JPH08386B2 - Robot controller - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION [Industrial field of application] The present invention enables to automatically move between two points by avoiding obstacles on the way only by designating a start point and an end point of an operation of an arm or the like. The present invention relates to a robot control device that operates.

[Prior Art] Conventionally, robots have been actively used in place of people for simple work and dangerous work. Although its structure and function differ depending on the required purpose and work content, it is desired to provide a robot capable of more advanced operations and complicated operations, and a robot capable of AI functions such as learning.

For example, JP-A-61-74399 is one of the former and JP-A-62-72004 is one of the latter.

[Problems to be Solved by the Invention] However, in a conventional robot, when a demonstration is performed, or when only a specific object is extracted from various objects in a selection process, the operation of an arm or the like is not performed. Specify start point and end point (or specify goods)
However, if you want to automatically move between the two points, it may not be possible to operate.

That is, when there are a plurality of obstacles between two designated points within the movement range, conventionally, the method of recognizing the features of the object has been adopted. There is a problem that the arm cannot be advanced.

It is an object of the present invention to provide a robot controller capable of automatically generating a locus between two designated points even if a plurality of similar objects exist in the operation range.

[Means for Solving the Problems] In order to achieve the above object, the present invention provides an image pickup means for outputting image information obtained by picking up an image of the operation range of a tool as an original image, and a binary image from the image information obtained by the image pickup means. , A labeling processing means for labeling each of a plurality of objects included in the original image, and a barycentric position, orientation and occupancy of the object based on a processing result by the means. An object occupation area calculation means for calculating an area and an object movement locus calculation means for calculating the movement locus of the tool based on the processing result by the means are provided.

[Operation] According to the above-mentioned means, the image within the operation range of the tool imaged by the imaging means is digitized, and a plurality of objects existing within the operation range are labeled.
The occupied area is calculated from this result, and the movement trajectory of the tool is further calculated. Therefore, even when there are a plurality of objects within the operation range, it is possible to grasp and move the target object.

Embodiments Embodiments of the present invention will be described below with reference to the drawings.

FIG. 1 is a block diagram showing an embodiment of the present invention. Here, the case where a joint type is used as the robot body is illustrated.

The embodiment shown here is a command control unit 1, which creates a program for operating the robot body 3 and inputs data.
Corresponding to the robot controller 2 and arms, hands, and fingers that are built in the robot body 3 and that perform control for driving the robot body 3 according to the instructions of the command controller 1 while exchanging data with the command controller 1. And a plurality of motors (not shown) as driving sources for the respective movable parts, which are configured by a robot main body 3 controlled by the robot controller 2.

The command control unit 1 is an operation panel 101 for inputting various data, set values, programs, etc., a television (TV) camera 102 having an operation range of the movable tip of the robot body 3, and an output signal of the operation panel 101. Input interface 103 that captures the data at a predetermined timing
U) 104, read-only memory (ROM) 105 that stores programs, random access memory (RAM) 106 that temporarily stores processing results and data, specific processing algorithms necessary for robot control, constants, etc. A table memory 107 to be stored, an image processing control section 108 for performing image processing and control for displaying an image on a monitor 109 and an image processing described later for moving the robot, display processing contents by the CPU 104 or a robot operating state. Monitor 109,
Output interface 1 for controlling the printer 111 described later
10, the printer control unit 2 sends data to the printer 111, which prints out processing results under the control of the output interface 110, and the robot control unit 2.
And a transmission input / output unit 112 that receives the operation state information of the robot body 3.

Further, each member to which data signals, control signals, address signals, etc. are connected to the CPU 104 and the CPU 104 are mutually connected by a bus 113. Furthermore, between the TV camera 102 and the bus 113, the TV camera 102
A binarization processing unit 114 for binarizing, ie, digitizing, the image output of is connected.

Although the TV camera 102 is described as being included in the command control unit 1 for the sake of explanation, it is actually installed near the robot body 3 so that its field of view is within the operation range of the movable portion of the robot body 3. Is set to.

