A full-automatic batch high-precision vision feeder correction method for a chip mounter

Abstract

The application discloses a kind of full-automatic batch high-precision vision fly correctors for patching machine, and the correction instrument receives the correction instruction of host computer, and automatically controls the sensor of fly machine and executes the correction instruction of host computer, completes the sensor error of automatic correction fly, and the actual position of material belt transmission is obtained in real time by high-speed camera, and the position is fed back to feeder to correct itself, high frame rate of high-speed camera and fast processing capacity of host computer are used, to achieve high-precision and efficient full-automatic correction, improve the correction precision and efficiency of fly, and the later maintenance of fly is more convenient.

Description

Technical Field

[0001] This invention relates to the field of feeder technology, and in particular to a fully automated, high-precision visual feeder calibration method for pick-and-place machines. Background Technology

[0002] The feeder of a pick-and-place machine is a crucial component. The feeder uses a motor and intermediate transmission gears to drive a material-pulling gear, which in turn pulls the material belt forward via various pulling teeth around the belt, thus feeding the material. However, due to factors such as motor differences, sensor errors, mechanical assembly errors, and transmission errors of mechanical parts, the various mechanisms that pull the material belt in the feeder have different errors. These errors can affect the feeding accuracy and performance of each feeder. Therefore, the feeder needs to have its various error factors adjusted and corrected to match the pick-and-place machine and achieve high-speed, precise feeding.

[0003] The intermediate transmission gear of the feeder also has errors during the transmission process. The mechanical errors caused by the gears cannot meet the feeder's feeding standards. As the feeder belt is being pulled forward, the feeder gears are constantly switching to pull the feeder belt. However, there are errors between the feeder gears and the feeder gears are not evenly distributed. Therefore, it is not feasible to improve the feeding accuracy by controlling the rotation angle of each feeding. It is also impossible to improve the accuracy of the actual position of the feeder belt after transmission. Summary of the Invention

[0004] To address the shortcomings of existing technologies, the purpose of this invention is to provide a fully automated, high-precision visual feeder calibration method for pick-and-place machines. This method automatically calibrates the feeder, improving calibration accuracy and efficiency, and making subsequent feeder maintenance more convenient.

[0005] To achieve the above objectives, the technical solution adopted by this invention is: a fully automated, high-precision, batch-scale vision feeder correction method for pick-and-place machines.

[0006] Step S1, Initialize the machine.

[0007] Power the calibrator and connect it to the host computer. The host computer establishes communication with the calibrator, configures relevant parameters on the host computer, initializes the high-speed camera, locates the feeder device, and connects to the feeder.

[0008] Step S2, Correct sensor error.

[0009] The host computer sends a calibration command to the feeder through the calibrator. The calibrator first sends a motor calibration signal to the feeder to calibrate the motor. After receiving the motor calibration signal, the feeder automatically runs the motor in open loop to obtain the parameters of the motor closed operation and sends them to the calibrator. After the motor calibration is completed, the feeder sends a generator calibration completion signal to the calibrator.

[0010] After the calibrator receives the signal that the calibrator motor has completed calibration, it sends a sensor calibration signal to the feeder for calibrating the sensor. Upon receiving the sensor calibration signal, the feeder automatically runs the motor in closed loop to obtain the sensor error parameters and sends them to the calibrator. After the sensor calibration is completed, the feeder sends a sensor calibration completion signal to the calibrator. After the calibrator receives the sensor calibration completion signal, the sensor error calibration process ends.

[0011] Step S3, correct mechanical errors.

[0012] First, the visual parameters are adjusted through the human-computer interaction interface with camera image interface set on the host computer. The visual parameters include light source parameters, image filtering parameters, binarization parameters, and recognition coordinate parameters. The light source parameters, filtering parameters, and binarization parameters are fine-tuned according to the filtered binarized image displayed by the host computer, and the crosshair center is adjusted to be aligned with the positioning line A of the feeder. The host computer starts to send commands to the feeder through the calibrator and receives the calibrator's reply. The host computer performs image recognition and calculation processing on the image captured by the high-speed camera in real time and sends the material trough position information, waiting for the calibrator's reply. If a signal indicating that the calibration is complete is received, the calibration ends.

