point exist in the middle of the main pipe P (5 m length).
d. When the size (inner diameter or outer diameter) of the main pipe P is determined, the sizes of the pipes at the T-shaped joint point, the Y-shaped joint point, and the like are also determined.
From the above, some feature points appear within 5 m in the axial direction in sewer pipes and the like. Therefore, the distance between each feature point and the autonomous traveling robot (in-pipe traveling device) 31 is set to 0 to 500 cm, and a table of the distance and the number of edge pixels corresponding to the distance is created in advance. If the tube size at the feature point is outside the specified range, it can be incorporated as prior information, so that it is possible to create a distance corresponding edge pixel number table corresponding thereto.

C. Calculation of calculation of self-position of autonomous traveling robot Obtained by edge processing of region of interest (ROI) of image data acquired by camera sensor 32 mounted on traveling autonomous traveling robot (in-pipe traveling device) 31 By comparing the edge pixel number N _edge with the edge pixel number NT _edge in the distance corresponding pixel number table, the self-position of the autonomous traveling robot (in-pipe traveling device) 31 can be calculated and calculated as the distance from each feature point. It should be noted that the selection of the edge pixel number NT _edge in the pixel number table to be compared is performed accurately because it can be easily determined from which region of interest (ROI) in FIG. 3 the edge pixel number N _edge is obtained. be able to.

2) Feature Point Type Detection by Distance Sensor As described above, the number of edge pixels N _edge calculated from the edge of the region of interest (ROI) in the image acquired by the camera sensor 32 and the pixels in the distance corresponding pixel number table It is possible to discriminate the self-position of the autonomous traveling robot (in-pipe traveling device) 31 and the type of feature point by comparing with several NT _edges , but there are disturbances such as cracks, dirt, and inclusions in the main pipe P. Even in this case, an edge is extracted. In such a case, in the present invention, the self-position of the autonomous traveling robot (in-pipe traveling device) 31 is corrected by the distance sensor 34.

A. Confirmation of Information Obtained by Distance Sensor 34 As shown in FIG. 2, the number of edge pixels N _edge obtained by extracting the region of interest (ROI) of the image obtained by the camera sensor 32 is smaller than the value T. When it becomes large, the information by the distance sensor 34 is confirmed. Detection of the type of feature point (manhole, etc.) by the distance sensor 34 is performed twice by moving the autonomously traveling robot (in-pipe traveling device) 31 slightly.

Here, the defined value T is the number of edge pixels when there is no disturbance when the autonomous traveling robot (in-pipe traveling device) 31 reaches the feature point. That is, when there is no disturbance, T is the maximum value of the number of edge pixels N _edge calculated from the edge of the region of interest (ROI) extracted from the image data acquired by the camera sensor 32.

I. Detection of Feature Point Type The feature point type is detected by the distance sensor 34 as follows.
a. As shown in FIG. 5, when the value of the distance sensor 34-1 indicates the maximum value, the feature point is a manhole M.
b. As shown in FIG. 6, when the value of the distance sensor 34-4 shows the maximum value, the feature point is a T-shaped junction (right side). Further, when the value of the distance sensor 34-2 shows the maximum value, the feature point is a T-shaped joint point (left side).
c. As shown in FIG. 7, when the value of the distance sensor 34-4 fluctuates when the autonomous traveling robot (in-pipe traveling device) 31 is moved slightly and measured twice, the characteristic point is the Y-shaped joint point. (Right). Further, when the value of the distance sensor 34-2 changes when the autonomous traveling robot (in-pipe traveling device) 31 is slightly moved and measured twice, the feature point is a Y-shaped junction (left side). is there.
d. As shown in FIG. 8, when the values of all the distance sensors 34-1 to 34-4 fluctuate when the autonomous traveling robot (in-pipe traveling device) 31 is slightly moved and measured twice, the feature point is It is a different diameter joint point (stepped portion) D.

3) Distance correction by the distance sensor 34 As described above, since the type of the feature point (manhole, etc.) in the sewer pipe can be reliably determined by the distance sensor 34, the autonomous traveling robot (inside the pipe) When traveling (navigating) the traveling device 31, the self-position correction of the autonomous traveling robot (in-pipe traveling device) 31 is performed by comparing each identified feature point with a previously created sewer piping map. Is possible.

