JPH04372335A - Workpiece direction discriminating method and device therefor - Google Patents

Abstract

PURPOSE:To enable the correct discrimination of workpiece direction dispensing with the preparation of a user program workpiece by workpiece by discriminating the direction of the workpiece on the basis of a vector having the center-of-gravity of the workpiece and that of a circumscribed rectangle as both end points. CONSTITUTION:The characteristic value is outputted from a characteristic value output part 11 for plural sample workpieces, and the center-of-gravity position of a workpiece and that of a circumscribed rectangle are selected by a first selecting part 12. The distance between the center-of-gravity of the workpiece and that of the circumscribed rectangle is computed by a first distance computing part 13. The distance average value and variance are then computed by an average distance-variance computing part 15, and whether the average distance is larger or smaller than the preset reference distance is discriminated by a distance discriminating part 16. In the case of discriminating the vector length to be longer than the specified length, the direction of the workpiece is discriminated on the basis of this vector, and in the case of being shorter than the specified length, another characteristic value is selected in place of the center-of-gravity of the circumscribed rectangle to obtain the direction vector.

Description

[Detailed description of the invention]

0001

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a method and apparatus for identifying the direction of a workpiece, and more specifically, it is particularly suitable for identifying the direction of workpieces sequentially fed out by a parts supply device such as a bowl feeder. The present invention relates to a direction identification method and device.

0002

2. Description of the Related Art Conventionally, in systems for grasping workpieces such as parts using industrial robots, a visual system consisting of an image detection device and an image processing device has been incorporated, focusing on the advantage of eliminating the need for positioning the workpiece. A system has been proposed that controls an industrial robot based on the position, posture, etc. of a workpiece obtained by a vision system ("Parts supply system using vision").
Hitachi, Ltd. Production Technology Laboratory Michio Takahashi 1932
January 29, 3rd year Japan Society for Precision Engineering Automatic Assembly Special Committee No. 6
(See 2nd research presentation).

[0003] In this system, the outline of a binary image area is approximated by a polygon, and data of the vertex sequence is used to approximate the outline of the binary image area.
More than 10 graphical features such as center of gravity, area, and perimeter are calculated, and these large numbers of graphical features are combined by a user program to determine the type, orientation, position, etc. of the workpiece. Therefore, if you create a user program that instructs how to combine graphical features, you can determine the type of workpiece, orientation, gripping position, etc. without any special work. Can be done.

0004

[Problem to be Solved by the Invention] However, as mentioned above, it is essential to create a user program, and moreover, it is necessary to create a user program according to the type of work, so there are many In the current situation where low-volume production of various types is becoming commonplace, not only does the amount of work required to create a user program significantly increase, but it is also necessary to change the user program in response to changes in the shape of the workpiece due to model changes, etc. is required and the user
This has the disadvantage that the amount of work required to change the program also increases.

[0005] The above describes the determination of the type of workpiece, the orientation,
Although a case has been described in which all the determination of the gripping position and the like are performed, the same disadvantages arise even when focusing only on identifying the orientation of the workpiece.

[0006]

[Object of the Invention] This invention has been made in view of the above problems, and provides a new workpiece direction that eliminates the need to create a user program for each workpiece and can accurately identify the direction of the workpiece. The object of the present invention is to provide an identification method and device.

[0007]

[Means for Solving the Problems] To achieve the above object, the method for identifying the direction of a workpiece according to claim 1 provides a method for identifying the direction of a workpiece, in which a first
obtain a vector, identify the direction of the workpiece based on the first vector when the length of the first vector is greater than or equal to a predetermined length, and identify the center of gravity of the workpiece when the length of the first vector is shorter than the predetermined length. This is a method of obtaining a second vector whose end points are the position and other feature points of the workpiece, and identifying the direction of the workpiece based on the second vector.

[0008] The method for identifying the direction of a workpiece according to a second aspect of the present invention provides a first
This method identifies the direction of the workpiece based on the sign of the inner product of the vector or the second vector and the principal axis of inertia vector. The workpiece direction identification device according to claim 3 includes: a workpiece gravity center calculation means for calculating the gravity center position of the work; a circumscribed rectangle gravity center calculation means for calculating the gravity center position of the circumscribed rectangle; and a workpiece gravity center calculation means for calculating the gravity center position of the circumscribed rectangle; a first vector calculating means for calculating a first vector having both end points at the center of gravity; a first vector determining means for determining whether the length of the calculated first vector is greater than or equal to a predetermined length; first direction identification means for identifying the direction of the workpiece based on the first vector in response to determining that the length of one vector is greater than or equal to a predetermined length; feature point calculating means for calculating the position of a feature point different from the center of gravity of the circumscribed rectangle in response to determining that the circumscribed rectangle is short;
a second vector calculating means for calculating a second vector having both end points the calculated center of gravity position of the workpiece and the positions of other feature points of the workpiece; and second direction identification means for identifying the direction of the workpiece based on the direction of the workpiece.