[0013] Step S4, test the calibration results.

[0014] After the mechanical error correction is completed, the correction result test is carried out. First, adjust the visual parameters in the image adjustment interface. According to the light source or the material of the strip, fine-tune the visual parameters of the high-speed camera to the appropriate parameters. Then, send the correction command to the feeder through the host computer. After receiving the feeder's reply, the host computer sends the material pit position information to the feeder in real time and waits for the feeder's reply. If a signal indicating that the correction is complete is received, the correction result test ends.

[0015] Furthermore, in step S3, during the high-speed feeding process of the feeder, the host computer performs image processing on the images captured by the high-speed photograph, and then transmits them to the feeder in real time via USB communication. The host computer uses a polygon fitting algorithm to fit a quadrilateral, identifies the position of the quadrilateral of the material pit, obtains the actual coordinates of the material pit, and uses the actual coordinates of the material pit to obtain the actual feeding distance, and automatically corrects the mechanical error of the feeder.

[0016] Furthermore, the image recognition and processing includes the following sub-steps: First, the original image captured by the high-speed camera is obtained by cropping according to the cross coordinates set by the host computer. Then, the original image is filtered and de-noised and binarized according to the filtering and binarization parameters set by the host computer. Next, the image is dilated to fill in the small gaps and obtain a smoother image. Then, the closed contour is scanned to obtain the pixel set of the closed contour. Then, the maximum bounding rectangle of the set is calculated. If the size of the calculated rectangle is within the empirical value range, the rectangle coordinates are output; otherwise, they are not output.

[0017] Furthermore, in step S1, the base of the calibrator is provided with a communication slot. The feeder, which is not connected to the conveyor belt, is inserted into the communication slot of the calibrator, and the calibrator and the feeder establish communication.

[0018] In step S2, the host computer sends a calibration command through the calibrator. The calibrator converts the calibration command sent by the host computer into a control signal for calibrating the feeder. The calibrator controls the motor and sensor of the feeder to execute the calibration command sent by the host computer. The host computer first sends a generator calibration command, and after receiving the motor calibration completion signal, the host computer sends a sensor calibration command to the calibrator.

[0019] Further, in step S3, after completing step S2, the feeder is first removed from the calibrator. After the feeder is connected to the material tape, it is then plugged back into the communication slot of the calibrator. The feeder is then identified again by the host computer. The light source of the calibrator is turned on and the position of the high-speed camera is adjusted. The high-speed camera takes pictures of the material tape and sends the captured visual images to the host computer. After the host computer acquires the visual images, it first performs image segmentation, image filtering, and visual recognition algorithm processing on the visual images to obtain the real-time material pit position of the component and the positioning line A of the feeder.

[0020] The feeder is controlled via a human-machine interface with a camera image interface set on the host computer, allowing for real-time monitoring of the calibration process to check for errors.

[0021] The host computer sets the feeder to calibration mode through the calibrator. The calibrator controls the feeder to move at a constant speed. After the components of the material belt move to the positioning line A, the calibrator sends the command from the host computer to the feeder. The host computer processes the data through an algorithm to obtain mechanical error correction information and sends it to the feeder to automatically calibrate the mechanical error.

[0022] Furthermore, the mainboard of the calibrator is equipped with a main control chip and a USB hub circuit. The feeder is connected to the main control chip of the calibrator via UART, realizing real-time communication between the calibrator and the feeder, enabling the calibrator to control the feeder and obtain the feeder's status. The main control chip obtains the electrical signals generated when the control buttons set on the outside of the calibrator are triggered in real time. The main control chip and the high-speed camera are respectively connected to the USB hub circuit of the mainboard via USB interface. The USB hub circuit is connected to the host computer via USB cable, reducing the number of USB connection cables used between the calibrator and the host computer.

[0023] Furthermore, the calibrator is equipped with a dimming circuit, which outputs a PWM wave to the dimming circuit to control and adjust the light source.

[0024] Furthermore, the housing covers the high-speed camera and the light source, and multiple control buttons are provided on the outside of the housing. The main control chip receives electrical signals from the control buttons and controls the feeder to execute corresponding instructions according to the received electrical signals. The human-machine interface is provided with an image adjustment interface, which adjusts the parameters of image segmentation and image filtering processing through real-time feedback of debugging images to obtain more accurate actual coordinates of the material pit.

