JPS58155192A - Robot hand - Google Patents

Abstract

(57) [Summary] This bulletin contains application data before electronic filing, so abstract data is not recorded.

Description

DETAILED DESCRIPTION OF THE INVENTION The present invention relates to an improvement in a robot hand that handles an object to be grasped based on information captured by a visual device.

As a conventional robot hand H1 of this type, there is one shown in FIG. 1, for example. In the figure, 1 is an arm, 2 is a wrist, 3 is a palm, and 4.4 is a pair of fingers, each of which corresponds to a human hand.

The arm 1 is configured to move in its axial direction, that is, in the direction of arrow a, and in the vertical and horizontal directions. The wrist 2 consists of a bifurcated first wrist 2a and a cylindrical second wrist 2b which is sandwiched between the first wrist 2a and rotates in the direction of the arrow around the shaft 5. is attached to arm 1 so as to be coaxially rotatable in the direction of arrow C. The palm 3 is mounted in the second wrist 2b so as to rotate coaxially therewith in the direction of arrow d.

The above-mentioned pair of fingers 4, 4 each consist of a $1 finger 4a and a second finger 4b, and both 4a
, 4b is a joint shaft 6 that can be opened and closed in the direction of arrow e.

The second fingers 4b of both fingers 4.4 are attached to the palm 3 and are opened and closed in the direction of arrow f.

7 is an industrial camera (hereinafter referred to as ITV camera) which is a visual device for controlling the robot hand H1 by sending video information of the object to be grasped to a control mechanism (not shown), and is directed toward the fingers 4, 4. Layer the bag on Palm 3 and melt. The operation of the robot hand H1 is controlled by information from the grasped object observed by the ITV camera 7 and sent to the control a mechanism.

Next, the operation of the robot hand HI will be explained with reference to FIG. 2. This figure shows a robot hand H on a plane.
The object to be grasped 8 (cube) No. 1 is placed, and an illumination device i19 that illuminates from the outside is set above it so that the object to be grasped 8 can be easily observed with the ITV camera 7 of the robot hand H0. Show me, Qru.

In this state, the image 11 observed on the screen 10 by the ITV camera 7 is as shown in Figure $3.

The following steps are taken from this state until the robot hand H1 is operated to grasp the object to be grasped 8.

First, the position of the center of gravity of the image 11 of the grasped object 8 on the screen 10F' is determined, and the robot hand H, is controlled so that the center of gravity is located at the center of the screen 10, as shown in FIG. Moving. The center of gravity here means a two-dimensional center of gravity, and is indicated by an x in Figure 1.

Next, even if the first center of gravity is at the center of the screen 10, as shown in FIG. .4, when the ITV is not directly facing the gripping direction q.
The camera 1, that is, the palm 3, further rotates in the direction of arrow d, and the image 11 on the screen 1O comes to the position shown in FIG. In other words.

The object to be grasped 8 shown in the image 11 comes to a position on the screen 10 where it can be grasped from the direction of the arrow q.

In this way, when the palm 3 reaches the gripping position on the screen 10, the palm 3 is adjusted in advance so that it moves in parallel a certain distance toward the gripping object 8, descends, and comes to a position where it can grip the object. Is it a robot hand? I descends as it is and grips the object 8 to be gripped.

In this way, the robot hand H, is equipped with the visual device 7 in the palm 3, so firstly.

The hand H8 can move and process information on the image 11 of the grasped object 8 observed to control itself. , has the advantage that it can be replaced by the mechanical movement of the robot hand H1.

However, the conventional robot hand H0 as described above
However, since the illumination system [9 is located far from the robot node H8, there are disadvantages such as the following. .

That is, the first drawback is that the accurate grasping of the grasped object 8 is often difficult due to the shadow caused by the robot hand H8 or the shadow caused by the grasped object 8 itself, which makes it difficult for the robot hand H8 to effectively grasp the grasped object 8. The second drawback is that information regarding the surface shape, such as the markings 13 on the object 12 shown in FIG.
If two parts are made of the same color, it cannot be easily extracted and understood, and to that extent, the robot hand Shizuku cannot be used effectively.

The present invention has been made by focusing on such conventional drawbacks, and includes a light projector, etc. that projects marks such as spot lights onto the object to be grasped, on the fingers of a robot hand equipped with the conventional visual device as described above. A light source is installed, and from the position of the light source, the projection direction of the mark, and the image position of the projected mark, the six-sided shape of the gripped object, which could not be identified conventionally, or the Fi loincloth rice shade. To provide a robot robot which can accurately grip an object by inputting information on the tooth portion which could not be identified due to the robot's operation.

The present invention will now be explained based on the embodiment shown in FIG. Note that the same parts as in FIG. 1 are denoted by the same reference numerals, and explanations thereof will be omitted.

First, to explain the configuration, the robot node H of this invention
1 is a conventional robot equipped with an ITv camera 7 as a visual device on the palm 3, as shown in FIG. The feature is that a light projector 14, which is a light source, is attached to one of the fingers 4.4, as shown in FIG.

Here, the light projector 14 is a robot hand H! This is a light source that projects a mark such as a spot light that serves as a recognition target for the ITV camera 7 onto an object to be gripped.

As shown in FIG. 7, the projector 14 is mounted so as to project a mark in the direction of the first finger 4a of one finger 4. Therefore, the first finger child 4a
By moving the mark, the projection direction of the mark can be changed.

