JP2002154080A - Automatic guided vehicle - Google Patents

Abstract

PROBLEM TO BE SOLVED: To resolve a problem that it takes a long time to process positional correction before starting transferring work after an automatic guided vehicle arrives at a station in a conventional system. SOLUTION: This automatic guided vehicle system is used for transporting articles by the automatic guided vehicle 1 provided with a robot arm 20 having an image recognition device 30. In this system, plural correction marks 60 are formed in the station 50 for transferring the articles, a reference mark 61 is formed at the center among the correction marks. When one of the correction marks is disposed at the center of the recognized image, all the correction marks and the reference mark are so set as to be included in the visual field region of the image recognition device, the correction marks and the reference mark are set bilaterally symmetrical and two of the correction marks are bilaterally symmetrically disposed with respect to the reference mark.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

The present invention relates to an automatic guided vehicle (AG)
The present invention relates to a technique for arranging a reference mark on a station in order to accurately and quickly operate a robot arm mounted on a station V) when performing a transfer operation at the station.

[0002]

2. Description of the Related Art For example, when articles are transferred by an unmanned carrier equipped with a robot arm in a factory or the like,
After the automatic guided vehicle stops at a predetermined work station, the robot arm mounted on the automatic guided vehicle is expanded and contracted to a predetermined position to perform the work, but the automatic guided vehicle performs a predetermined operation based on the map information. Even if the vehicle stops at the position, it is inevitable that a slight error occurs between the wheel and the normal stop position due to the friction between the wheels and the road surface, the effectiveness of the brake, and the like. In order for the robot arm to perform work accurately after the automatic guided vehicle stops, it is necessary to correct an error between the stop position and the operation reference position of the robot arm.

In order to correct the position error, conventionally, two marks are arranged at predetermined positions of a station, and the mark is recognized by a visual recognition means provided on a robot arm to calculate the error. The work was performed by correcting the position data taught in advance for the work.

[0004]

However, in a system for recognizing two marks, if the stop error of the automatic guided vehicle is large, the obtained image may include only one mark. . At this time, conventionally,
The robot arm mounted on the automatic guided vehicle is moved horizontally so that the recognized mark is located at the center of the image,
Two marks can be recognized in an image. The moving coordinates are stored. Then, after calculating the shift angle of the two marks with respect to the x-ray (y = 0) of the reference coordinates and the coordinates (shift coordinates) of the center position between the two marks, the two coordinates are calculated from the moving coordinates and the shift coordinates. The correction value was calculated.

As described above, when only one mark can be recognized, a two-stage process is required, and the robot arm also needs to be driven, so that it takes time to search for a reference position. That is, it takes time to correct the position before the automatic guided vehicle arrives at the station and starts the transfer work,
Shortening of the correction processing time was desired. Therefore, an object of the present invention is to make it possible to further easily correct the stop position error of the automatic guided vehicle.

[0006]

According to one aspect of the present invention, there is provided an unmanned transport system for transporting articles by an unmanned transport vehicle equipped with a robot arm having an image recognition device. A plurality of correction marks are provided on a station on which the correction mark is mounted, and a reference mark is provided at the center between the correction marks.

According to a second aspect of the present invention, when any one of the correction marks according to the first aspect is arranged at the center of the recognized image, all the correction marks and the reference marks are aligned with the field of view of the image recognition apparatus. It was set to be within the area.

In the present invention, the reference mark and the correction mark are symmetrical, and two correction marks are arranged symmetrically with respect to the reference mark.

[0009]

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS Hereinafter, an embodiment of an automatic guided vehicle system according to the present invention will be described with reference to the drawings. FIG. 1 is a diagram showing a schematic configuration of the automatic guided vehicle system of the present embodiment, and FIG. 2 is a conceptual diagram showing a control configuration of the automatic guided vehicle system. The automatic guided vehicle 1 controlled by the automatic guided vehicle system according to the present invention includes a traveling carriage unit 10 and a robot arm unit 2.
0 and an image recognition unit (image recognition device) 30.

The automatic guided vehicle 1 moves over a plurality of work stations 50. These work stations 50
Is given as map information in advance. The automatic guided vehicle 10 can stop at a predetermined stop position SP (FIG. 3) facing the predetermined work station 50 based on the map information. The position of the stop position SP is
It is represented by the absolute coordinate system sigma _A. The movement of the automatic guided vehicle is controlled by the automatic guided vehicle control unit 11 (FIG. 2).