[0009] In the workpiece direction identification device according to claim 4, the first
The direction identification means and the second direction identification means each identify the direction of the workpiece based on the sign of the inner product of the corresponding vector and the principal axis of inertia vector.

0010

[Operation] In the workpiece direction identification method according to claim 1, regardless of the type of the workpiece, the first method that determines the center of gravity of the workpiece and the center of gravity of the circumscribed rectangle as both end points is based on the position of the center of gravity of the workpiece and the center of gravity of the circumscribed rectangle. Get a vector. And the first
The direction of the workpiece is identified based on the first vector when the length of the vector is greater than or equal to a predetermined length. Conversely, if the length of the first vector is shorter than the predetermined length, the accuracy of identifying the direction based on the first vector will be significantly reduced. A second vector having the points as both endpoints is obtained, and the direction of the workpiece is identified based on the second vector.

[0011] Therefore, the user
There is no need to create a program, the amount of work for creating and changing programs can be greatly reduced, and moreover, highly accurate direction identification can be achieved for all workpieces. Also,
Since the direction of the workpiece can be identified based on the first vector unless the contour of the workpiece and the circumscribed rectangle are not significantly different, the load for identifying the direction of the workpiece can be greatly reduced.

According to the workpiece direction identification method of claim 2, by calculating the inner product and determining whether the sign is positive or negative, it is possible to identify whether or not the principal axis of inertia angle can be directly adopted as the direction of the workpiece. The processing load and control load of industrial robots, etc. can be significantly reduced. Furthermore, by determining whether the value of the inner product is close to 0, it is possible to identify whether the image is horizontally long or not.

[0013] In the workpiece direction identification device according to claim 3, the workpiece gravity center calculation means calculates the gravity center position of the workpiece, the circumscribed rectangle gravity center calculation means calculates the gravity center position of the circumscribed rectangle, and the calculated gravity center position of the workpiece is calculated. Based on the position of the center of gravity of the circumscribed rectangle, the first vector calculation means calculates a first vector having both end points the center of gravity of the workpiece and the center of gravity of the circumscribed rectangle. Then, the first vector determining means determines whether the length of the calculated first vector is equal to or greater than a predetermined length.

If the length of the first vector is determined by the first vector determining means to be greater than or equal to the predetermined length, sufficient direction identification accuracy can be obtained. , identifying the orientation of the workpiece based on the first vector. On the other hand, if the length of the first vector is determined to be shorter than the predetermined length by the first vector determining means, sufficient direction identification accuracy cannot be obtained as it is, so the feature point calculating means The position of the feature point that is different from the center of gravity of the circumscribed rectangle is calculated, and based on the calculated center of gravity position of the workpiece and the positions of other feature points of the workpiece, the second vector calculation means moves the center of gravity of the workpiece and the corresponding feature point at both ends. A second vector is calculated as a point. Then, the second direction identifying means identifies the direction of the workpiece based on the second vector.

As is clear from the above explanation, the direction of the workpiece can be identified based on the first vector unless the contour of the workpiece and the circumscribed rectangle are very different. The load can be greatly reduced. With the workpiece orientation identification device according to claim 4, it is possible to determine whether or not the principal axis of inertia angle can be directly adopted as the workpiece direction by simply calculating the inner product and determining whether the sign is positive or negative, thereby reducing the processing load. , the control load on industrial robots can be further reduced. Furthermore, by determining whether the value of the inner product is close to 0, it is possible to identify whether the image is horizontally long or not.

