Fanuc Corporation

Taro Arimatsu

Atsushi Watanabe

Kazunori Ban

Ryuichi Hara

Kazuhiko Akiyama

Masaru Oda

Fumikazu Warashina

Ichiro Kanno

Tatsuo Karakama

Keisuke Watanabe

Yoshinori Ochiishi

Hiroaki Kubota

An object picking device, which is inexpensive and capable of speedily and accurately picking one object at a time from a random pile state. A target detecting part of an image processing part processes an image captured by a camera and detects objects. A target selecting part selects an object among the detected objects based on a certain rule. A view line direction calculating part calculates the direction of a view line extending to the selected object. A target position estimating part estimates the position including the height of the selected object based on size information of the object in the image. Then, a grip correction calculating part calculates an amount of correction of the movement of a robot so as to grip the object by using the robot.

Object picking device

A hand as an end effector. The hand includes a base, a hook element associated with the base and capable of hooking and lifting an object, a holding element associated with the base and cooperating with the hook element to hold the object therebetween, and a drive section causing a relative movement between the hook element and the holding element. For example, the hook element is arranged movably in a direction toward and away from the holding element on the base, and the drive section drives the hook element. Alternatively, the holding element is arranged movably in a direction toward and away from the hook element on the base, and the drive section drives the holding element. A handling robot includes an arm and the above-described hand attached to the arm.

Hand and handling robot

A workpiece W serving as an object for detection is fixed in place. A camerais mounted to the end of a robot RB. The camera is turned about an axis, which passes through the center position of the workpiece W and is perpendicular to the optical axis of the camera, to take an image of the workpiece W at a plurality of positions in different directions. A teaching model is generated on the basis of each produced image data. The relative position and posture of the workpiece to the camerais also stored in association with the teaching model. Thus, it is possible to easily generate the teaching model of the workpiece regardless of three-dimensional variations in posture.

Teaching model generating device

A robot system having an image processing function capable of detecting position and/or posture of individual workpieces randomly arranged in a stack to determine posture, or posture and position of a robot operation suitable for the detected position and/or posture of the workpiece. Reference models are created from two-dimensional images of a reference workpiece captured in a plurality of directions by a first visual sensor and stored. Also, the relative positions/postures of the first visual sensor with respect to the workpiece at the respective image capturing, and relative position/posture of a second visual sensor to be situated with respect to the workpiece are stored. Matching processing between an image of a stack of workpieces captured by the camera and the reference models are performed and an image of a workpiece matched with one reference model is selected. A three-dimensional position/posture of the workpiece is determined from the image of the selected workpiece, the selected reference model and position/posture information associated with the reference model. The position/posture of the second visual sensor to be situated for measurement is determined based on the determined position/posture of the workpiece and the stored relative position/posture of the second visual sensor, and precise position/posture of the workpiece is measured by the second visual sensor at the determined position/posture of the second visual sensor. A picking operation for picking out a respective workpiece from a randomly arranged stack can be performed by a robot based on the measuring results of the second visual sensor.

Robot system having image processing function

A robot control unit 42 and an image processing control unit (control unit of the image processing apparatus) 43 are incorporated into a robot controller 40. A camera CM is connected to the image processing control unit 43. A main body 1 of a robot is connected to the robot control unit 42 through an amplifier 41. A portable robot teaching pendant 80 connected to the robot control unit 42 is provided with a monitor display, and functions also as a teaching pendant of the image processing apparatus. Therefore, by using the teaching pendant 80, manipulation of image processing, and issuing of an instruction to a program for processing an image can be performed. Furthermore, an image obtained by a camera CM, and information relevant to the manipulation of the image processing apparatus such as an operation menu, etc. can be displayed on the monitor display. Therefore, an operator can efficiently perform all operations relevant to the robot, the camera, processing an image, etc. while watching a monitor screen.

Image processing apparatus for robot

An image processing apparatus capable of detecting position and posture of individual workpieces randomly arranged in a stack and having identical shapes. Reference models are created from two-dimensional images of a reference workpeace captured in a plurality of directions by a camera and stored. Also, the relative positions/postures of the workpiece with respect to the camera at the respective image capturing are stored. An image of a stack of workpieces is captured by the camera to obtain a two-dimensional image and the position/posture of the camera at the image capturing is stored. An image of a workpiece matched with one reference model is selected by matching processing of the reference model with the captured image. A three-dimensional position/posture of the workpiece with respect to the camera is obtained from the image of the selected workpiece, the selected reference model and position/posture information associated with the reference model. A picking operation for picking out a respective workpiece from a randomly arranged stack can be performed by a robot based on the position/posture of the workpiece.

Image processing apparatus

There is provided a measurement device being capable of obtaining an image being free from a distortion even if the position of a measurement object varies in image pickup and being capable of performing a precise measurement according to the image. Camera calibration is executed by using a dot pattern or the like, and parameters of a camera model are stored (S ). The image of a reference object is fetched (S ), a corrected reference image being free from a lens distortion and a distortion caused by image pickup in a diagonal direction is formed on the basis of the equation of the camera model (S ), and parameters for detecting the measurement object are set in accordance with the image (S ). In a system operation, the image of the measurement object, the position of which varies, is acquired (S ), and the corrected reference image being free from a distortion as in S is formed (S ). The measurement object is detected by using the parameters for detecting the measurement object (S , S ), necessary data for an application is formed and output (S ). In failure to perform detection, an appropriate message is output (S , S ).

Measurement apparatus

A tool mounted on the wrist end of a robot is placed in a predetermined photographing position, and the front and side faces of the tool are photographed by cameras to obtain three-dimensional shape data of the tool. The extent of deformation of the tool is detected by comparing the obtained shape data and reference shape data, and a programmed movement path for the robot is corrected depending on the extent of deformation.

Apparatus for correcting movement path of a robot and a method therefor

A taught data setting method in a visual sensor system, capable of automatically and rapidly setting optimum taught data. An image of a single circular sample is subject to image processing to determine formal setting values of first to sixth parameters (P1-P6), respectively corresponding to circle diameter, circumference, area and deformation degree, and types of a smoother filter and an emphasizing filter. Many sample images are then subject to image processing which utilizes the formal setting values, to thereby determine set center positions of the respective images. An optimum value of the first parameter enough to minimize the sum of the squares of the errors between the set center positions and image center positions is automatically determined (100), the image center positions being detected while the first parameter is changed, with the second to fourth parameters kept fixed at their formal setting values. Similarly, optimum values of the second to fourth parameters are determined (100-103). Image processing is further performed while changing filter types and keeping the first to fourth parameters fixed at their optimum values, thereby determining optimum values of the fifth and sixth parameters (108,109).

Growing community of inventors

Yamanashi, Japan

Taro Arimatsu