The robot arm unit 20 includes a traveling carriage unit 10
Mounted on top. As shown in FIG. 1, the robot arm unit 20 has a plurality of degrees of freedom, and has a hand 25 at the end thereof so that articles can be transferred. Due to these degrees of freedom, the tip of the robot arm 20 is
It can be moved to a predetermined position. The movement of the distal end of the robot arm unit 20 is performed based on work position data taught in advance. This position data is obtained by teaching the position of the end portion (end effector) of the robot arm unit 20 by, for example, a teaching operation of the operator. These operations are controlled by the robot arm control unit 21 (FIG. 2). The position data of the distal end of the robot arm portion 20, in the robot coordinate system sigma _B, it is described. The robot coordinate system Σ _B is a coordinate system fixed to the traveling carriage 10 serving as a base of the robot arm 20.

Further, as shown in FIG.
Has an image input unit 32 and an image processing unit 34. The image input unit 32 has a camera 32a (FIG. 1) fixed near the tip of the robot arm unit 20. As the camera 32a, a two-dimensional CCD camera or the like can be used. The image captured by the image input unit 32 is
The processing is performed by the image processing unit 34. By this image processing, correction marks 60, 60 and a reference mark
Is recognized, and a correction value is obtained from the position of the reference mark 61.

After the automatic guided vehicle 1 stops, the traveling carriage unit 1
0 robot arm portion 20 which is mounted on, first, on the basis of the position data described above the robot coordinate system sigma _B, the initial position WP0 (5, 7) moving the tip up to.

FIG. 3 is a plan view showing the relationship between the automatic guided vehicle 1 and the work station 50. In FIG. 3, after the automatic guided vehicle 1 moves in the direction of the arrow AR1 and stops at the predetermined stop position SP, the robot arm unit 20 moves to the initial position W.
This shows a state in which it has moved to P0. On the work station 50, a work 70 to be worked is placed. Further, two correction marks 60 and 60 and a reference mark 61 are fixed in the vicinity of the work 70. As will be described later, the positions of these marks are checked, and the error of the stop position of the automatic guided vehicle 1 is corrected. The robot arm unit 20 can perform the work on the work 70 at an accurate position. Conventionally, since only two correction marks 60 are provided, it takes a long time for the correction control after the tip of the robot arm 20 has moved to the predetermined initial position WP0.

Next, a correction mark 6 used for position correction
0, the reference mark 61 will be described. As shown in FIG. 3, two correction marks 6 are provided on a predetermined portion in the work space, in this embodiment, on the upper surface of the work station 50 on the automatic guided vehicle 1 side.
0 and a reference mark 61 are arranged. The two correction marks 60 are formed to have the same shape and the same size, and one reference mark 61 is configured to have a different shape or size from the correction mark 60.
Are arranged at exactly intermediate positions between the correction marks 60. That is, assuming that a line connecting the centers (centers of gravity) of the correction marks 60 is the left-right direction, the correction marks 60 are arranged at the left and right centers on the line. In other words, the correction marks 60 and 60 and the reference mark 61 are arranged symmetrically left and right, and the correction marks 60 and 60 are arranged so as to be symmetrical with respect to the reference mark 61. By doing so, it is possible to prevent the installation from being performed in a different orientation.Furthermore, the installation operator can install the device without recognizing the orientation of the mark. Can be shortened.

The correction marks 60 and 60 are circular marks of the same size, and the reference mark 61 is a circular mark smaller than the correction mark 60 and has a cylindrical shape protruding above the work station 50. Make up. In this embodiment, the reference mark 61 is smaller than the correction mark 60, but the correction mark 6
It can also be a variant configured to be larger than zero. Here, a mark having anisotropy in its shape itself,
For example, the position of the mark itself is specified by making a polygon such as a triangle or a quadrangle, or a star shape, or the position of the mark itself is changed by changing the shape and size between three marks. It is also possible. However, it is preferable to use two non-special marks of the same shape. The use of three marks of a simple shape has an advantage that the position can be easily recognized by an image. In addition, installation and manufacture can be simplified, and the configuration can be made at low cost.

Next, a method of recognizing the correction mark 60 and the reference mark 61 and correcting a stop error of the automatic guided vehicle 1 will be described. The solid line in FIG. 5 represents a captured image when the initial position WP0 of the distal end of the robot arm 20 exactly matches the theoretical position (reference position) WP1. In other words, it shows a captured image of the image input unit 32 when there is no position error.