Further, the robot control unit 2 receives a signal from the transmission input / output unit 112, and transmits the information regarding the operation state of the robot body 3 and the like, the transmission input / output unit 201, the transmission input / output unit
A control unit 202 that outputs control information for driving a motor based on a signal from 201 and manages information sent to the command control unit 1 side, and each motor of the robot main body 3 according to a command from the control unit 202 And a motor drive unit 203 that drives a drive source such as.

The control unit 202 is configured to include a CPU and its peripheral circuits,
It operates as a slave to the CPU 104 of the command control unit 1 which is a master.

Next, regarding the operation of the embodiment having the above configuration,
This will be described with reference to the flowchart in the figure. In addition, in the following, the step of the flow is described as S.

First, an execution start command is keyed in on the operation panel 101, the apparatus is started and initialized, and the condition for operating the robot is specified. The input setting conditions are stored in the table memory 107. Also, the operation range is photographed by the TV camera 102 and the image output is captured. The CPU 104 displays on the monitor 109 the contents that can be operated by the robot and waits. In addition, necessary information is transmitted to the robot controller 2 via the transmission input / output unit 112.

In the robot control unit 2, the control unit 202 executes processing according to the flow shown in FIG. 2 and exchanges information with the command control unit 1.

First, when the power is turned on, the robot body 3 is initialized (S21), and data regarding an object within the operation range is initialized (S22). Then, the command control unit 1 is requested via the transmission input / output unit 201 to request a robot motion selection of what motion should be performed (S23).

In response to the robot operation selection request from the robot controller 2, the command controller 1 prompts the operator to select from the menu (S32) displayed on the monitor 109, and the operation selection result selected by the operation panel 101. Is returned to the robot controller 2. Receiving this, the control unit 202 issues an image processing request to the command control unit 1, and in the command control unit 1 receiving this, the image processing control unit 108 processes the image information captured from the television camera 102, and The image processing content (S33) is returned to the robot controller 2.

Further, the control unit 202 subsequently issues an image processing result request (S25). For this image processing, a plurality of reciprocal exchanges are performed between the command control unit 1 and the robot control unit 2, and the command control unit 1 repeats the image processing result 34.

When the result of the image processing is obtained, the control unit 202 plans the action to be taken by the robot (that is, how to move between the start point and the end point) in response to the given action selection request and image content. Yes (S26). Based on the contents of this plan, the control unit 202 sends a control command to the motor drive unit 203 to start driving the arm of the robot body 3.

When one operation ends, the command control unit 1 is inquired whether or not there is a next process to be executed (S28). In response to this inquiry, the operator selects the same operation as the previous operation or another different operation while looking at the menu screen (S35) on the monitor 109. This selection content is transmitted to the transmission input / output unit 201 of the robot control unit 2 via the transmission input / output unit 112 and the transmission path, and is taken in by the control unit 202. Control unit 20
2 judges the received contents (S29), and if it is the re-execution of the same operation, the operation after step 27 is repeatedly executed again.

In addition, in response to a request to execute a different operation, it is determined whether or not it corresponds to the same type of operation (S30),
If they are the same, the processing after step 24 is repeatedly executed. On the other hand, if it is not the same type of operation, it is determined whether or not the operation request made in step 23 is requested again to the command control unit 1 (S31), and if an operation reselection instruction is determined, step 23 The subsequent processing is executed again. If no operation reselection is displayed, all processing is terminated.

Next, the details of the image processing will be described with reference to FIGS. 3 (A), (B), and (C). In the above description, the start point and the end point are instructed to move the arm or the like, but here, one feature of a plurality of objects within the motion range is given as data, and the specified object is specified. A case will be described in which the object is grasped by performing image processing and searching.

First, as shown in FIG. 3 (A), the television camera 102 captures an image of a subject within the operating range of the tool. This is the original image (S41). There are a plurality of objects within this operating range (= viewing range of the camera), and
One is the object to be gripped by the robot.