The figure which shows the self-position detection apparatus of the robot by the camera sensor and distance sensor which concern on one Example of this invention (a) is a side view, (b) is a front view. The block diagram which shows the processing flow of the self-position detection method of the robot by the camera sensor and distance sensor which concern on one Example of this invention 1 is a block diagram showing steps of calculating the number of edge pixels of a region of interest from an acquired image in a robot self-position detection method using a camera sensor and a distance sensor according to an embodiment of the present invention; FIG. 4 is a schematic diagram showing a calculation calculation stage of a self-position of an autonomous traveling robot and a detection position of a feature point by a distance sensor in a robot self-position detection apparatus and method using a camera sensor and a distance sensor according to an embodiment of the present invention; The schematic diagram which shows the pattern of the identification of the manhole by a distance sensor based on one Example of this invention The schematic diagram which shows the pattern of the identification of the T-shaped junction by a distance sensor based on one Example of this invention The schematic diagram which shows the pattern of the identification of the Y-shaped junction by a distance sensor based on one Example of this invention The schematic diagram which shows the pattern of the identification of the different diameter joint point by the distance sensor based on one Example of this invention The figure which shows the image (manhole part) acquired by the camera sensor based on one Example of this invention. The figure which shows the image (manhole vicinity) after the edge process of the region of interest of the image (manhole part) acquired by the camera sensor based on one Example of this invention. The figure which shows the image (manhole distant) after the edge process of the region of interest of the image (manhole part) acquired by the camera sensor based on one Example of this invention The top view which shows the autonomous traveling type robot (in-pipe traveling apparatus) based on one Example of this invention. The front view which shows the autonomous traveling type robot (in-pipe traveling apparatus) based on one Example of this invention.

Explanation of symbols

DESCRIPTION OF SYMBOLS 1 In-pipe traveling apparatus 2 Frame 3 1st arm 4 2nd arm 5 3rd arm 6 4th arm 7 1st drive motor 8 2nd drive motor 9 3rd drive motor 10 4th drive motor DESCRIPTION OF SYMBOLS 11 1st driving wheel 12 2nd driving wheel 13 3rd driving wheel 14 4th driving wheel 15 1st spring 16 2nd spring 17 Payload 18 Rotating shaft 19 Rotating shaft 20 Rotating shaft 21 times Dynamic axis 22 Tube 31 Autonomous traveling robot (in-pipe traveling device)
32 Camera sensor 34 Distance sensor M Manhole J Joint part T T-shaped joint Y Y-shaped joint P Main pipe D Different diameter joint

Claims (2)

  A method of detecting the self-position of a robot that autonomously travels inside a sewer pipe and detects a defect or the like inside a pipe tub, wherein a camera sensor and a distance sensor are mounted on the autonomous traveling robot. Identify feature points such as manholes, T-shaped joints, Y-shaped joints, different-diameter joints, joints, etc. in sewer pipes, and create the number of edge pixels in the image data acquired by the camera sensor in advance. A camera sensor that calculates and calculates the self-position of the autonomous mobile robot as a distance from the feature point by comparing with the number of edge pixels corresponding to the distance from each feature point. A robot self-position detection method using a distance sensor. An apparatus for detecting a self-position of a robot that autonomously travels inside a sewer pipe and detects a defect or the like inside a pipe tub, the autonomous traveling robot and a camera sensor mounted on the autonomous traveling robot And the distance in the radial direction of the sewer pipe or the like from the autonomous traveling robot is measured, and based on the result, the manhole, the T-shaped joint, the Y-shaped joint, the different-diameter joint, the joint, etc. A distance sensor for identifying a feature point; an image processing means for performing edge processing on an image acquired by the camera sensor; and the number of edge pixels in a table of edge pixels corresponding to a distance from a feature point, The number of edge pixels obtained by edge processing of the image acquired by the camera sensor is compared, and the self-position of the autonomous traveling robot is defined as the distance from each feature point. Self-position detecting apparatus for a robot by the camera sensor and the distance sensor; and a processor for calculating calculated.

Images