[0016]

Embodiments Hereinafter, embodiments will be described in detail with reference to the accompanying drawings. FIG. 7 is a diagram schematically showing the configuration of a component supply device to which the workpiece direction identification method of the present invention is applied.
A part of the parts conveyance track 1a of the circulating bowl feeder 1 is constituted by a translucent plate 1b, and lighting devices 2 and 3 are arranged above and below the translucent plate 1b. . A half mirror 3a is placed between the translucent plate 1b and the lower lighting device 3, and a CCD is placed at a predetermined position where it can receive the light reflected by the half mirror 3a.
An image detection section 4 consisting of a camera and the like is arranged. Furthermore, the industrial robot 5 is equipped with a suction head 5a for suctioning the workpiece. Also, there is an image processing section 4a which performs binarization processing etc. by inputting the image signal taken in from the image detection section 4, and a robot which receives the processing result signal from the image processing section 4a and gives operation commands to the industrial robot 5. - It has a controller 4b.

FIGS. 1 and 2 are flowcharts showing an embodiment of the method for identifying the direction of a workpiece according to the present invention.
2 shows a teaching process based on a sample work, and FIG. 2 shows a direction identification process based on a work to be identified. First, the flowchart in FIG. 1 will be explained. In step SP1, wait until the sample work is set within the field of view of the image detection unit 4, and in step SP2, capture the image of the sample work, and perform binarization processing by the image processing unit 4a to obtain a sample work image. , In step SP3, the image processing unit 4a calculates the position of the feature value of the sample work based on the binarized image. Then, in step SP4, it is determined whether or not a predetermined number of sample works have been processed, and if the predetermined number has not been reached, the process in step SP1 is performed again. Conversely, if it is determined that the predetermined number has been reached, the center of gravity of the sample work and the center of gravity of the circumscribed rectangle are selected from among the feature values of the sample work in step SP5, and the center of gravity of the sample work and the center of gravity of the circumscribed rectangle are selected in step SP6. Calculate the average value and variance of the distance between and the center of gravity of the circumscribed rectangle,
In step SP7, it is determined whether the average value of the distances is shorter than a predetermined distance. If it is determined that the average value of the distances is shorter than the predetermined distance, then in step SP8, a feature value that provides a distance longer than the predetermined distance is selected instead of the center of gravity of the circumscribed rectangle. Examples of the feature points selected here include the center of gravity of the largest hole, the center of gravity of the smallest hole, the largest radius circle passing point, the smallest radius circle passing point, the largest angle point, the smallest angle point, and the like.

If it is determined in step SP7 that the average value of the distances is not shorter than the predetermined distance, or if the process in step SP8 is performed, step SP9
In step SP10, the principal axis of inertia direction vector and the direction identification vector having the center of gravity of the sample workpiece as its starting point and the selected feature point as its end point are obtained.In step SP10, the inner product of both vectors is calculated, and in step SP11, the absolute value of the value of the inner product is calculated. and a predetermined value. Then, only when it is determined that the absolute value of the inner product value is smaller than a predetermined value, that is, when it is determined that the inner product value is close to 0, the corresponding sample work is identified as being horizontally long in step SP12. Then, the series of processing ends.

Next, the flowchart of FIG. 2 will be explained. In step SP1, an image of the workpiece to be identified is captured, in step SP2, the position of the center of gravity of the workpiece and the position of the feature point (center of gravity of a circumscribed rectangle or other feature point) selected in the flowchart of FIG. Determine whether the workpiece is horizontally long or not. If it is determined that the object is horizontally long, the principal axis of inertia angle is increased by 90 degrees in step SP4. However, the principal axis of inertia angle may be reduced by 90°.

Thereafter, in step SP5, a direction identification vector is calculated with the center of gravity of the workpiece to be identified as the starting point and the selected feature point as the end point, and in step SP6, the inner product of the direction identification vector and the principal axis of inertia direction vector is calculated, In step SP7, the sign of the inner product is determined. If it is determined in step SP7 that the inner product is negative, the principal axis of inertia angle is set to 180 in step SP8.
decrease by °. That is, the direction of the principal axis of inertia is reversed.

If it is determined in step SP7 that the inner product is positive, or if the process in step SP8 is performed, the principal axis of inertia angle is output as the direction angle of the workpiece to be identified in step SP9, and a series of processes is performed. finish. This will be explained in more detail with reference to FIGS. 3 to 5. Figure 3 (A) shows the case where the shape of the sample work is an isosceles triangle, and Figure 3 (B) shows the case where the shape of the sample work is a trapezoid. The center of gravity of each rectangle is indicated by C. As is clear from these figures, in the case of an isosceles triangle, the distance between the centers of gravity G and C is large, and in the case of a trapezoid, the distance between the centers of gravity G and C is small. Therefore, when comparing the two, in the case of an isosceles triangle, sufficient direction identification can be achieved by adopting a direction identification vector that starts from the center of gravity G and ends at the center of gravity, but in the case of a trapezoid (especially when the upper side If the difference from the bottom side is small and the lengths of the side sides are similar), the length of the direction identification vector starting from the center of gravity G and ending at the center of gravity C becomes significantly shorter, and the effects of calculation errors etc. As the size increases, the accuracy of direction identification decreases.