[0018] In the state this initial position WP0 are no errors, as shown in FIG. 5, in the camera coordinate system sigma _C in the field of view region R, the reference mark 61 _{X C = 0, Y C =} 0
, And the two correction marks 60 are located side by side on a straight line Y _C = 0 parallel to the X _C axis at the center in the Y _C direction. Note that X _C = 0 and Y _C = 0 are the origin positions. This, Y _C direction stop error of the automatic guided vehicle 1 X _C
It is based on the fact that it is smaller than the stop error in the direction. By arranging the two marks in a direction in which the stop error is relatively large, the possibility that at least one mark is imaged becomes higher. For example, the distance S between marks in the _XC direction
Either of the two marks can be included in the field of view even if a position error 1.5 times of the above occurs, but the two correction marks 60 are moved to X _C = 0 (Y _C direction). Tile arranged, because disappear from two to as the viewing area when 1.5 times of the position error of the distance between marks S occurs in the X _C direction.

The field of view will be described. In FIG. 4, the distance between the marks of the two correction marks 60 (between the positions of the centers of gravity) is represented by S, and the length L of one side of the field of view is defined as L.
Assuming that the mark radius of the correction mark 60 is r, the visual field region L corresponding to the image is determined so as to satisfy the following condition. L ≧ 2 × (S + r) Equation 1

As shown in FIG. 4, the center of gravity of one correction mark 60a is set at the center of the visual field (X _C = 0, Y _C =
When it is positioned at 0), the center of gravity of the other correction mark 60b exists on the circle having the radius S. Therefore, if the condition of Expression 1 is satisfied, one of the correction marks 60 (for example, 60
If only a) can be recognized, one of the correction marks 60a is positioned at the center of the screen, and the other correction mark 60b is also included in the field of view. Can calculate.

In the above equation 1, a condition for the radius r is added. This assumes that the position of the center of gravity of the correction mark 60 cannot be specified unless the whole of the correction mark 60 exists in the visual field region R. Therefore, when the position of the center of gravity of the correction mark 60 can be specified by the fact that only a part of the correction mark 60 exists in the visual field region R, the value that can be taken by the length L of one side of the visual field region R can be determined. It can be even smaller.

The image input section 32 can be moved by moving the robot arm section 20 horizontally.
As shown in FIG. 7, the image input unit 32 is disposed so as to be substantially opposed to a portion where two correction marks 60 and a reference mark 61 are set. That is, as in this embodiment, the upper surface of the work station 50 is a horizontal surface,
And two correction marks 60 and a reference mark 6 on the horizontal plane.
When 1 is set, the attitude of the image input unit 32 at the initial position is substantially downward.

In this state, when only one mark (for example, 60a) is picked up in the field of view, in this embodiment, the image input section 32 is horizontally moved in the direction of the arrow AR1. Thereby, the two correction marks 60 and the reference mark 61 are put in the field of view without lowering the measurement position accuracy based on the image recognition. It should be noted that the image input unit 32 is vertically upward (arrow A
When moved in the direction of (R2), the two correction marks 60 can be put in the field of view, but the accuracy of the measurement position is reduced. Therefore, only the horizontal direction is used without moving in the vertical direction. .

The two correction marks 60 and the reference mark 6
1 may be set in the work station 50 as in this embodiment, or may be attached to the work 70 itself. In any case, two correction marks are provided at the initial position of the image input unit 32. It is in a posture substantially opposed to a virtual line (hereinafter referred to as “inter-mark direction”) connecting 60. If the image input unit 32 is moved in a plane substantially perpendicular to the imaging axis AX1 in this state, the two correction marks 60 can be simultaneously placed in the field of view without lowering the measurement position accuracy.

In such a configuration for recognizing a field of view and a mark, even when there is a large deviation, correction can be made. That is, when only one of the correction marks 60 is within the visual field region R, the position of the reference mark 61 cannot be specified, so that the robot arm unit 20 is operated to horizontally move the image input unit 32 as described above. Adjustment is made so that the correction mark 60 is located at the origin. This is called preliminary adjustment. With this preliminary adjustment, the two correction marks 60 and 60 and the reference mark 61 can be recognized, and a correction value can be obtained based on the position and inclination of the reference mark 61.

