KR102205040B1 - Intelligent robot controling method of turning machining using machine vision - Google Patents

Abstract

Disclosed is an intelligent robot control method using a machine vision, comprising: a step where a robot control unit controls a robot arm unit to move to a preset reference position; a step where the robot control unit controls a vision camera unit to photograph a subject to be extracted; a step of detecting a cap-shaped disc in a vision image photographed by the vision camera unit; a step where the cap-shaped disc controls and moves the robot arm unit in accordance with an offset value between the detected vision position and the preset reference position; a step where the robot control unit controls an extraction tool unit to extract the cap-shaped disc, and controls the robot arm unit to transfer the extracted cap-shaped disc; and a step where, when the cap-shaped disc is completely transferred, the robot control unit controls a slip sheet adsorption unit to adsorb and remove a slip sheet. The present invention is able to effectively extract, transfer, and remove the subject to be extracted and the slip sheet from a pallet by using a robot.

Description

Intelligent robot control method using machine vision {INTELLIGENT ROBOT CONTROLING METHOD OF TURNING MACHINING USING MACHINE VISION}

The present invention (Disclosure) relates to an intelligent robot control method using machine vision, and specifically, a vision position detected through a vision camera provided in the robot, and an extraction position according to a preset reference position and an offset distance. Is determined, and after taking out and transferring the take-out object through the robot's position control according to the take-out position, by repeatedly performing the process of adsorbing and removing the slip sheet, effectively taking out and transferring the take-out object and the slip sheet from the pallet using the robot. And an intelligent robot control method using removable machine vision.

Here, background technology related to the present invention is provided, and these do not necessarily mean known technology (This section provides background information related to the present disclosure which is not necessarily prior art).

Wheels used in automobiles and trains and discs used in brakes are made of cast iron.

Here, the wheel, wheel and disk require continuous inspection due to uneven wear, corrosion, etc. occurring during the manufacturing stage or during use, and the work of post-processing the outer surface to maintain a certain quality is required.

In addition, the wheels of the train may be periodically detached according to wear and then precisely processed, or the wheels may be loaded and unloaded into the processing machine through position control using a robot capable of transferring the wheels, which are heavy objects to perform diagnosis.

These wheel wheels and disks (hereinafter referred to as'disks') are difficult to transport or work for a long time due to their high weight, so the transport and processing of the disks are performed in consideration of productivity improvement and worker safety. It is mainly carried out by an automated system.

In the automated system, the disk may be loaded on a pallet provided with a preset standard, and the disk loaded on the pallet may be transferred to a processing device by a robot for processing.

However, the roughed disk can be loaded onto the pallet in a stacked type in which a layer of cap-type disks is stacked again after applying a certain pattern on the slip sheets to put them into the turning process (or polishing process). In order to take out the disc and the slip sheet, the process of removing the slip sheet must be repeatedly performed after the disk is taken out according to the pattern, but the position control is difficult, and there is a problem that a separate robot for taking out the disk and taking out the slip sheet is required.

In addition, conventionally, since materials are handled in pallet units, there may be cases where it is not possible to apply standard patterns such as 3*3, 4*4, and various patterns such as 11 and 17 can be used per layer. The same non-standardized pattern has a problem of acting as a variable in the automated process because the pattern is distorted during transportation using a freight truck, forklift, etc., and the material cannot be taken out from an accurate location.

In addition, conventionally, when a material is input by a forklift, a tolerance must occur, and when a problem occurs in system operation, the operator randomly assigned an offset by software to operate the robot, but the material is outside the allowable range for each layer on which the material is stacked. In this case, the same problem occurs repeatedly, and there is a problem in continuous production using an automated process and its productivity.

1. Korean Registered Patent Publication No. 10-0457312 (registered on November 5, 2004)

The present invention (Disclosure) determines the take-out position according to the vision position detected through a vision camera provided in the robot, a preset reference position and an offset distance, and controls the position of the robot according to the take-out position. Provides an intelligent robot control method using machine vision that can effectively take out, transfer and remove the take-out object and the slip sheet from the pallet by using a robot by repeatedly performing the process of adsorbing and removing the slip sheet after taking out and transporting the take-out object. Is for work purposes.

