A weld seam tracking and temperature measurement device and method based on image recognition

Abstract

This invention relates to the field of weld seam tracking, specifically disclosing a weld seam tracking and temperature measurement device and method based on image recognition. The device includes a camera, a temperature sensor, a controller, and a motor. The drive end of the motor is fixedly connected to the camera and temperature sensor, driving them to rotate horizontally. The controller is electrically connected to the motor, temperature sensor, and camera. By acquiring the position of the stirring head in real time through the camera and using an algorithm for control, the temperature sensor is kept aligned with the stirring head in real time, achieving temperature monitoring. The image recognition-based weld seam tracking and temperature measurement method can react instantly to the movement of the stirring head and control the motor to drive the rotating platform so that the temperature sensor always faces the stirring head. This results in a faster response, smaller temperature measurement deviation, and the ability to measure the temperature of the stirring head without contact, avoiding the malfunction of the stirring head caused by contact temperature measurement.

Description

Technical Field

[0001] This invention relates to the field of weld seam tracking, specifically to a weld seam tracking and temperature measurement device and method based on image recognition. Background Technology

[0002] There are two common methods for weld seam tracking technology. One is the target tracking method based on vision sensors. This method uses sensors to detect the position of the FSW stirring head and measure the temperature. However, due to various unforeseen circumstances such as harsh welding environments, the sensors may malfunction, and the welding accuracy cannot be guaranteed. The other is the target tracking method based on a given trajectory. This method requires specifying the welding path, which results in low production efficiency.

[0003] Commonly used weld seam tracking sensors include CCD vision sensors, arc sensors, and laser two-dimensional contour scanning sensors. CCD vision sensors and arc sensors are used in arc welding, while laser two-dimensional contour scanning sensors are used in weld seam contour modeling. However, various uncertainties during the welding process can affect the weld seam and cause deviations. Relying on sensors to extract deviation information can greatly increase the likelihood of weld seam tracking failure. Summary of the Invention

[0004] To address the technical problems of low efficiency in trajectory target tracking and high failure rate of tracking sensors in existing technologies, this invention provides a weld seam tracking and temperature measurement device and method based on image recognition.

[0005] The technical solution adopted by the present invention to achieve the above objectives is: a weld seam tracking and temperature measurement device based on image recognition, including a camera 3, a temperature sensor 4, a controller 5, and a motor 6. The driving end of the motor 6 is fixedly connected to the camera 3 and the temperature sensor 4, and is used to drive the camera 3 and the temperature sensor 4 to rotate in the horizontal direction. The controller 5 is electrically connected to the control motor 6, the temperature sensor 4, and the camera 3 respectively.

[0006] Preferably, the device also includes a bracket 1, and a rotating platform 2 is provided on the top of the bracket 1. The camera 3 and the temperature sensor 4 are fixed on the rotating platform 2. The drive end of the motor 6 is fixedly connected to the bottom of the rotating platform 2 and is used to drive the rotating platform 2 to rotate in the horizontal direction. The bracket 1 is also provided with a power supply 7, which is used to supply power to the camera 3, the temperature sensor 4, the controller 5, and the motor 6.

[0007] Preferably, the controller 5 is an STM32 microcontroller;

[0008] Preferably, the bracket 1 is a hollow frustum, and the power supply 7 and the controller 5 are fixed to the inner wall of the bottom surface of the hollow frustum;

[0009] A weld seam tracking and temperature measurement method based on image recognition, characterized by employing an image recognition-based weld seam tracking and temperature measurement device, specifically including the following steps:

[0010] S1: Acquire images of the moving stirring head position using an image recognition-based weld seam tracking and temperature measurement device;

[0011] S2: The obtained image of the stirring head position is enhanced by histogram equalization;

[0012] S3: Train the YOLO object detection network using backpropagation; then feed the image enhanced by histogram equalization into the trained YOLO object detection network for object detection and output the coordinate information of the stirring head in the image;

[0013] S4: During the movement of the stirring head, the coordinate information of the stirring head is obtained again by repeating steps S1-S3. The position information of the stirring head is compared with the position information of the stirring head in the previous two steps by using the position offset algorithm to calculate the moving distance of the stirring head.

[0014] S5: By combining the motor PID control algorithm with the moving distance of the stirring head, the motor 6 is controlled in real time to drive the temperature sensor 4 to track the stirring head for temperature measurement.