However, with the current stopping accuracy of the automatic guided vehicle 1, it is difficult to stop all the correction marks 60 and the reference marks 61 at positions within the visual field region R in most cases. The mark 60 and the reference mark 61 can be stopped at a position within the visual field region R. Therefore, in the present invention, when the positions of the one correction mark 60 and the reference mark 61 can be recognized, the coordinates (X ₀ , Y ₀ ) where the reference mark 61 is located, the center of gravity of the one correction mark 60 and the reference mark 61 By finding the angle θ of the line connecting the centers of gravity of
Easily find the correction value, shorten the time required for correction control,
They are ready to move on to the next task.

Next, the control method of the present invention will be described with reference to FIG. FIG. 6 is a flowchart showing a control algorithm when the actual work at each work station is performed after teaching.

First, the automatic guided vehicle 1 stops at a predetermined stop position facing a predetermined work station 50 (step SP10). Next, the robot arm unit 20
Operates relative to the stopped automatic guided vehicle 1,
The tip of the robot arm 20 is moved to a predetermined initial position WP0.
(Step SP20). Then, the image input unit 32 fixed to the robot arm unit 20 captures an image of an area corresponding to the positions of the two correction marks 60 and the reference mark 61 (step SP30).

The image captured by the image input unit 32 is processed by the image processing unit 34. This process, the center of gravity of the mark represented by the camera coordinate system sigma _C is obtained. Here, as shown in FIG. 1, the camera coordinate system Σ _C is a coordinate system fixed to an image captured by the image input unit 32. As a result of the above processing by the image processing unit 34, the center of gravity position is recognized for the number N of correction marks (N = 0, 1, 2) and the number M of reference marks (M = 0, 1). In the following, depending on the number N of the correction marks 60 and the number M of the reference marks 61 whose center of gravity is recognized,
Since the processing is different, the number of marks N and M are determined in step SP40, and the corresponding processing is performed.

(1) First, a case where the number N of correction marks 60 photographed in an image is 1 or more (N ≧ 1) and the number M of reference marks 61 is 1 (M = 1) will be described. I do. That is, when the automatic guided vehicle 1 is stopped, in most cases, the correction mark 60 and the reference mark 61 can be recognized, and this control is performed. One correction mark 60 and reference mark 61
Is recognized in the image, the working position of the robot arm unit 20 can be corrected based on the position. Specifically, as shown in FIG. 5, when stopping at a position shifted by a two-dot chain line, the correction mark 60a ′ and the reference mark 6
When 1 ′ is recognized, the displacement amount (X ₀ , Y ₀ ) of the reference mark 61 with respect to the origin position of the camera coordinate system Σ _{C and} the displacement amount θ (−90 (deg) ≦ θ ≦ 90 (deg)) related to the attitude angle ) And ask. These shift amounts are corrected with respect to the work position data in the control algorithm taught in advance by teaching (step SP60). That is,
The reference mark 61 'is positioned at the reference mark 61 and the angle θ
A simple correction just by rotating is sufficient. Subsequently, the work is started (step SP7).
0).

In the case of the arrangement as in this embodiment, it is practically impossible that the attitude of the automatic guided vehicle 10 is shifted by 90 degrees or more. Therefore, the shift amount θ with respect to the posture angle can be specified in the range of −90 (deg) ≦ θ ≦ 90 (deg). In other words, the position of the other correction mark 60 can be determined based on the magnitude of the corresponding coordinate value in the captured image where the reference mark 61 is located, and the corresponding relationship can be specified only by the one correction mark 60 and the reference mark 61.

Although the following control is hardly performed, the following control can also be performed because it is assumed that a maintenance time such as abrasion of the brake is near or a position shift occurs due to contact with an obstacle or the like. Like that. In the following cases, the stop accuracy is reduced, so that control for issuing an alarm may be accompanied. (2) The case where the number N of the correction marks 60 is 1 (N = 1) and the number M of the reference marks 61 is 0 (M = 0) will be described. FIG. 8 is a diagram illustrating the relationship between the image and the position of the correction mark 60 in this case. As shown in FIG. 8, when only one mark is recognized, the robot arm unit 20 is re-moved in the lateral direction so that the other mark is included in the visual field region R of the camera 32a (step SP50). .
That is, preliminary adjustment is required. For example, a position WP where the position of one of the already recognized correction marks 60a exists at the center position (X _C = 0, Y _C = 0) of the screen.
By moving the robot arm unit 20 to 1 again,
The other correction mark can also be reliably captured in the viewing area. This is based on the fact that the distance S between marks and the length L of one side of the viewing area satisfies the relationship shown in Expression 1.