Further, the present invention (Disclosure) obtains an offset value of a robot corresponding to a specific position by subtracting a reference position corresponding to a specific position from a vision position detected through a vision image at a specific position, and calculates the reference position and the offset value. Providing an intelligent robot control method using machine vision that can accurately eject the disk through precise position control of the robot according to the detection position of the disk to be ejected by determining the ejection position reflecting the vision offset value at a specific position by adding up Is for work purposes.

Here, a summary of the present invention is provided, and this section provides a general summary of the disclosure and is not a comprehensive disclosure of its full scope or all of its features).

In order to solve the above problems, the intelligent robot control method using machine vision according to any one of the various aspects describing the present invention, in which the robot control unit controls the robot arm to move to a preset reference position. step; Photographing an object to be taken out by controlling the vision camera unit in the robot control unit; Detecting a cap-type disk from the vision image captured through the vision camera unit; Controlling and moving the robot arm according to an offset value between the detected vision position of the cap-shaped disk and the preset reference position; Controlling the take-out tool part by the robot control part to take out the cap-type disk, and then controlling the robot arm part to transfer the taken out cap-type disk; And when the transfer of the cap-type disk is completed, the robot control unit controls the separation sheet to be sucked and removed.

In an intelligent robot control method using machine vision according to an aspect of the invention, after the step of photographing the object to be taken out, a detection error that fails to detect the cap-type disk in the vision image is counted to become a detection error threshold. In this case, the method may further include controlling the robot arm to change the search location.

In the intelligent robot control method using machine vision according to an aspect of the invention, the step of providing an abnormal state alarm when a threshold value of the number of times of change is reached while changing the search location; may further include.

In the intelligent robot control method using machine vision according to an aspect of the invention, the step of controlling and moving the robot arm unit includes: when the cap-type disk is detected in the vision image, the reference position and the vision position are When the offset distance is less than the offset distance threshold, the robot arm may be controlled to move to the vision position.

In the intelligent robot control method using machine vision according to an aspect of the invention, when the cap-type disk is not detected for each pattern in the step of detecting the cap-type disk, the exposure time correction and the illumination intensity adjustment of the vision camera are performed. Step; may further include.

In the intelligent robot control method using machine vision according to an aspect of the invention, the step of controlling and moving the robot arm unit includes the reference when the moving distance of the robot arm unit converges to the offset distance within an error range. The offset value corresponding to the position may be obtained.

In the intelligent robot control method using machine vision according to an aspect of the invention, the step of controlling and moving the robot arm includes determining an extraction position reflecting the vision offset value by summing the reference position and the offset value. The robot arm can be moved.

In the intelligent robot control method using machine vision according to an aspect of the invention, the step of controlling and moving the robot arm part may change the coordinate system of the robot arm part to the take-out coordinate system before controlling the robot arm part to move. have.

In the intelligent robot control method using machine vision according to an aspect of the present invention, the step of transferring the cap-type disk includes removing the cap-type disk by actuating a tong tool of the take-out tool part, and then controlling the robot arm part. It can be transferred to the target position.

In the intelligent robot control method using machine vision according to an aspect of the invention, the step of adsorbing and removing the slip sheets comprises: after sequentially taking out and transferring the cap-shaped disk, the suction tool of the slip sticking portion is lowered to lower the slip sheet. Can be removed by adsorption.

According to the present invention, an extraction position is determined according to a vision position detected through a vision camera provided in a robot, a preset reference position and an offset distance, and an object to be taken out through the position control of the robot according to the extraction position. After taking out and transporting the slippers, by repeatedly performing the process of adsorbing and removing the slip sheets, the object to be taken out and the slip sheets from the pallet can be effectively taken out, transferred, and removed using a robot.

And, according to the present invention, by subtracting a reference position corresponding to a specific position from a vision position detected through a vision image at a specific position, an offset value of the robot corresponding to a specific position is obtained, and the reference position and the offset value are summed. By determining the ejection position reflecting the vision offset value at a specific position, the corresponding disk can be accurately ejected through the precise position control of the robot according to the detection position of the disk to be ejected.