[0015] Preferably, in step S2, the histogram equalization formula is:

[0016]

[0017] Where MN is the total number of image pixels, n j Let L be the histogram distribution of the j-th pixel, L be the number of gray levels in the image, j be the number of pixels in the image, k be the gray level of the image, and S be the gray level of the image. k The grayscale value after equalization;

[0018] Preferably, the algorithm obtained after training the YOLO object detection network in step S3 is as follows:

[0019] H(x)=G[b i +W i s(b i-1 +W i-1 x)]

[0020] Where H(*) is the YOLO object detection network, x is the input vector, W is the neuron connection weight, b is the bias, i is the number of network layers, and G and s are activation functions;

[0021] Preferably, the expression for outputting the coordinate information of the stirring head in the image in step S3 is:

[0022] I(x,y)=H(img)

[0023] Where img is the enhanced image, x and y represent the pixel positions of the stirring head in the image, and I(x,y) is the coordinate information of the stirring head in the image;

[0024] Preferably, the expression for calculating the moving distance of the stirring head according to the position offset algorithm in step S4 is as follows:

[0025] D = α[I2(x,y) - I1(x,y)]

[0026] Where α is the ratio of pixel displacement in the image to actual machine displacement, I2 is the coordinate information of the next stirring head in the image, I1 is the coordinate information of the previous stirring head in the image, x and y represent the pixel position of the stirring head in the image, and D is the actual movement distance.

[0027] Preferably, the expression for controlling the motor 6 according to the motor PID control algorithm and tracking the stirring head for temperature measurement in step S5 is as follows:

[0028] v = PID(D)

[0029] Where v is the motor speed after control, and D is the actual distance traveled.

[0030] The beneficial effects of this invention are that it can react instantly when the stirring head moves and control the motor to drive the rotary table so that the temperature sensor always faces the stirring head. The reaction is faster, the temperature measurement deviation is smaller, and the temperature of the stirring head can be measured without contact with the stirring head, thus avoiding the inability of the stirring head to work properly due to contact temperature measurement. Attached Figure Description

[0031] Figure 1 This is a schematic diagram of the structure of a weld seam tracking and temperature measurement device based on image recognition according to the present invention.

[0032] Figure 2 This is a flowchart of a weld seam tracking and temperature measurement method based on image recognition according to the present invention.

[0033] Figure 3 This is a flowchart of the operation of a weld seam tracking and temperature measurement device based on image recognition according to the present invention.

[0034] In the diagram: 1: Support; 2: Rotary table; 3: Camera; 4: Temperature sensor; 5: Controller; 6: Motor; 7: Power supply. Detailed Implementation

[0035] The embodiments of the present invention will be described in further detail below with reference to the accompanying drawings and examples. The following examples are for illustrative purposes only and should not be construed as limiting the scope of the invention.

[0036] In the description of this invention, it should be noted that the terms "upper", "lower", "bottom", etc., indicate the orientation or positional relationship based on the orientation or positional relationship shown in the accompanying drawings. They are only for the convenience of describing this invention and simplifying the description, and do not indicate or imply that the device or element referred to must have a specific orientation, or be constructed and operated in a specific orientation. Therefore, they should not be construed as limiting this invention.

[0037] In the description of this invention, it should be noted that, unless otherwise explicitly specified and limited, the term "connection" should be interpreted broadly. For example, it can be a fixed connection, a detachable connection, or an integral connection; it can be a mechanical connection or an electrical connection; it can be a direct connection or an indirect connection through an intermediate medium; it can be a connection within two components. Those skilled in the art can understand the specific meaning of the above terms in this invention based on the specific circumstances. Furthermore, in the description of this invention, unless otherwise stated, "a plurality of" means two or more.

[0038] like Figure 1 As shown, a weld seam tracking and temperature measurement device based on image recognition includes a camera 3, a temperature sensor 4, a controller 5, and a motor 6. The drive end of the motor 6 is fixedly connected to the camera 3 and the temperature sensor 4, and is used to drive the camera 3 and the temperature sensor 4 to rotate in the horizontal direction. The controller 5 is electrically connected to the control motor 6, the temperature sensor 4, and the camera 3 respectively.