Then, basic image processing is executed (S42), and a binary image is created for this (S43). Labeling processing is performed on the binarized image thus obtained (S44). In this labeling process, image information in which shading is applied to the input binarized image is made, and each object is numbered and labeled.

Next, as shown in FIG. 3 (B), a binarized image is created for the grayscale image subjected to the labeling process (S4
6) Then, the object position, the object direction, and the occupied area of each object are calculated (S47). This processing is achieved by calculating each of the following based on the labeling processing information.

Object position: Calculate the barycentric coordinates of the object Orientation of the object: The angle between the principal axis of inertia and the X axis Occupied area of the object: Maximum and minimum coordinates in the X and Y directions. Convert to a certain world coordinate system. X = a * x + c (mm) Y = -d * y + f (mm) with respect to the origin x, y of the image coordinate system (where a is the measurement length (mm) of one pixel in the x direction, and c is the world coordinate system) Within the image system, x is the origin, d is the measurement length (mm) of one pixel in the y direction, and f is the image system y origin position (mm) in the world coordinate system) Next, the geometric attributes of the object are calculated, that is, the object The feature amount of is calculated (S48). A scalar feature defined in the analysis constant table is issued for one object, and its geometric attribute value is calculated. The value thus obtained is stored in the geometric attribute work table.

Next, as shown in FIG. 3C, the number of the object is determined (S49). This decision is made based on the latest value in the geometric attribute table. Then, it is determined whether or not the object number is the object to be grasped (S50). If the numbers match, the gripping position and gripping direction of the object are stored as data in the object gripping method table and the object number correspondence record (S51).

Then, it is checked whether or not the object number and the moving target object number match (S52). In addition, step 5
In 0, if the object number and the gripping target object number do not match, the process proceeds to step 52. When it is determined in step 52 that the object number and the movement target object number match, the data of the object outline (length, width, height, etc.) is input and stored in the object gripping method table (S53). .

Further, the data is stored in the object recognition result table (S5
Four). After that, the process returns to step 45 of FIG. 3 (B) to execute the subsequent processes. When it is determined in step 45 that the processing has been completed for all the objects, the robot control unit 2 sequentially executes the processing of step 26 and subsequent steps shown in FIG. 2 based on the object recognition result, and determines the shortest distance to the designated object. To approach and hold it.

EFFECTS OF THE INVENTION As described above, according to the present invention, the image pickup means for outputting the image information obtained by picking up the operation range of the tool as an original image, and the binarization processing for making the binary image from the image information by the image pickup means. Means, a labeling processing means for labeling each of a plurality of objects included in the original image, and a barycentric position of the object based on a processing result by the means,
An object occupation area calculation means for calculating the orientation and the occupation area,
Since the object movement locus calculation means for calculating the movement locus of the tool based on the processing result by the means is provided, the movement locus is automatically created even when a plurality of objects exist within the operation range. , It is possible to grasp and move a target object.

[Brief description of drawings]

FIG. 1 is a block diagram showing an embodiment of the present invention, FIG. 2 is a flow chart for explaining the outline of the processing of the command controller 1 and the robot controller 2, and FIGS. 3 (A), (B), (C). 10 is a flowchart showing details of image processing in the present invention. In the figure. 1: Command control unit, 2: Robot control unit 3: Robot main body, 4: External storage device 101: Control panel, 102: TV camera 104: CPU, 107: Table memory 108: Image processing control unit 109: Monitor, 112 : Transmission input / output unit 114: Binarization processing unit, 201: Transmission input / output unit 202: Control unit, 203: Motor drive unit

Claims (1)

[Claims] 1. An image pickup means for outputting image information obtained by picking up an image of an operation range of a tool as an original image, a binarization processing means for producing a binarized image from the image information by the image pickup means, and the original image in the original image. Labeling processing means for labeling each of a plurality of included objects, object occupancy area calculation means for calculating the barycentric position, orientation, and occupied area of the object based on the processing result by the means, and processing by the means A robot controller, comprising: an object movement locus calculation means for calculating the movement locus of the tool based on the result.