In addition, according to the present invention, when performing position control using machine vision, materials can always be taken out from a certain position, so that a separate device such as a disk loading device and an alignment device is not required, as well as various applications other than a magnetic tool. The tool can be designed and reflected, and material can be taken out and transferred from a certain position at all times through the position control of the robot using machine vision, providing a stable operating environment, and offsets deviated from the machine vision for each position to be taken out. The value can be calculated automatically to maximize productivity.

On the other hand, according to the present invention, it is possible to control the position of the robot using machine vision, so that a task combined with other processes can be performed, thereby improving economic efficiency and efficiency in performing an automated process.

1 is a flowchart showing an intelligent robot control process using machine vision according to an embodiment of the present invention,
2 is a block diagram of a control device that performs an intelligent robot control process using machine vision according to an embodiment of the present invention.
3 is a diagram illustrating an object to be taken out loaded on a pallet according to an embodiment of the present invention,
4 is a diagram for explaining a reference position, a vision position, and an offset distance according to an embodiment of the present invention,
5 and 6 are diagrams illustrating a control device for performing an intelligent robot control process using machine vision according to an embodiment of the present invention,
7 and 8 are views illustrating a take-out tool part and a sticking part according to an embodiment of the present invention,
9 is a flow chart for explaining in detail an intelligent robot control process using machine vision according to an embodiment of the present invention.

Hereinafter, an embodiment of implementing a method for controlling an intelligent robot using machine vision according to the present invention will be described in detail with reference to the drawings.

However, the intrinsic technical idea of the present invention cannot be said to be limited by the embodiments to be described below, and the intrinsic technical idea of the present invention is given below by a person skilled in the art. It turns out to cover the range that can be easily proposed by a method of substitution or change of the embodiment described in FIG.

In addition, since the terms used below are selected for convenience of description, in grasping the intrinsic technical idea of the present invention, it is not limited to the dictionary meaning and is appropriately interpreted as a meaning consistent with the technical idea of the present invention. Should be.

Hereinafter, exemplary embodiments of the present invention will be described in detail with reference to the accompanying drawings.

1 is a flowchart showing an intelligent robot control process using machine vision according to an embodiment of the present invention, and FIG. 2 is a block configuration of a control device for performing an intelligent robot control process using machine vision according to an embodiment of the present invention. Is also.

Referring to FIGS. 1 and 2, the robot control unit 150 may control the robot arm unit 110 to move to a preset reference position (step 101).

Here, the robot arm unit 110 moves the position of the robot in the XYZ axis according to the control of the robot control unit 150, and a horizontal rotation axis in charge of movement in the X axis and Y axis among spatial coordinates, Among the spatial coordinates, a plurality of vertical rotation axes in charge of movement to the Z axis, and a plurality of joints provided between the horizontal rotation axis and each vertical rotation axis may be included.

The robot arm unit 110 may refer to an equipment provided with multi-joints that are freely moved for each axis on a spatial coordinate.

The reference position can be represented by RPi (xi, yi, ri), which is a preset position corresponding to the i position, where the i position is the number of materials in one layer in which the object to be taken out is loaded, and the i-th position (for example, 1-17 Etc.), xi and yi denote coordinates, and ri denotes a misaligned angle of the detected pattern.

In addition, the robot control unit 150 may control the vision camera unit 120 to take a picture of an object to be taken out (step 102).

Here, the vision camera unit 120 is to photograph the take-out object loaded on the pallet P under the control of the robot control unit 150, and is displayed on the pallet through a vision camera provided at the end of the robot arm unit 110. The loaded object to be taken out may be photographed and the vision image may be provided to the robot control unit 150.

The pallet (P) as described above will be provided in a state in which the cap-shaped disk 10 is loaded on the interlayer paper 20 as shown in FIG. 3 in order to put the rough-processed disk into the turning process (or polishing process). I can.

Next, the robot control unit 150 may detect the cap-shaped disk from the vision image captured through the vision camera unit 120 (step 103). Here, the cap-type disk may be detected by comparing with reference patterns previously stored for each pattern of the cap-type disk 10 (ie, a concentric edge shape corresponding to the cap shape).

