JPH0771978A - One's own position recognition system - Google Patents

Abstract

PURPOSE:To determine one's own position at a high speed even during movement by computing the current position in accordance with the results of computations by first and second target-position computing means. CONSTITUTION:A first drive means 8 controls the directions of a first CCD camera 6 and a first laser pointer 7 in accordance with directions from a first target-position recognition means 1. A first target-position computing means 1 computes a first target position from the directions of the camera 6 and the laser pointer 7 at the time when the first target is caught. Further, a second drive means 11 controls the directions of a second CCD camera 9 and a second laser pointer 10 in accordance with directions from a second target-position recognition means 2. A second target-position computing means 4 computes a second target position from the directions of the camera 9 and the least pointer 10 when the second target is caught. A current-position computing means 5 computes the current position in accordance with the results of computations by the means 3 and 4.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a self-position recognizing device, and more particularly to a self-position recognizing device suitable for positioning an unmanned traveling vehicle or an autonomous mobile robot.

[0002]

2. Description of the Related Art In a conventional self-position recognizing device, as shown in FIG.
A CD camera is used to track each target object, and the moving direction and the moving distance are calculated from the relative positions of the target object before and after the movement.

[0003]

However, in the above-mentioned conventional example, since the target object is specified by the shape recognition, as shown in FIG. 6, the target object becomes invisible or the shape changes depending on the direction. However, there is an inconvenience that it may not be recognized.

Further, since the shape recognition requires a large amount of data, a long processing time is required, and there is a problem that the position cannot be recognized during movement.

[0005]

SUMMARY OF THE INVENTION It is an object of the present invention to provide a self-position recognizing device which can improve the inconveniences of the conventional example and can measure the position of the user at high speed especially during movement.

[0006]

Therefore, in the present invention, a first CCD camera that captures a first target, a first laser pointer that emits a laser spot to the first target, and a first CCD camera. Target position recognizing means for scanning a characteristic portion using a cross-shaped window on the image from FIG. 1, a first CCD camera from the first target position recognizing means, and a first target position recognizing means.
Target position calculating means for calculating the position of the first target from the angle information of the laser pointer, the second CCD camera for capturing the second target, and the second laser beam for emitting the laser spot to the second target. Laser pointer, second target position recognition means for scanning a characteristic portion using a cross-shaped window on the image from the second CCD camera, and second CCD camera from the second target position recognition means. The second target position calculating means for calculating the second target position from the angle information of the second laser pointer, and the present result based on the calculation results of the first target position calculating means and the second target position calculating means It is configured to include a current position calculation means for calculating the position. This aims to achieve the above-mentioned object.

[0007]

The first target position recognizing means 1 scans the image of the first CCD camera by using a cross-shaped window, and scans the image as an object having a characteristic brightness difference in the window as the first target. recognize.

Then, the direction of the first laser pointer is changed so that the bright spot due to the reflected light of the first laser pointer is located at the center of the window.

The first target position calculating means 3 detects angle information of the first CCD camera 6 and the first laser pointer when the first target sent from the first target position recognizing means 1 is captured. From, the first target position is calculated.

The second target position recognizing means 2 uses the second CC
The cross-shaped window is used for the image of the D camera 9 to scan the inside of the image, and an object having a characteristic brightness difference in the window is recognized as the second target.

Then, the direction of the second laser pointer is changed so that the bright spot due to the reflected light of the second laser pointer is located at the center of the window.

The second target position calculating means 4 detects the angle information of the second CCD camera 9 and the second laser pointer when the second target sent from the second target position recognizing means 2 is captured. From this, the second target position is calculated.

The current position calculating means 5 calculates the current position based on the first target position obtained by the first target position calculating means 3 and the second target position obtained by the second target position calculating means 4. To do.

[0014]

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS An embodiment of the present invention will be described below with reference to FIGS.

The embodiment of FIG. 1 has a first goal of capturing a first goal.
CCD camera 6, a first laser pointer 7 that emits a laser spot to a first target, and a first target that scans a characteristic portion using a cross-shaped window on the image from the first CCD camera 6. The position recognition means 1, the first drive means 8 for controlling the directions of the first CCD camera 6 and the first laser pointer 7 according to the instruction from the first target position recognition means 1, and the first target are captured. It is provided with a first CCD camera 6 and a first target position calculation means 3 for calculating the position of the first target from the direction of the first laser pointer 7 when it is opened.