[0025] Furthermore, the motor parameters of the motor are self-calibrated through software algorithms. After receiving the start calibration command from the host computer, the feeder runs its own internal calibration code to obtain the calibration parameters.

[0026] Furthermore, the feeder is connected to the calibrator, and the firmware inside the feeder is replaced by communicating with the calibrator through the host computer. A bracket is installed on the base of the calibrator, and a camera position adjustment mechanism is installed on the top of the bracket. The high-speed camera is installed on one side of the camera position adjustment mechanism, and the high-speed camera is located above the feeder slot. The calibrator is equipped with a three-axis dovetail groove manual slide, and the knobs that drive the movement of each axis extend out of the machine housing to facilitate the adjustment and focusing of the high-speed camera and avoid the influence of errors caused by the mechanical parts of the calibrator.

[0027] The advantages of this invention compared to existing technologies are: the calibration instrument acquires the actual position of the conveyor belt in real time through a high-speed camera and feeds the position back to the feeder for self-calibration. By utilizing the high frame rate of the high-speed camera and the fast processing capability of the host computer, it achieves high-precision and efficient fully automatic calibration, improving the calibration accuracy and efficiency of the feeder, and making the subsequent maintenance of the feeder more convenient. Attached Figure Description

[0028] Figure 1 This is a schematic diagram of the structure of the calibration instrument of the present invention.

[0029] Figure 2This is a schematic diagram of the structure of the calibrator's bracket, which is equipped with a three-axis dovetail groove manual slide.

[0030] Figure 3 This is a partial structural diagram of the Benfeda.

[0031] Figure 4 This is a flowchart illustrating the overall operation of the calibration instrument of this invention.

[0032] Figure 5 This is a flowchart of the sensor calibration process of the present invention.

[0033] Figure 6 This is a flowchart illustrating the correction of mechanical errors in this invention.

[0034] Figure 7 This is a flowchart of the image recognition algorithm of the present invention.

[0035] Figure 8 This is a flowchart of the test and correction results of the present invention.

[0036] Figure 9 This is a circuit block diagram of the calibration instrument motherboard of the present invention.

[0037] Figure 10 This is a communication block diagram of the calibration instrument motherboard of the present invention.

[0038] Figure 11 This is a diagram of the main interface of the Feida host computer of the present invention.

[0039] Figure 12 This is a diagram of the calibration mode interface of the feeder host computer of the present invention.

[0040] Figure 13 This is a diagram of the feeder host computer image adjustment interface of the present invention.

[0041] Marked in the image:

[0042] 1. High-speed camera

[0043] 2. Light source

[0044] 3. Feeder Communication Slot

[0045] 4. Fixing components

[0046] 5. Main control circuit board of the calibrator

[0047] 6 Z-axis dovetail groove manual slide

[0048] 7 Y-axis dovetail groove manual slide

[0049] 8 brackets

[0050] 9 X-axis dovetail groove manual slide

[0051] 10. Base. Detailed Implementation

[0052] A fully automated, high-precision, batch-process vision feeder calibration method for pick-and-place machines. Figures 1 to 13 As shown,

[0053] Step S1, Initialize the machine.

[0054] Power the calibrator and connect it to the host computer. The base 10 of the calibrator is equipped with a feeder communication slot 3. First, insert the feeder that is not connected to the material belt into the feeder communication slot 3 of the calibrator. The calibrator and the feeder establish communication. The host computer establishes communication with the calibrator. Configure relevant parameters on the host computer, initialize the high-speed camera 1, find the feeder device and connect to the feeder.

[0055] Step S2, Correct sensor error.

[0056] The host computer sends a calibration command through the calibrator. The calibrator converts the calibration command sent by the host computer into a control signal for calibrating the feeder. It controls the feeder motor and sensor to execute the calibration command sent by the host computer. The host computer first sends a generator calibration command. After receiving the motor calibration completion signal, the host computer sends a sensor calibration command to the calibrator.