If the cap-type disk is not detected for each pattern previously stored in step 105, the robot controller 150 may perform exposure time correction and illumination intensity adjustment of the vision camera.

On the other hand, the robot control unit 150 counts the detection errors that failed to detect the cap-type disk in the vision image, and controls the robot arm unit 110 to change the search position when the detection error threshold is reached. The search location can be changed by a preset distance (e.g., +100mm, -100mm, etc.), and an abnormal status alarm can be provided when the change frequency threshold is reached while changing the search location. An audible alarm such as a buzzer or a warning sound can be output to restart after a cause analysis of a hazardous device is stopped, or a visual alarm that displays a warning screen on the administrator terminal can be output.

In addition, when the cap-type disk is detected, the robot control unit 150 may control and move the robot arm unit 110 according to an offset value between the vision position at which the cap-type disk is detected and a preset reference position (step 104).

For example, the reference position can be expressed as RPi (xi, yi, ri), which is a preset position corresponding to the i position, and the vision position is VPi (xi, yi, ri), which is the position detected in the vision image of the i position. And the offset distance (D) can be expressed as |VPi-RPi|. At this time, the i position is the number of materials in one layer in which the object to be taken out is loaded, and is the i-th position (for example, 1-17, etc.), xi and yi refer to coordinates, and ri refers to the wrong angle of the detected pattern. , As shown in FIG. 4, a reference position (RPi), a vision position (VPi), and an offset distance (D) may be indicated.

Here, in step 104, if the offset distance between the reference position and the vision position is less than the offset distance threshold value, the robot control unit 150 may control the robot arm unit 110 to move to the vision position. When the moving distance of) converges to the offset distance within the error range, an offset value corresponding to the reference position can be obtained. For example, when the moving distance of the robot arm 110 converges to the offset distance within the error range (eg, ±1mm, etc.) (that is, when the offset distance converges to 0), corresponding to the reference position The offset value can be obtained.

Since the robot control unit 150 controls the robot arm unit 110 to always come to the focal center of the light source axis by tracking the cap-shaped disk 10 until the offset value converges to almost zero, the detection object is detected. A tracking algorithm that can reduce errors can be applied.

Accordingly, if the detection object is out of the preset reference position range, the detection object cannot be retrieved, so that the detection target within the preset reference position range is tracked and the robot arm unit 110 is controlled to move back to the preset reference position. By measuring later, the position where the offset value is converged to zero can be calculated as the optimal position.

In addition, in step 104, the robot control unit 150 may move the robot arm unit 110 by summing the reference position and the offset value to determine an extraction position reflecting the vision offset value. At this time, the robot control unit 150 may change the coordinate system of the robot arm unit 110 to the take-out coordinate system before controlling the robot arm unit 110 to move.

Here, the offset value of the robot at the i position can be expressed as "OPi(xi, yi, ri)=VPi(xi, yi, ri)-RPi(xi, yi, ri)", and the vision offset value at the i position The take-out position, which is the position to be picked up reflecting the value, may be expressed as "GPi(xi, yi, ri)=RPi(xi, yi, ri) + OPi(xi, yi, ri)".

Next, the robot control unit 150 controls the take-out tool unit 130 to take out the cap-shaped disk 10, and then controls the robot arm unit 110 to transfer the taken out cap-shaped disk 10 (step 105).

In this step 105, under the control of the robot control unit 150, the tongs tools of the take-out tool parts 130A and 130B are operated to take out the cap-shaped disk 10, and then the robot arm part 110 is controlled and transferred to the target position. The take-out tool parts 130A and 130B take out the take-out object under the control of the robot control unit 150, and the take-out tool for taking out the take-out object may be provided with a forceps tool or the like.

For example, as shown in FIG. 7, in the case of the take-out tool 130A having one tongs tool, the connection block 131a connected to the end of the robot arm part 110 and the lower part of the connection block 131b The provided connection post (132a) and the cylinder block (133a) provided under the connection post (132a) to drive the clamp jaw (134a), and inward or outward according to the driving of the cylinder block (133a) It may include a clamp jaw (134a) to move and take out the cap-shaped disk (10).