Further, in this embodiment, a second CCD camera 9 for capturing a second target, a second laser pointer 10 for emitting a laser spot to the second target, and an image from the second CCD camera 9 are provided. Second target position recognizing means 2 for scanning a characteristic portion using a cross-shaped window, and a second CCD camera 9 according to an instruction from the second target position recognizing means 2.
And the second drive means 11 for controlling the direction of the second laser pointer 10, the second CCD camera 9 when the second target is captured, and the second target from the direction of the second laser pointer 10. A second target position calculation means 4 for calculating the position, and a current position calculation means 5 for calculating the current position based on the calculation results of the first target position calculation means 3 and the second target position calculation means 4. Are equipped with.

Here, as shown in FIG. 2A, the CCD camera and the laser pointer are installed at predetermined intervals on the same supporting column which stands upright on the vehicle body. The CCD camera and laser pointer are shown in FIG.
As shown in (b), the rotary table installed on the column allows simultaneous rotation. 2C shows an example of the optical system of the first laser pointer.

Next, the operation of this embodiment will be described.

(A). Determination of target position:

First, two targets are found in a stopped state before traveling.

[0021]. The first target landmark is found with the first CCD camera 6.

Here, the first target position recognizing means 1 is
While controlling the first driving means 8 to rotate the first CCD camera 6 horizontally with respect to the ground, a cross-shaped window is set for the image from the first CCD camera 6,
As shown in FIG. 3A, an object having a large difference in brightness between the center of the window and the arm is targeted.

Next, the first target position recognition means 1
By controlling the driving means 8 of the first laser pointer 7 to rotate the mirror of the first laser pointer 7 perpendicularly to the ground, and the bright spot by the laser reflected light shown in FIG.
In the image from, it is located in the center of the cross-shaped window.

Then, the first target position recognizing means 1 detects the angle of the first CCD camera 6 with respect to the ground, the horizontal rotation angle of the first CCD camera 6 with respect to the traveling direction, and the first laser pointer 7 with respect to the ground. To the first target position calculation means 3.

[0025]. Horizontal distance L to the first landmark
Ask for 10.

Here, the distance between the first CCD camera 6 and the first laser pointer 7 is d1, the first C with respect to the ground.
Assuming that the angle of the CD camera 6 is β10 and the mirror rotation angle of the first laser pointer 7 with respect to the ground surface is α10, the first target position calculating means 3 is, as shown in FIG. 4B, a principle of triangulation. Then, the horizontal distance to the first landmark is calculated by the following equation (1).

L10 = d1 / (1 / cotα10-1 / cotβ10) (1)

.. The coordinates of the first landmark are determined from the horizontal rotation angle γ10 with respect to the traveling direction of the first CCD camera 6 and the horizontal distance L10 to the first landmark.

If the current position of the vehicle is the origin, the traveling direction is the X direction, and the direction orthogonal to the traveling direction is the Y direction, then the first
The target position calculation means 3 calculates the coordinates (X10, Y10) of the first landmark by the following equation (2).

X10 = L10.cosγ10

Y10 = L10 · sin γ10 (2)

.. The second CCD camera 9 finds the target second landmark.

Here, the second target position recognition means 2 is
The second driving means 11 is controlled to control the second CCD camera 9
While rotating horizontally with respect to the ground, a cross-shaped window is set for the image from the second CCD camera 9. As shown in FIG. 3A, the center of the window and the arm portion are set. The target is one with a large difference in brightness from.

Next, the second target position recognition means 2
The driving means 11 for controlling the second laser pointer 10
The mirror is rotated perpendicularly to the ground, and the bright spot due to the laser reflected light is positioned at the center of the cross-shaped window in the image from the second CCD camera 9 as shown in FIG. 4 (a). To do so.