This operation will be described. FIG. 8 is a diagram illustrating a case where one correction mark 60a exists in the visual field region R, and the other correction mark 60b and the reference mark 61 exist outside the visual field region R. Assuming that the captured image when the camera 32a is moved to the initial position WP0 is shown in FIG. 8, only one correction mark 60a is recognized, so the center of gravity of the other correction mark 60b is Presumed to be external.

Next, one of the correction markers already recognized
Robot so that the position of the mark 60a is at the center of the screen.
The arm 20 is moved again. That is, the recognized
The position of the center of gravity of the correction mark 60a from the center of the image
Move again based on the amount. Of the robot arm unit 20
The amount of movement is based on the camera coordinate system Σ_CVector quantity at
(X ₁, Y₁). Robot arm unit 20
When the camera 32a moves again, the camera 32a and the X_C, Y_CPlane
Robot arm
Move 20 again.

The positional relationship between the two correction marks 60 and the reference mark 61 after the robot arm unit 20 is re-moved is indicated by a two-dot chain line in FIG. It will be in R. FIG. 7 shows the position WP0 of the image input unit 32,
FIG. 3 is a diagram showing WP1. The position WP0 is the initial position of the robot arm unit 20, and the field of view R at that time is shown in FIG.
Is shown in On the other hand, the reference position WP1 is a position after the robot arm unit 20 has re-moved.

After the re-movement operation in the oblique horizontal direction, the flow returns to step SP30, and after the image is captured again, the number N, M of marks is determined in step SP40. In this case, all the marks are moved to the field of view R
Therefore, N ≧ 1 and M = 1, and the operation (1) described above is performed. However, in calculating the correction amount of the position data, it is necessary to consider the movement amount (X ₁ , Y ₁ ) at the time of the re-movement.

(3) The number N of the correction marks 60 is 0
(N = 1) and the number M of the reference marks 61 is 1 (M =
0) will be described. In this case, it is possible, although very rare. At this time, the robot arm 20 is operated so that the reference mark 61 is located at the origin. The movement of the robot arm portion 20 is represented as a vector amount in the camera coordinate system _{Σ C (X 2, Y 2} ). When the reference mark 61 is at the origin position, the two correction marks 60a and 60b can be recognized. Therefore, if the angle of a line passing through the position of the center of gravity is calculated, it is only necessary to correct the angle. In calculating the correction amount of the position data, it is necessary to consider the movement amount (X ₂ , Y ₂ ) at the time of the re-movement.

(4) The number of marks N, M is 0 (N
= 0, M = 0), it hardly occurs,
If this occurs, most of the time is when the arm or traveling device breaks down, gets stuck, or an unintended accident occurs. Become.

Next, the correctable area and the control speed will be described. FIG. 9 is a diagram showing a region where the visual field region R can be corrected at the initial position WP1.
As shown in FIG. 9, the interval between the two correction marks 60 is the same as that of the related art.
By providing the reference mark 61 at an intermediate position of 60, if the visual field region R is the same, the robot arm 20
Can be expanded without moving (preliminary adjustment), and the efficiency can be increased.

That is, in FIG. 9, the range U1 corresponds to the range in which the automatic guided vehicle 1 reaches the work station 50, expands and contracts the robot arm 20, and moves the camera 32a to the initial position W.
When positioned at P0, two correction marks 60a ".
When only one of the reference marks 60b "is in the visual field region R, the maximum error can be recognized. The preliminary adjustment is performed to recognize the reference mark 61 and the correction mark 60, and the above-described correction control is performed. Although the range is the same regardless of whether the mark 61 "is provided or not, the control after the preliminary adjustment can be made faster as described later.

The range U2 in FIG. 9 is a correction range that does not require preliminary adjustment, and the two correction marks 60a and 60b and the reference mark 61 are in the field of view (L × L). In this case, conventionally, it is necessary to recognize the position of the center of gravity of the two correction marks 60 and 60, specify the coordinates of the intermediate position between the two correction marks, and use the coordinates as the correction value together with the inclination between the centers of gravity. In the present invention, if the reference mark 61 is recognized,
By obtaining the coordinates of the position of the reference mark 61, that is, obtaining the vector amount from the origin, and obtaining the inclination between the reference mark 61 and the correction mark 60, the correction value can be obtained. Thus, the number of instructions can be reduced and the control speed can be increased.