In addition, IT is calculated to determine the surface shape of the object to be grasped.
The distance from the V camera 7 to the mark is calculated from the following factors. That is, the projector 14
V(', which can be calculated from the relative position of the mark with respect to the i'rv camera 7, the projection direction of the mark with respect to the optical axis of the ITV camera 7, and the image position of the mark observed by the ITV camera 7).

Then, the calculated distance information to the valley mark is finally input to the control structure of the robot hand H2 (not shown),
The configuration is such that it is processed as drive information for the hand H2.

Note that the calculation of the above-mentioned distance is based on the principle of triangular side quantity, and will be explained with reference to FIGS. 8 to 10 as follows.

FIGS. 8 and 9 show a state in which spot light from the light projector 14 of the finger 4 of the robot hand H2 is observed by the Tv camera 1. In the figure, Yo is I
'r Distance from the tip of the V camera 7 to the floodlight 14,
Xo is the distance from the optical axis of the Tv camera 7 to the projector 14. This Xo.

The relative position of the projector 14 with respect to the ITV camera 7 is determined by Yo.

π/4−α indicates the angle between the optical axis of the ITv camera 7 and the optical axis of the spot light, and the projection direction of the thumbnail mark.

FIG. 10 shows the grasped object 1 in FIGS. 8 and 9.
9 shows the screen 16 when the mark projected on the screen 5 is observed by the ITV camera 7, and the ψ mark 17 is the 8th
The mark 18 indicates the appearance in the case of FIG. 9. X is mark 1 located at the center of screen 16
1 to the mark 18, that is, the image position of the mark.

Under the above conditions, mark 17.1 from ITV camera 7
Find the distance to 8. In the case of Figure 8, Y, = Yo + X. tan (g/j-a') In the case of Figure 9, if the magnification of the ITV camera 7 is K, then Y trout = Y6 + (X6 Kit) tai (K
/ 4 (!), respectively.

Next, the operation based on the above configuration will be explained.

For convenience of explanation, 1. A gripping object 12 having a depression 13 on the surface shown in FIG. 6 at the position of the gripping object 8 shown in FIG.
Consider the case where there is.

In this case, as in the past, the IT of the robot hand H2
The gripping position of the gripping object 12 cannot be determined only from the contour information obtained by the V-camera 7. This is because it is necessary to determine the gripping direction and the gripping position by identifying the orientation of the gripping object 12, that is, which side of the video screen the recess 13 is on.

Therefore, this invention's robot hand H! As in the conventional case, first, as shown in FIGS. 3 to 5, the position of the grasped object 12 determined only by the outline is determined.

However, in the image state shown in FIG. 5, although the outline and position of the grasped object 12 can be determined, only the external illumination provided by the illumination device 8 will cause the image 20 in FIG. It is not possible to identify which side of the screen 10 1 and 3 are located, that is, the A position or the B position.

Therefore, in the next step, the recess 13 is A.

This means checking at one point at which position B is located. In other words, the fingers 4, 4 including the rotation of the palm 3
The distance from the tip of the ITV camera 7 to the mark is calculated from the position of the light projector 14 and the image position of the mark observed by the ITV camera 7 while controlling the direction in which the mark is projected by the light projector 14 by the movement of .

Information regarding the surface shape of the object to be grasped 12 obtained at the end of the distance information to each of the projected marks is then manually inputted to the control mechanism, and is used as drive information for the robot hand H2.

As described above, the present invention is constructed by attaching a light source such as a projector that projects a mark such as a spot light to the finger of a robot hand equipped with a visual device in the palm, which is difficult to identify using only conventional external layers. The unevenness of the object to be grasped,
It is now possible to grasp information about the surface shape such as depressions, or the shadow parts of the object to be grasped, which could not be observed with conventional lighting alone. Even if the object to be gripped has a depression or the like, it can be handled accurately and easily.

[Brief explanation of the drawing]

Figure 1 is a perspective view of a conventional robot hand, and Figure 2 is a perspective view of a conventional robot hand.
Fig. 1 is a perspective view, Fig. 3 shows how the robot hand is used.
5 @ Figures 4 and 5 are views showing images of the object to be grasped observed with the visual device of the conventional robot hand, Figure 6 is a perspective view of the object to be grasped with a ridge, and Figure 7 is the image of the object to be grasped according to the present invention. FIGS. 8 to 10 are diagrams for explaining the principle of distance measurement in the robot hand of this invention, and FIG. 11 is a perspective view of a grasped object observed by the visual device of the robot hand of this invention. A diagram showing the video of Ruru. HI... Conventional robot hand H2...
......Robot hand according to this invention 3...
・・・・・・－・−Palm 4・・・・・・・・・Finger 7・・・・・・
...ITV camera 12, which is a visual device...
... Grasped object 14 ...... Representative of the floodlight that is the light source Satoru Kuzuno (and one other person) Fig. 1 540 Fig. 5 Fig. 9 Fig. 10 Fig. 71

Claims (1)

[Claims] A finger of a robot hand equipped with a visual device in its palm is provided with a light source that projects a mark that is a recognition target of the visual device onto an object to be grasped, and a relative position of the light source with respect to the visual device; A robot hand characterized in that the shape of the object to be gripped is grasped based on the projection direction of the mark with respect to the optical axis of the visual device and the image position of the mark observed from the visual device.