FIG. 4 is a diagram illustrating a direction identification vector based on feature points that can be adopted in place of the center of gravity C of a circumscribed rectangle when the length of the direction identification vector is short. The center of gravity G is the starting point. In FIG. 4, V1 is a direction identification vector whose end point is the center of gravity of the largest hole, V2 is a direction identification vector whose end point is the center of gravity of the smallest hole, V3 is a direction identification vector whose end point is the maximum radius circle passing point, V4 is a direction identification vector whose end point is the minimum radius circle passage point, V5 is a direction identification vector whose end point is the maximum angle point, and V6 is a direction identification vector whose end point is the minimum angle point.

FIG. 5 is a diagram for explaining the relationship between the workpiece direction identification vector and the principal axis of inertia vector. If the inner product of the workpiece direction identification vector and the principal axis of inertia direction vector is close to 0, the sample workpiece is horizontally long. Therefore, it is only necessary to increase or decrease the principal axis of inertia angle by 90 degrees, and the value of the inner product of both vectors after the principal axis of inertia angle is increased or decreased by 90 degrees can be made sufficiently large. Furthermore, if the sign of the inner product of both vectors is positive as shown in Fig. 5(A), the principal axis of inertia angle and the direction angle of the workpiece match, and as shown in Fig. 5(B), the sign of the inner product of both vectors is positive. When the sign of the inner product is negative, the supplementary angle of the principal axis of inertia angle and the direction angle of the workpiece match.

Therefore, after teaching based on the flowchart of FIG. 1, the directional angle of the workpiece can be easily obtained by performing the processing based on the flowchart of FIG.

[0025]

[Embodiment 2] FIG. 6 is a block diagram showing an embodiment of the workpiece orientation identification device of the present invention, which has a feature of obtaining and outputting feature values of the workpiece based on the image of the workpiece taken in from the image detection unit 4. a value output section 11; a first selection section 12 that selects the center of gravity of the workpiece and the center of gravity of the circumscribed rectangle from among the feature values output from the feature value output section 11; A first distance calculation unit 13 that calculates the distance to the center of gravity, a feature value output unit 11, a first selection unit 12, and a first distance for a sample work of n samples in response to the teaching mode being instructed. an iterative control unit 14 that repeatedly operates the calculation unit 13; an average distance/variance calculation unit 15 that calculates the average value and variance of the distances calculated by the first distance calculation unit 13 for n sample works; The distance determining unit 16 determines whether the average distance is greater than or equal to a preset reference distance, and the distance determining unit 16 operates based on a determination signal from the distance determining unit 16 indicating that the average distance is shorter. Based on the second selection section 17 which selects a feature value whose distance from the center of gravity of the workpiece is longer than the reference distance from among the other feature values output from the feature value output section 11, and the discrimination signal from the discrimination section 16. select the center of gravity position of the workpiece selected by the first selection unit 12 and the center of gravity position of the circumscribed rectangle or the feature value selected by the second selection unit 17, and set the center of gravity of the workpiece as the starting point and select any selected feature. A first vector calculation unit 18 that calculates a workpiece direction identification vector with the value as the end point, and a first inner product calculation that calculates the inner product of the workpiece direction identification vector calculated by the first vector calculation unit 18 and the principal axis of inertia direction vector. Part 1
9, an inner product value determination unit 20 that determines whether the absolute value of the inner product calculated by the first inner product calculation unit 19 is larger or smaller than a preset predetermined value, and a determination indicating that the absolute value of the inner product is smaller. The landscape instruction data output unit 21 outputs instruction data indicating that the corresponding sample work is landscape based on the results, and the feature value output unit 11 responds to the instruction of the work orientation identification mode. Among the output feature values, the third selection unit 22 selects the same feature value as the center of gravity position of the workpiece and the feature value selected by the second selection unit 17, and the horizontal orientation data output unit 21 indicates that the work is horizontally long. a first principal axis of inertia angle correction unit 23 that increases or decreases the principal axis of inertia angle by 90° only when instruction data indicating that A second vector calculation unit 24 that calculates and outputs a workpiece direction identification vector with the characteristic value as the end point.
and a second inner product calculation unit 25 that calculates the inner product of the workpiece direction identification vector calculated by the second vector calculation unit 24 and the principal axis of inertia direction vector, and the sign of the inner product calculated by the second inner product calculation unit 25. and a work direction angle output section 27 that obtains the principal axis of inertia angle or an angle obtained by reducing the principal axis of inertia angle by 180 degrees based on the sign discrimination result by the sign discrimination section 26 and outputs it as a work direction angle. It has