[0057] Specifically, the host computer first sends a calibration command to the feeder through the calibrator. The calibrator first sends a motor calibration signal to the feeder to calibrate the motor. After receiving the motor calibration signal, the feeder automatically runs the motor in open loop, obtains the parameters of the motor's closed-loop operation, and sends them to the calibrator. After the motor calibration is completed, the feeder sends a generator calibration completion signal to the calibrator.

[0058] After the calibrator receives the signal that the calibrator motor has completed calibration, it sends a sensor calibration signal to the feeder for calibrating the sensor. Upon receiving the sensor calibration signal, the feeder automatically runs the motor in closed loop to obtain the sensor error parameters and sends them to the calibrator. After the sensor calibration is completed, the feeder sends a sensor calibration completion signal to the calibrator. After the calibrator receives the sensor calibration completion signal, the sensor error calibration process ends.

[0059] The motor parameters of the Feida motor are self-calibrated through the built-in software algorithm of Feida. After Feida receives the start calibration command from the host computer, Feida runs the corresponding calibration code inside itself to obtain the calibration parameters.

[0060] The mainboard of the calibrator is equipped with a main control chip and a USB hub circuit. The feeder is connected to the main control chip of the calibrator via UART, enabling real-time communication between the calibrator and the feeder. This allows the calibrator to control the feeder and obtain its status. The main control chip receives the electrical signals generated when the control buttons set on the outside of the calibrator are triggered. The main control chip and the high-speed camera 1 are connected to the USB hub circuit of the mainboard via USB interfaces. The USB hub circuit is connected to the host computer via a USB cable, reducing the number of USB cables used between the calibrator and the host computer.

[0061] The main control circuit board 5 of the calibrator is installed on the bracket 8. A Y-axis dovetail manual slide 7 is installed on the bracket 8. An X-axis dovetail manual slide 9 is installed on the Y-axis dovetail manual slide 7. A Z-axis dovetail manual slide 6 is installed on the X-axis dovetail manual slide 9. The high-speed camera 1 and the light source 2 are fixed on the Z-axis dovetail manual slide 6 by the fixing component 4 installed on the upper end of the Z-axis dovetail manual slide 6. The Y-axis position of the high-speed camera 1 is adjusted by rotating the manual adjustment knob of the Y-axis dovetail manual slide 7. The X-axis position of the high-speed camera 1 is adjusted by rotating the manual adjustment knob of the X-axis dovetail manual slide 9. The Z-axis position of the high-speed camera 1 is adjusted by rotating the manual adjustment knob of the Z-axis dovetail manual slide 6.

[0062] Step S3, correct mechanical errors.

[0063] After completing step S2, the feeder is removed from the calibrator. After the feeders are connected to the material strips, they are plugged back into the feeder communication slot 3 of the calibrator. The feeders are identified again by the host computer. The light source 2 of the calibrator is turned on and the position of the high-speed camera 1 is adjusted. The high-speed camera 1 takes pictures of the material strip and sends the captured visual images to the host computer. After the host computer acquires the visual images, it performs image segmentation, image filtering, and visual recognition algorithm processing on the visual images to obtain the real-time material pit position of the components and the positioning line A of the feeder. The feeder is controlled through the human-machine interface with a camera image interface set on the host computer. The calibration process is observed in real time to check for errors. The host computer sets the feeder to calibration mode through the calibrator. The calibrator controls the feeder to move at a constant speed. After the components of the material strip move to the positioning line A, the calibrator sends the command from the host computer to the feeder. The host computer processes the data through the algorithm to obtain mechanical error correction information and sends it to the feeder to automatically calibrate the mechanical error.

[0064] Specifically, firstly, the visual parameters are adjusted through the human-computer interaction interface with camera image interface set on the host computer. The visual parameters include light source 2 parameters, image filtering parameters, binarization parameters, and recognition coordinate parameters. The light source 2 parameters, filtering parameters, and binarization parameters are fine-tuned according to the filtered binarized image displayed by the host computer, and the crosshair center is adjusted to be aligned with the positioning line A of the feeder. The host computer starts sending commands to the feeder through the calibrator and receives the calibrator's reply. The host computer performs image recognition and calculation processing on the image captured by high-speed camera 1 in real time and sends the material trough position information, waiting for the calibrator's reply. If a signal indicating that the calibration is complete is received, the calibration ends.