Then, the second target position recognizing means 2 detects the angle of the second CCD camera 9 with respect to the ground, the horizontal rotation angle of the second CCD camera 9 with respect to the traveling direction, and the second laser pointer 10 with respect to the ground. The mirror rotation angle of
The target position calculation means 4 is notified.

.. Horizontal distance L to the second landmark
Ask for 20.

Here, the distance between the second CCD camera 9 and the second laser pointer 10 is d2, the angle of the second CCD camera 9 with respect to the ground is β20, and the mirror rotation angle of the second laser pointer 10 with respect to the ground. Is set to α20, the second target position calculation means 4 is, as shown in FIG.
Applying the principle of triangulation, the horizontal distance to the second landmark is calculated by the following equation (3).

L20 = d2 / (1 / cotα20-1 / cotβ20) (3)

.. The coordinates of the second landmark are determined from the horizontal rotation angle γ20 with respect to the traveling direction of the second CCD camera 9 and the horizontal distance L20 to the second landmark.

If the current position of the vehicle is the origin, the traveling direction is the X direction, and the direction orthogonal to the traveling direction is the Y direction, then the second
The target position calculating means 4 calculates the coordinates (X20, Y20) of the second landmark by the following equation (4).

X20 = L20.cosγ20

Y20 = L20.sinγ20 (4)

(B). Behavior on the move:

.. The first CCD camera 6 tracks the first landmark.

Here, the first target position recognizing means 1 is
As shown in FIG. 3B, the window is moved so that the cross-shaped window data in the captured image has the same characteristics as when the first landmark is set.

When the characteristic of the first landmark cannot be found even if the window is moved, the first target position recognizing means 1 controls the first driving means 8 so that the first CCD camera 6 is activated. Rotate and look for features of the first landmark.

When the first target position recognizing means 1 finds the characteristics of the first landmark, it controls the first driving means 8 so that the first landmark is located at the center of the image, The first CCD camera 6 is moved.

Then, the first target position recognizing means 1 makes the angle of the first CCD camera 6 with respect to the ground, the horizontal rotation angle of the first CCD camera 6 with respect to the traveling direction, and the first laser pointer 7 with respect to the ground. To the first target position calculation means 3.

Even if the first CCD camera is moved,
If the feature of the first landmark cannot be found, the position of the first landmark is returned to search for the feature of the first landmark.

.. The second CCD camera 9 tracks the second landmark.

Here, the second target position recognizing means 2 is
As shown in FIG. 3B, the window is moved so that the cross-shaped window data in the captured image has the same characteristics as when the second landmark is set.

When the feature of the second landmark cannot be found even if the window is moved, the second target position recognizing means 2 controls the second driving means 11 to make the second CCD.
Rotate the camera 9 to look for features of the second landmark.

When the second target position recognizing means 2 finds the feature of the second landmark, it controls the second driving means 11 so that the second landmark is located at the center of the image, The second CCD camera 9 is moved.

Then, the second target position recognizing means 2 detects the angle of the second CCD camera 9 with respect to the ground, the horizontal rotation angle of the second CCD camera 9 with respect to the traveling direction, and the second laser pointer 10 with respect to the ground. The mirror rotation angle of
The target position calculation means 4 is notified.

If the feature of the second landmark cannot be found even if the second CCD camera 9 is moved, the position of the second landmark is returned to search for the feature of the second landmark.

(C). Operation at the correction position:

When the moving direction is changed, the current position is temporarily stopped and the current position is confirmed, and the position is corrected in some cases.

.. Horizontal distance L to the first landmark
Ask for 11.

Here, the angle of the first CCD camera 6 with respect to the ground is β11, and the first laser pointer 7 with respect to the ground.
Assuming that the mirror rotation angle of α11 is α11, the first target position calculation means 3 applies the principle of triangulation to the first landmark up to the first landmark as shown in FIG. 4 (b). Find the horizontal distance of.

L11 = d1 / (1 / cotα11-1 / cotβ11) (5)

.. The coordinates of the first landmark are obtained from the horizontal rotation angle γ11 of the first CCD camera 6 with respect to the traveling direction and the horizontal distance L11 to the first landmark.

The first target position calculating means 3 calculates the coordinates (X11, Y11) of the first landmark by the following equation (6).