The operation of the workpiece direction identification device having the above structure is as follows. When the teaching mode is instructed, the feature value output unit 11 outputs a feature value for each of the n sample workpieces, and the first selection unit 12 selects the center of gravity position of the workpiece and the center of gravity position of the circumscribed rectangle. is selected, and the first distance calculation unit 13 calculates the distance between the center of gravity of the workpiece and the center of gravity of the circumscribed rectangle. Then, the average distance/variance calculating section 15 calculates the average value and variance of the distance, and the distance determining section 16 determines the magnitude of the average distance and a preset reference distance. That is, it is determined whether the distance is such that sufficient direction identification accuracy can be obtained. As a result of this determination, if it is determined that the average distance is shorter, the second selection section 17 selects one of the other feature values output from the feature value output section 11 instead of the center of gravity of the circumscribed rectangle. Select a feature value whose distance from the center of gravity is longer than the reference distance. Thereafter, the first vector calculation section 18 calculates the center of gravity position of the workpiece and the center of gravity position of the circumscribed rectangle selected by the first selection section 12 based on the discrimination signal from the discrimination section 16 or the center of gravity position of the circumscribed rectangle selected by the second selection section 17. A feature value is selected, and a workpiece direction identification vector is calculated with the center of gravity of the workpiece as the starting point and one of the selected feature values as the ending point. Then, the first inner product calculation unit 19 calculates the inner product of the workpiece direction identification vector and the principal axis of inertia direction vector, and the inner product value determination unit 20 determines whether the absolute value of the inner product is larger or smaller than a preset predetermined value. Then, only when it is determined that the absolute value of the inner product is smaller, the landscape instruction data output unit 21 outputs instruction data indicating that the corresponding sample work is horizontally long.

After the sample workpiece has been taught as described above, it is only necessary to instruct the workpiece direction identification mode. First, the third selection section 22 selects the feature value output section 1.
1, selects the same feature value as the center of gravity position of the workpiece and the feature value selected by the second selection unit 17, and instructs the horizontal orientation data output unit 21 that the workpiece is horizontal. The first principal axis of inertia angle correction unit 23 adjusts the principal axis of inertia angle to 9 only when the instruction data is output.
Increase or decrease by 0°. Next, the second vector calculation unit 24 calculates and outputs a workpiece direction identification vector with the workpiece gravity center as the starting point and the feature value selected by the third selection unit 22 as the end point, and the second inner product calculation unit 25 calculates and outputs the workpiece direction identification vector. The inner product of the workpiece direction identification vector and the principal axis of inertia direction vector is calculated, and the sign of the inner product is determined by the sign determining unit 26. Then, the workpiece direction angle output unit 27 reduces the principal axis of inertia angle by 180° only when the inner product is negative, and when the inner product is positive, the principal axis of inertia angle is directly adopted and output as the workpiece direction angle.

As is clear from the above explanation, for most workpieces, inner product calculation, sign discrimination, etc. are performed based on the workpiece direction identification vector and the principal axis of inertia direction vector obtained based on the workpiece's center of gravity and the center of gravity of the circumscribed rectangle. By simply performing this process, the direction of the workpiece can be identified easily and with high accuracy. Also, only when the distance between the center of gravity of the workpiece and the center of gravity of the circumscribed rectangle is short, instead of the center of gravity of the circumscribed rectangle, another feature value whose distance to the center of gravity of the workpiece is sufficiently large can be selected to obtain the workpiece direction identification vector. The direction of the workpiece can be identified with high accuracy.

It should be noted that the present invention is not limited to the above-mentioned embodiments; for example, the process associated with calculating the inner product may be omitted and the direction of the workpiece may be identified solely based on the workpiece direction identification vector. In addition, various design changes can be made without changing the gist of the invention.