[0065] Image recognition and processing includes the following sub-steps: First, the original image captured by the high-speed camera 1 is cropped according to the cross coordinates set by the host computer. Then, the original image is filtered to remove noise and binarized according to the filtering and binarization parameters set by the host computer. Next, the image is dilated to fill in the small gaps and obtain a smoother image. Then, the closed contour is scanned to obtain the pixel set of the closed contour. Then, the maximum bounding rectangle of the set is calculated. If the size of the calculated rectangle is within the empirical value range, the rectangle coordinates are output; otherwise, they are not output.

[0066] The calibrator is equipped with a dimming circuit. The calibrator outputs a PWM wave to the dimming circuit, which then controls and adjusts the light source 2.

[0067] The calibrator's casing encloses the high-speed camera 1 and the light source 2, preventing the external light source 2 from affecting the actual image quality and avoiding eye strain from excessively bright supplementary lighting. It also protects the internal components. Multiple control buttons are located on the outside of the casing. The main control chip receives electrical signals from these buttons and controls the feeder to execute corresponding instructions based on these signals. The calibrator of this invention features control buttons, which are more convenient than the membrane buttons on the feeder. The host computer's human-machine interface includes an image adjustment interface. Real-time feedback of the debugging image allows adjustment of the parameters for image segmentation and image filtering to obtain more accurate actual coordinates of the material pit.

[0068] The human-machine interface of the feeder is relatively simple, with four membrane buttons. These four membrane buttons are used to adjust and calibrate various errors, which is labor-intensive and inefficient. Furthermore, when using the manual membrane buttons to obtain the actual walking step length of the feeder, the feeder body shakes due to the human pressing the membrane buttons, and the low accuracy of the human eye and slow human reaction result in the accuracy not meeting the requirements.

[0069] The calibration instrument uses a high-speed camera 1 to acquire the actual position of the conveyor belt in real time and feeds the position back to the feeder for self-calibration. With the high frame rate of the high-speed camera 1 and the fast processing of the PC, it can achieve high-precision and efficient fully automatic calibration.

[0070] Errors in mechanical parts and during mechanical assembly can cause variations in the actual step length of the feeder during feeding. For example, if the feeder's minimum step length is 2mm, and the pull wheel of the feeder travels one revolution, the corresponding 2mm step length corresponds to 72 pieces of material. Therefore, it is necessary to obtain the actual step length of 72 pieces of material and use this parameter to make corrections during the feeding process so that the feeder's feeding accuracy meets the actual mounting requirements.

[0071] During the high-speed feeding process of the feeder, the host computer uses image processing technology to process the images captured by high-speed photography, and then transmits them to the feeder in real time via USB communication. The host computer uses a polygon fitting algorithm to fit a quadrilateral, identifies the position of the quadrilateral of the material pit, obtains the actual coordinates of the material pit, uses the actual coordinates of the material pit to obtain the actual feeding distance, and automatically corrects the mechanical error of the feeder.

[0072] The host computer sets the feeder to calibration mode through the calibrator. The calibrator controls the feeder to move at a constant speed. After the components of the material belt move to the positioning line A, the calibrator sends the command from the host computer to the feeder. The host computer processes the data through an algorithm to obtain mechanical error correction information and sends it to the feeder to automatically calibrate the mechanical error.

[0073] Step S4, test the calibration results.

[0074] After the mechanical error correction is completed, the correction result test is carried out. First, adjust the visual parameters in the image adjustment interface. According to the material of the light source 2 or the material strip, fine-tune the visual parameters of the high-speed camera 1 to the appropriate parameters. Then, send the correction command to the feeder through the host computer. After receiving the feeder's reply, the host computer sends the material pit position information to the feeder in real time and waits for the feeder's reply. If a signal indicating that the correction is complete is received, the correction result test ends.

[0075] Figures 11 to 13 As shown, by adjusting the parameters of image segmentation and image filtering through real-time feedback debugging images, the coordinates are made more accurate when positioning the material trough. At the same time, you can choose to automatically or manually adjust the recognition calibration coordinates. The sliders from top to bottom correspond to the binarization parameters, image filtering parameters, and the size of the light source 2, respectively. Clicking or long-pressing the direction arrow can adjust the recognition coordinate center.