In addition, as shown in FIG. 8, in the case of the take-out tool 130B having three tongs tools, a connection block 131b connected to the end of the robot arm unit 110, and a connection block 131b provided under the connection block 131b. The connection post 132b and the cylinder block 133b provided under the connection post 132b to drive the clamp jaw 134b, and the cylinder block 133b move inward or outward according to the drive. The clamp jaw 134b for taking out the cap-shaped disk 10 and the connecting block 131b are provided on the side of the cap-shaped disk 10 to grip the cap-shaped disk 10 and transfer it to a polishing machine, or align it in a different posture, Alternatively, it may include a first gripper (135b) and a second gripper (136b), etc. used for taking out the processed material. Of course, the grippers may be provided in various numbers as needed.

Subsequently, when the transfer of the cap-shaped disk 10 is completed, the robot control unit 150 may control the intercalation adsorption unit 140 to adsorb and remove the intercalation paper (step 106).

In this step 106, after the cap-type disk 10 is sequentially taken out and transferred, the suction tool of the intercalation adsorption unit 140 may be lowered to adsorb and remove the separating paper. The intercalation adsorption unit 140 is the robot control unit 150 It can be adsorbed and removed according to the control of.

For example, as shown in Figs. 7 and 8, the gripping part 140 has one end fixed to the take-out tool parts 130A and 103B to rotate up and down, and the other end of the rotating cylinder 141 The suction support block 142 which is fixed and rotates up and down, the rotary shaft piece 143 connected to both ends of the suction support block 142 and the extraction tool parts 130A and 103B, and the suction support block 142 A plurality of adsorption pads 144 that are provided at the bottom of the device to adsorb the separators 20, and the adsorption support block 142 at the center of the plurality of adsorption pads 144 serves as a buffer for the separators 20 to be adsorbed. It may include a buffer pad 145 to perform.

When the cap-type disk 10 is taken out, the interlayer suction unit 140 is maintained in an upright state in the vertical direction, and after all the cap-type disks 10 are taken out and transferred, the rotating cylinder 141 protrudes and extends. The adsorption support block 142 is rotated in a horizontal state along the rotation axis, and the separation paper 20 is stably held through the plurality of adsorption pads 144 and buffer pads 145 provided under the adsorption support block 142. It can be removed by adsorption.

Accordingly, the present invention determines the take-out position according to the vision position detected through the vision camera provided in the robot, the preset reference position and offset distance, and the take-out object through the position control of the robot according to the take-out position. After taking out and transporting the slippers, by repeatedly performing the process of adsorbing and removing the slip sheets, the object to be taken out and the slip sheets from the pallet can be effectively taken out, transferred, and removed using a robot.

In addition, the present invention obtains an offset value of a robot corresponding to a specific position by subtracting a reference position corresponding to a specific position from a vision position detected through a vision image at a specific position, and summing the reference position and the offset value to obtain a specific position By determining the ejection position reflecting the vision offset value, the corresponding disk can be accurately ejected through the precise position control of the robot according to the detection position of the disk to be ejected.

Meanwhile, FIG. 9 is a flowchart for explaining in detail an intelligent robot control process using machine vision according to an embodiment of the present invention, and specifically describes an intelligent robot control process using machine vision according to an embodiment of the present invention. I will do it.

Referring to FIG. 9, the robot control unit 150 may control the robot arm unit 110 to move to a preset reference position, which is a vision camera measurement position (ie, a photographing position of a vision camera) (step 901 ). Here, in the case of a calculation method or a non-uniform (or irregular, unspecified) pattern according to the total number of objects to be taken out, the robot arm unit 110 may be controlled to move to a preset reference position (ie, a measurement position).

In addition, the robot control unit 150 may reset the detection error count for counting when the pattern of the object to be retrieved from the vision image cannot be detected (step 802).

In addition, the robot control unit 150 may control the vision camera unit 120 at a preset reference position to take a picture of an object to be taken out (step 903).

Next, the robot control unit 150 checks whether or not the cap-shaped disk 10 is detected in the vision image photographed with the taken-out video object (step 904).