[0030]

As described above, the invention of claim 1 eliminates the need to create a user program for each type of work, and can greatly reduce the amount of work for creating and changing programs. Since it is possible to achieve highly accurate orientation identification of the workpiece and to identify the direction of the workpiece based on the first vector, except in cases where the outline of the workpiece and the circumscribed rectangle are not very different, it is possible to identify the orientation of the workpiece. This has the unique effect of greatly reducing the load on the machine.

According to the second aspect of the invention, by calculating the inner product and determining whether the sign is positive or negative, it is possible to identify whether or not the principal axis of inertia angle can be directly adopted as the direction of the workpiece, thereby reducing processing load and industrial robots. It has the unique effect of being able to significantly reduce the control load. In the invention of claim 3, since the direction of the workpiece can be identified based on the first vector except when the outline of the workpiece and the circumscribed rectangle are not much different, the load for identifying the direction of the workpiece can be greatly reduced. It has the unique effect of being able to reduce the

The invention of claim 4 can determine whether or not the principal axis of inertia angle can be directly adopted as the direction of the work simply by calculating the inner product and determining whether the sign is positive or negative. This has the unique effect of further reducing the control load on the robot.

[Brief explanation of the drawing]

FIG. 1 is a flowchart showing teaching processing based on a sample work in an embodiment of the work direction identification method of the present invention.

FIG. 2 is a flowchart showing direction identification processing based on a workpiece to be processed in an embodiment of the workpiece direction identification method of the present invention.

FIG. 3 is a diagram showing the relationship between the center of gravity of a sample work and the center of gravity of a circumscribed rectangle when the shape of the sample work is an isosceles triangle and a trapezoid.

FIG. 4 is a diagram illustrating a direction identification vector based on feature points that can be adopted instead of the center of gravity of a circumscribed rectangle when the length of the direction identification vector is short.

FIG. 5 is a diagram illustrating the relationship between the direction identification vector of the workpiece and the principal axis of inertia vector.

FIG. 6 is a block diagram showing an embodiment of the workpiece direction identification device of the present invention.

FIG. 7 is a diagram schematically showing the configuration of a component supply device to which the workpiece orientation identification method of the present invention is applied.

[Explanation of symbols]

11 Feature value output section 12 First selection section
15 Average distance/variance calculation section 16 Distance discrimination section 17 Second selection section
18 First vector calculation unit

Claims (4)

[Claims] Claim 1: Obtain a first vector having one of the center of gravity of the workpiece and the center of gravity of the circumscribed rectangle as a starting point and the other as an end point based on the center of gravity position of the workpiece and the center of gravity of the circumscribed rectangle, and calculate the length of the first vector. is longer than a predetermined length, the first
The direction of the workpiece is identified based on the vector, and if the length of the first vector is shorter than a predetermined length, a second vector is obtained whose end points are the center of gravity position of the workpiece and other feature points of the workpiece, and the second vector is A method for identifying the direction of a workpiece, characterized by identifying the direction of the workpiece based on a vector. 2. The method for identifying the direction of a workpiece according to claim 1, wherein the direction of the workpiece is identified based on the sign of the inner product of the first vector or the second vector and the principal axis of inertia vector. 3. Workpiece gravity center calculation means (11, 12) for calculating the gravity center position of the workpiece, circumscribed rectangle gravity center calculation means (11, 12) for calculating the gravity center position of the circumscribed rectangle, and the calculated gravity center position of the workpiece. and a first vector calculation means (18) for calculating a first vector whose end points are the centroid position of the circumscribed rectangle; 1 vector determining means (15, 16), and a first direction identifying means for identifying the direction of the workpiece based on the first vector in response to determining that the length of the first vector is greater than or equal to a predetermined length. (22
, 23, 24, 25, 26, 27) and a feature point for calculating the position of the feature point that is different from the centroid position of the circumscribed rectangle in response to determining that the length of the first vector is shorter than a predetermined length. calculation means (11, 17), and a second vector calculation means (18) for calculating a second vector whose end points are the calculated center of gravity position of the workpiece and the positions of other feature points of the workpiece.
) and second direction identification means (22, 23, 24, 25, 26, 27) for identifying the direction of the workpiece based on the second vector in response to the second vector being calculated. Characteristic workpiece orientation identification device. Claim 4: First direction identification means (22, 23, 24
, 25, 26, 27) and second direction identification means (22,
4. The workpiece orientation identification device according to claim 3, wherein the workpiece directions are identified based on the sign of the inner product of the corresponding vector and the principal axis of inertia vector.