[0076] Connect the feeder to the calibrator and use the communication between the host computer and the calibrator to replace the firmware inside the feeder.

[0077] A bracket 8 is installed on the base 10 of the calibrator, and a camera position adjustment mechanism is installed above the bracket 8. A high-speed camera 1 is mounted on one side of the camera position adjustment mechanism, positioned above the feeder slot. The calibrator is equipped with a three-axis dovetail manual slide, with knobs for driving each axis extending outside the housing to facilitate adjustment and focusing of the high-speed camera 1, thus avoiding the influence of errors in the mechanical parts of the calibrator. This invention provides a stable mechanical structure through the three-axis dovetail manual slide, reducing errors caused by vibration during feeder feeding or calibration.

[0078] This invention improves the factory output efficiency of feeders in mass production by using a calibration instrument, while also improving the feeding accuracy of the feeders. Furthermore, it can be used for the later maintenance of feeders, making the later maintenance of feeders simpler, more convenient, and faster.

Claims

1. A fully automated, high-precision, batch-process vision feeder calibration method for a pick-and-place machine, characterized in that: Step S1, Initialize the machine. Power the calibrator and connect it to the host computer. The host computer establishes communication with the calibrator, configures relevant parameters on the host computer, initializes the high-speed camera, locates the feeder device, and connects to the feeder. The mainboard of the calibrator is equipped with a main control chip and a USB hub circuit. The feeder is connected to the main control chip of the calibrator through UART to realize real-time communication between the calibrator and the feeder, enabling the calibrator to control the feeder and obtain the feeder's status. The main control chip obtains the electrical signals generated when the control buttons set on the outside of the calibrator are triggered in real time. Step S2, Correct sensor error. The host computer sends a calibration command to the feeder through the calibrator. The calibrator first sends a motor calibration signal to the feeder to calibrate the motor. After receiving the motor calibration signal, the feeder automatically runs the motor in open loop to obtain the parameters of the motor closed operation and sends them to the calibrator. After the motor calibration is completed, the feeder sends a generator calibration completion signal to the calibrator. After the calibrator receives the signal that the calibrator motor has completed calibration, it sends a sensor calibration signal to the feeder for calibrating the sensor. Upon receiving the sensor calibration signal, the feeder automatically runs the motor in closed loop to obtain the sensor error parameters and sends them to the calibrator. After the sensor calibration is completed, the feeder sends a sensor calibration completion signal to the calibrator. After the calibrator receives the sensor calibration completion signal, the sensor error calibration process ends. The motor parameters of the Feida motor are self-calibrated through the built-in software algorithm of Feida. After Feida receives the start calibration command from the host computer, Feida runs the corresponding calibration code inside itself to obtain the calibration parameters. Step S3, correct mechanical errors. After completing step S2, the feeders are first removed from the calibrator. After the feeders are connected to the material tape, they are plugged back into the communication slot of the calibrator. The host computer then identifies the feeders again, turns on the calibrator's light source, adjusts the position of the high-speed camera, takes pictures of the material tape with the high-speed camera, and sends the captured visual images to the host computer. After the host computer acquires the visual image, the visual parameters, including light source parameters, image filtering parameters, binarization parameters, and recognition coordinate parameters, are adjusted through the human-computer interaction interface with a camera image interface on the host computer. The light source parameters, filtering parameters, and binarization parameters are fine-tuned based on the filtered binarized image displayed on the host computer, aligning the crosshair center with the feeder's positioning line A. This enables the host computer to perform image segmentation, image filtering, and visual recognition algorithm processing on the visual image, obtaining the real-time component pit position and the feeder's positioning line A. The feeder is controlled through a human-machine interface with a camera image interface set on the host computer. The calibration process is monitored in real time to check for errors. Then, the host computer sets the feeder to calibration mode through the calibrator. The calibrator controls the feeder to move at a constant speed. After the components of the material belt move to the positioning line A, the calibrator sends the command from the host computer to the feeder. After the host computer processes the data through the algorithm, it obtains the mechanical error correction information and sends it to the feeder to achieve automatic calibration of mechanical errors. Among them, automatic calibration of mechanical error starts when the host computer sends a command to the feeder through the calibrator and receives a reply from the calibrator. The host computer performs image recognition and calculation on the image captured by the high-speed camera in real time and sends the material trough position information, waiting for the calibrator to reply. If a signal indicating that the calibration is complete is received, the calibration ends. Step S4, test the calibration results. After the mechanical error correction is completed, the correction result test is carried out. First, adjust the visual parameters in the image adjustment interface. According to the light source or the material of the strip, fine-tune the visual parameters of the high-speed camera to the appropriate parameters. Then, send the correction command to the feeder through the host computer. After receiving the feeder's reply, the host computer sends the material pit position information to the feeder in real time and waits for the feeder's reply. If a signal indicating that the correction is complete is received, the correction result test ends.