As a result of the check in step 904, when a detection error that fails to detect the cap-type disk 10 occurs, the robot control unit 150 may count the detection error (step 905).

Next, while performing the process of step 904 again, the robot control unit 150 increases the detection error count and checks whether the detection error count becomes a detection error threshold value (step 906).

As a result of the check in step 906, when the detection error count becomes the detection error threshold value, the robot control unit 150 may control the robot arm unit 110 to change the search position (step 907). Here, the robot control unit 150 controls the movement of the robot arm unit 110 to change the search location by a preset distance (eg, +100mm, -100mm, etc.) in the X-axis, Y-axis, etc. After the search location is changed, the controller 150 may control the vision camera unit 120 to perform step 903 again.

Next, the robot control unit 150 checks whether the number of changes becomes the threshold value of the number of changes (step 908). Here, when the number of changes does not reach the threshold value of the number of changes, the robot control unit 150 may control the search location to be changed to perform the processes of steps 902 to 907 again.

As a result of the check in step 908, when the number of changes becomes the threshold value of the number of changes, the robot controller 150 may provide an abnormal state alarm (step 909). Here, the robot control unit 150 may output an audible alarm such as a buzzer, a warning sound, etc., or a visual alarm displaying a warning screen on the manager terminal so that the device is stopped for cause analysis and then restarted after analysis of the cause. .

On the other hand, as a result of the check in step 904, when the cap-shaped disk 10 is detected, the robot control unit 150 pre-stored reference patterns (ie, concentric circles corresponding to the cup shape) for each pattern of the cap-shaped disk 10. Shape, etc.), it can be checked whether it is detected (step 910).

As a result of the check in step 910, if the pattern of the cap-type disk 10 is not detected, the robot control unit 150 can control the vision camera unit 120 to correct the exposure time and adjust the illumination intensity of the vision camera. Yes (step 911). After that, the robot control unit 150 may control the vision camera unit 120 to perform step 903 again.

On the other hand, as a result of the check in step 910, when detecting each pattern of the cap-shaped disk 10, the robot control unit 150 determines the detected vision position of the cap-shaped disk 10 and an offset distance between the preset reference position. It is checked whether it is less than the offset distance threshold (step 912).

Here, the reference position can be expressed as RPi (xi, yi, ri), which is a preset position corresponding to the i position, and the vision position is VPi (xi, yi, ri), which is the position detected in the vision image of the i position. It can be represented, and the offset distance (D) can be expressed as |VPi-RPi|.

As a result of the check in step 912, if the offset distance exceeds the offset distance threshold value, the robot control unit 150 may control the vision camera unit 120 to perform step 903 again.

On the other hand, as a result of the check in step 912, if the offset distance is less than the offset distance threshold, the robot control unit 150 can control the robot arm unit 110 to move to the vision position of the cap-type disk 10. There is (913).

In addition, in the robot control unit 150, when the moving distance of the robot arm unit 110 converges to the offset distance within an error range (for example, ±1mm, etc.) (that is, when the offset distance converges to 0), the reference An offset value corresponding to the position may be obtained (step 914).

Since the robot control unit 150 controls the robot arm unit 110 to always come to the focal center of the light source axis by tracking the cap-shaped disk 10 until the offset value converges to almost zero, the detection object is detected. A tracking algorithm that can reduce errors can be applied.

Accordingly, if the detection object is out of the preset reference position range, the detection object cannot be retrieved, so that the detection target within the preset reference position range is tracked and the robot arm unit 110 is controlled to move back to the preset reference position. By measuring later, the position where the offset value is converged to zero can be calculated as the optimal position.

In addition, the robot control unit 150 may control the robot arm unit 110 to move by summing the reference position and the offset value to determine the take-out position reflecting the vision offset value (step 915). At this time, the robot control unit 150 may change the coordinate system of the robot arm unit 110 to the take-out coordinate system before controlling the robot arm unit 110 to move.

Here, the offset value of the robot at the i position can be expressed as "OPi(xi, yi, ri)=VPi(xi, yi, ri)-RPi(xi, yi, ri)", and the vision offset value at the i position The take-out position, which is the position to be picked up reflecting the value, may be expressed as "GPi(xi, yi, ri)=RPi(xi, yi, ri) + OPi(xi, yi, ri)".