A range U3 in FIG. 9 indicates a range in which correction can be made faster and easier than in the prior art. That is, as in the present invention, the reference mark 61 is provided between the two correction marks 60.
Can be corrected without preliminary adjustment if one of the correction mark 60a 'and the reference mark 61' is included in the visual field region R.
The range in which the correction can be performed without operating (the range in which the preliminary adjustment is unnecessary) is defined by the correction mark 60a 'and the reference mark 61'.
Can be expanded beyond the range U2 by the distance between, and the efficiency of the correction operation can be increased.

As described above, the work position data is corrected in step SP60. Thereafter, the robot arm unit 20 performs a predetermined operation on the work 70 based on the corrected position data for the operation. In addition,
In the above embodiment, the visual field region R is a square, but is not limited to this. For example, the viewing area may be rectangular. In the above embodiment, one reference mark 61 is arranged at the center between the two correction marks 60. However, a small circular reference mark is vertically arranged at the center between the correction marks 60. It is also possible to adopt a configuration in which two are arranged. In this case, if two reference marks can be recognized, a correction value can be obtained without recognizing a correction mark. In this case, the same correction control as in the case where two conventional correction marks are recognized at once is performed. Further, the correction mark 60 may be arranged at the vertex of the regular polygon, and the reference mark may be arranged at the center of the regular polygon. Also in this case, the same correction control as in the above embodiment can be performed by recognizing one correction mark and one reference mark.

[0044]

As described above, according to the first aspect of the present invention, there is provided an unmanned transport system for transporting an article by an unmanned transport vehicle equipped with a robot arm having an image recognition device, wherein a correction mark is provided at a station for transferring the article. By providing an automatic guided vehicle system with a reference mark at the center between the correction marks, if one correction mark and the reference mark can be recognized, the robot arm will be used until all the correction marks are recognized. There is no need to move, and the time required for recognition can be reduced, and the processing time can be reduced. In addition, the correctable range that does not require preliminary adjustment can be expanded.

Further, when any one of the correction marks is arranged at the center of the recognized image, all the correction marks and the reference marks are located within the field of view of the image recognition apparatus. In the normal case, the reference mark can be recognized in addition to the correction mark in the first image recognition, and the position can be corrected in the next step. The required time can be shortened. Even if the reference mark cannot be recognized, all correction marks and reference marks can be recognized by moving any one correction mark to the center of the image.

Further, since the reference mark and the correction mark are symmetrical and two correction marks are arranged symmetrically with respect to the reference mark, all the marks are arranged symmetrically. Therefore, it is possible to prevent the mounting direction from being changed, and the operator can mount without recognizing the direction of the mark. Can be shortened.

[Brief description of the drawings]

FIG. 1 is a diagram showing a schematic configuration of an automatic guided vehicle system of the present embodiment.

FIG. 2 is a conceptual diagram showing a control configuration of the automatic guided vehicle system.

FIG. 3 is a plan view showing a relationship between the automatic guided vehicle 1 and a work station 50.

FIG. 4 is a diagram illustrating a relationship between a distance S between marks of two correction marks 60 and a length L of one side of a visual field region.

FIG. 5 is a diagram showing a captured image in a case where the position of the distal end of the robot arm unit 20 exactly matches the theoretical position of the initial position WP0.

FIG. 6 is a flowchart showing a control algorithm when an actual work is performed at each work station after teaching.

FIG. 7 shows a position WP0 of the image input unit 32. , WP1 FIG. 5 is a diagram illustrating a relationship between the two correction marks 60.

FIG. 8 is a diagram illustrating a case where one correction mark 60a exists in a visual field region R, and the other correction mark 60b and a reference mark 61 exist outside the visual field region R.

FIG. 9 is a diagram showing a region where correction is possible for a visual field region R at an initial position WP1.

[Explanation of symbols]

 DESCRIPTION OF SYMBOLS 1 Unmanned guided vehicle 20 Robot arm part 30 Image recognition part 50 Station 60 Correction mark 61 Reference mark

Claims (3)

[Claims] 1. An unmanned transport system for transporting an article by an unmanned transport vehicle equipped with a robot arm having an image recognition device, wherein a plurality of correction marks are provided at a station for transferring the articles, and a center is provided between the correction marks. An automated guided vehicle system with fiducial marks. 2. One of the correction marks is:
2. The automatic guided vehicle system according to claim 1, wherein all the correction marks and the reference marks are set so as to be within the field of view of the image recognition device when arranged at the center of the recognized image. 3. The automatic guided vehicle system according to claim 1, wherein the reference mark and the correction mark are symmetrical, and two correction marks are arranged symmetrically with respect to the reference mark.