2. The fully automated batch high-precision vision feeder correction method for a pick-and-place machine according to claim 1, characterized in that: In step S3, during the high-speed feeding process of the feeder, the host computer performs image processing on the images captured by the high-speed photo and then transmits them to the feeder in real time via USB communication. The host computer uses a polygon fitting algorithm to fit a quadrilateral, identifies the position of the quadrilateral in the material pit, obtains the actual coordinates of the material pit, uses the actual coordinates of the material pit to obtain the actual feeding distance, and automatically corrects the mechanical error of the feeder.

3. The fully automated batch high-precision vision feeder calibration method for a pick-and-place machine according to claim 1, characterized in that: The image recognition and processing includes the following sub-steps: First, the original image captured by the high-speed camera is obtained by cropping according to the cross coordinates set by the host computer. Then, the original image is filtered and de-noised and binarized according to the filtering and binarization parameters set by the host computer. Next, the image is dilated to fill in the small gaps and obtain a smoother image. Then, the closed contour is scanned to obtain the pixel set of the closed contour. Then, the maximum bounding rectangle of the set is calculated. If the size of the calculated rectangle is within the empirical value range, the rectangle coordinates are output; otherwise, they are not output.

4. The fully automated batch high-precision vision feeder correction method for a pick-and-place machine according to claim 1, characterized in that: In step S1, the base of the calibrator is provided with a communication slot. The feeder, which is not connected to the conveyor belt, is inserted into the communication slot of the calibrator, and the calibrator and the feeder establish communication. In step S2, the host computer sends a calibration command through the calibrator. The calibrator converts the calibration command sent by the host computer into a control signal for calibrating the feeder. The calibrator controls the motor and sensor of the feeder to execute the calibration command sent by the host computer. The host computer first sends a generator calibration command, and after receiving the motor calibration completion signal, the host computer sends a sensor calibration command to the calibrator.

5. The fully automated batch high-precision vision feeder correction method for a pick-and-place machine according to claim 1, characterized in that: The main control chip and the high-speed camera are respectively connected to the USBHUB circuit of the motherboard via USB interface, and the USBHUB circuit is connected to the host computer via USB cable.

6. The fully automated batch high-precision vision feeder correction method for a pick-and-place machine according to claim 1, characterized in that: The calibrator is equipped with a dimming circuit. The calibrator outputs a PWM wave to the dimming circuit, which then controls and adjusts the light source.

7. The fully automated batch high-precision vision feeder correction method for a pick-and-place machine according to claim 1, characterized in that: The housing encloses the high-speed camera and the light source. Multiple control buttons are provided on the outside of the housing. The main control chip receives electrical signals from the control buttons and controls the feeder to execute corresponding instructions based on the received electrical signals. The human-machine interface is provided with an image adjustment interface, which adjusts the parameters of image segmentation and image filtering processing through real-time feedback of the debugging image to obtain more accurate actual coordinates of the material pit.

8. The fully automated batch high-precision vision feeder correction method for a pick-and-place machine according to claim 1, characterized in that: Connect the feeder to the calibrator. A bracket is installed on the base of the calibrator. A camera position adjustment mechanism is installed on the bracket. The high-speed camera is installed on one side of the camera position adjustment mechanism. The high-speed camera is located above the feeder slot. The calibrator is equipped with a three-axis dovetail groove manual slide. The knobs that drive the movement of each axis extend out of the machine housing, which facilitates the adjustment and focusing of the high-speed camera and avoids the influence of errors caused by the mechanical parts of the calibrator.

Images