In addition, the robot control unit 150 may control the robot arm unit 110 to be transferred to a target position after actuating the tongs tools of the take-out tool units 130A and 130B to take out the cap-shaped disk 10 (step 916). ).

Next, the robot control unit 150 may take out and transfer the total number of the detected cup-shaped disks 10 sequentially for each pattern through the above-described process (step 917).

Subsequently, when the transfer of the cap-shaped disk 10 is completed, the robot control unit 150 may control the slipper suction unit 140 to adsorb and remove the slippery paper 20 (step 918).

For example, when the transfer of the cap-type disk 10 is completed, the rotation cylinder 141 protrudes and extends so that the suction support block 142 rotates in a horizontal state along the rotation axis, and the suction support Through the plurality of adsorption pads 144 and buffer pads 145 provided under the block 142, the interlayer paper 20 may be stably adsorbed and removed.

Accordingly, the present invention determines the take-out position according to the vision position detected by the vision camera provided in the robot, the preset reference position and offset distance, and takes out the take-out object through the position control of the robot according to the take-out position. After the transfer, by repeatedly performing the process of adsorbing and removing the slip sheets, the object to be taken out and the slip sheets from the pallet can be effectively taken out, transferred, and removed using a robot.

In addition, the present invention does not require a separate device such as a disk loading device and an alignment device, since it is possible to always take out material from a certain position when performing position control using machine vision. It can be designed and reflected, and through the position control of the robot using machine vision, the material can always be taken out and transferred from a certain position, providing a stable operating environment, and offset values that are wrong in the machine vision for each location to be taken out It can be calculated automatically to maximize productivity.

In addition, according to the present invention, the position of the robot can be controlled using the machine vision, so that a task combined with other processes can be performed, thereby improving economic efficiency and efficiency in performing an automated process.

Claims (10)

Controlling the robot arm in the robot control unit to move to a preset reference position;
Photographing an object to be taken out by controlling the vision camera unit in the robot control unit;
Detecting a cap-type disk from the vision image captured through the vision camera unit;
Controlling and moving the robot arm according to an offset value between the detected vision position of the cap-shaped disk and the preset reference position;
Controlling the take-out tool part by the robot control part to take out the cap-type disk, and then controlling the robot arm part to transfer the taken out cap-type disk; And
When the transfer of the cap-type disk is completed, the robot control unit controls the separation sheet to remove the separation sheet by adsorbing,
Counting a detection error that fails to detect the cap-type disk in the vision image and changing a search location by controlling the robot arm when a detection error threshold is reached after the step of photographing the take-out object;
Providing an abnormal state alarm when a threshold value of the number of times of change is reached while changing the search location; And
In the step of detecting the cap-type disk, when the cap-type disk is not detected for each pattern, performing exposure time correction and illumination intensity adjustment of the vision camera; further comprising,
The step of controlling and moving the robot arm may include controlling the robot arm when the cap-type disk is detected in the vision image and an offset distance between the preset reference position and the vision position is less than an offset distance threshold value. An intelligent robot control method using machine vision for moving to a position and obtaining the offset value corresponding to the preset reference position when the moving distance of the robot arm unit converges to the offset distance within an error range. delete delete delete delete delete The method according to claim 1,
In the step of controlling and moving the robot arm part, the intelligent robot control method using machine vision for moving the robot arm part by summing the preset reference position and the offset value to determine a take-out position reflecting the vision offset value. The method of claim 7,
The step of controlling and moving the robot arm part may include changing a coordinate system of the robot arm part to a take-out coordinate system before controlling the robot arm part to move. The method of claim 8,
In the step of transferring the cap-type disk, after the cap-type disk is removed by actuating the clip tool of the take-out tool part, the intelligent robot control method using machine vision controls the robot arm part to transfer it to a target position. The method according to any one of claims 1 and 7 to 9,
In the step of adsorbing and removing the slip sheets, the cap-type disk is sequentially taken out and transferred, and then the suction tool of the slip sticking portion is lowered to adsorb and remove the slip sheets.

Images