X11 = L11.cosγ11

Y11 = L11 · sinγ11 (6)

.. Horizontal distance L to the second landmark
Ask for 21.

Here, the angle of the second CCD camera 9 with respect to the ground is β21, and the second laser pointer 1 with respect to the ground.
Assuming that the mirror rotation angle of 0 is α21, the second target position calculating means 4 applies the principle of triangulation to the second landmark according to the following equation (7) as shown in FIG. 4 (b). Find the horizontal distance to.

L21 = d2 / (1 / cotα21-1 / cotβ21) (7)

.. The coordinates of the second landmark are calculated from the horizontal rotation angle γ21 with respect to the traveling direction of the second CCD camera 9 and the horizontal distance L21 to the second landmark.

The second target position calculating means 4 calculates the coordinates (X21, Y21) of the second landmark by the following equation (8).

X21 = L21.cosγ21

Y21 = L21 · sinγ21 (8)

.. The current position calculation means 5 determines the coordinates of the first landmark (X10, Y10) and the coordinates of the second landmark (X20, Y20) when the target position is determined, and the first landmark at the correction position. The current position (X, Y) of the vehicle is calculated using the coordinates (X11, Y11) and the coordinates (X21, Y21) of the second landmark.

X = X11-X10

Y = Y11−Y10 (9)

X = X21-X20

Y = Y21−Y20 (10)

Here, the coordinates obtained by the equation (9) and (1
0) The average of the coordinates obtained from the equation
Y).

Then, the vehicle is moved in a predetermined direction with the present position (X, Y) as a new initial position.

After that, the vehicle can be guided to a predetermined position by repeatedly performing the above processes (B) and (C).

As described above, since it is not necessary to recognize the shape of the target object, high speed processing is possible. Moreover, since two targets are used, accurate positioning is possible.

Further, the distance to the target can be obtained by using a laser range finder, a stereo camera or the like.

[0082]

Since the present invention is constructed and functions as described above, according to this, it is possible to calculate the positions of two objects without shape recognition, which makes it possible to perform high speed and high movement even when moving. It is possible to provide an unprecedented excellent self-position recognizing device that enables self-position recognition with high accuracy.

[Brief description of drawings]

FIG. 1 is a configuration diagram showing an embodiment of the present invention.

2A and 2B are explanatory views for explaining the positional relationship between the CCD camera and the laser pointer in FIG. 1, FIG. 2A shows an overall perspective view, and FIG. 2B shows one CCD camera and the laser pointer. 2C shows the positional relationship with FIG.
It is explanatory drawing which shows the optical system of the laser pointer in (b).

FIG. 3 is an explanatory diagram for explaining the operation of the target position recognition means in FIG.

FIG. 4 is an explanatory diagram for explaining the operation of the target position calculation means in FIG.

FIG. 5 is a configuration diagram showing a conventional example.

6 is an explanatory diagram for explaining an operation of the conventional example of FIG.

[Explanation of symbols]

 1 1st target position recognition means 2 2nd target position recognition means 3 1st target position calculation means 4 2nd target position calculation means 5 present position calculation means 6 1st CCD camera 7 1st laser pointer 8 First driving means 9 Second CCD camera 10 Second laser pointer 11 Second driving means

Claims (2)

[Claims] 1. A first CCD camera that captures a first target, a first laser pointer that emits a laser spot to the first target, and a predetermined window for an image from the first CCD camera. A first target for calculating the position of the first target from first target position recognizing means for scanning a part and angle information of the first CCD camera and the first laser pointer from the first target position recognizing means. A second CCD camera that has a position calculation means and captures a second target, a second laser pointer that emits a laser spot to the second target, and a predetermined window for an image from the second CCD camera. The second target position recognition means for scanning the characteristic portion using the second target position recognition means, and the second target position recognition means based on the angle information of the second CCD camera and the second laser pointer from the second target position recognition means. A second target position calculation means for calculating the present position, and a current position calculation means for calculating the current position based on the calculation results of the first target position calculation means and the second target position calculation means. A self-position recognition device. 2. The self-position recognition device according to claim 1, wherein the window is a cross-shaped window.