[0039] A weld seam tracking and temperature measurement device based on image recognition includes a bracket 1, which is a hollow truncated cone. A rotating platform 2 is rotatably connected to the top of the bracket 1. A camera 3 and a temperature sensor 4 are fixed on the rotating platform 2. A motor 6 is fixed to the bottom of the rotating platform 2, and the drive end of the motor 6 is fixedly connected to the bottom of the rotating platform 2 to drive the rotating platform 2 and the camera 3 and temperature sensor 4 on it to rotate in the horizontal direction. A power supply 7 and a controller 5 are fixed to the inner wall of the bottom surface of the hollow truncated cone. The power supply 7 is electrically connected to and supplies power to the camera 3, temperature sensor 4, motor 6, and controller 5. The controller 5 can be an STM32 microcontroller. The controller 5 controls the motor 6 to drive the rotating platform 2 to rotate. At the same time, the controller 5 also controls the temperature sensor 4 and the camera 5.

[0040] Camera 5 transmits the captured image information to controller 5 for calculation. Controller 5 drives motor 6, which in turn drives temperature sensor and camera 4 to always be aligned with stirring head and perform real-time temperature measurement. The temperature information is then sent to PC via the wireless transmission module on controller 5.

[0041] like Figure 2-3 As shown, a weld seam tracking and temperature measurement method based on image recognition includes the following steps:

[0042] S1: Acquire images of the moving stirring head position using an image recognition-based weld seam tracking and temperature measurement device;

[0043] S2: The obtained image of the stirring head position is enhanced by histogram equalization;

[0044] The formula for histogram equalization is:

[0045]

[0046] Where MN is the total number of image pixels, n j Let L be the histogram distribution of the j-th pixel, L be the number of gray levels in the image, j be the number of pixels in the image, k be the gray level of the image, and S be the gray level of the image. k This is the grayscale value after equalization.

[0047] S3: Train the YOLO object detection network using backpropagation, and iterate the YOLO object detection network using backpropagation until the evaluation result reaches the optimal condition (the difference between two iterations is less than 0.001) or reaches the maximum number of iterations (1000 times); then feed the image enhanced by histogram equalization into the trained YOLO object detection network for object detection, and output the coordinate information of the stirring head in the image;

[0048] The algorithm obtained after training the YOLO object detection network is:

[0049] H(x)=G[b i +W i s(b i-1 +W i-1 x)]

[0050] Where H(*) is the YOLO object detection network, x is the input vector, W is the neuron connection weight, b is the bias, i is the number of network layers, and G and s are activation functions.

[0051] The expression for the coordinate information of the stirring head in the image is:

[0052] I(x,y)=H(img)

[0053] Where img is the enhanced image, x and y represent the pixel positions of the stirring head in the image, and I(x,y) is the coordinate information of the stirring head in the image.

[0054] S4: During the movement of the stirring head, the coordinate information of the stirring head is obtained again by repeating steps S1-S3. The position information of the stirring head is compared with the position information of the stirring head in the previous two steps by using the position offset algorithm to calculate the moving distance of the stirring head.

[0055] The expression for calculating the moving distance of the stirring head based on the position offset algorithm is as follows:

[0056] D = α[I2(x,y) - I1(x,y)]

[0057] Where α is the ratio of pixel displacement in the image to actual machine displacement, I2 is the coordinate information of the subsequent stirring head in the image, I1 is the coordinate information of the previous stirring head in the image, x and y represent the pixel positions of the stirring head in the image, and D is the actual movement distance.

[0058] S5: By combining the motor PID control algorithm with the moving distance of the stirring head, the motor 6 is controlled in real time to drive the temperature sensor 4 to align with the stirring head, and all the above steps are repeated to achieve closed-loop control and realize the tracking and temperature measurement of the stirring head.

[0059] The expression for controlling the motor and tracking the temperature of the stirring head using the motor PID control algorithm is as follows:

[0060] v = PID(D)

[0061] Where v is the motor speed after control, and D is the actual distance traveled.

[0062] Work process: such as Figure 3 As shown, the temperature measurement method and the control method of camera 3, motor 6, and temperature sensor 4 are deployed in the controller 5 of the temperature measurement device. The temperature measurement device is fixed on the friction stir welding workbench. Camera 3 and temperature sensor 4 face the stirring head. When the stirring head starts to move, the motor 6 is controlled by the temperature measurement method to drive the rotary table 2 to rotate, so that camera 3 and temperature sensor 4 on the rotary table 2 are always aligned with the stirring head and measure the temperature of the stirring head. Temperature sensor 4 then transmits the obtained temperature information to controller 5. Controller 5 transmits the information to PC wirelessly to complete the temperature measurement of the weld. The above steps are repeated to achieve closed-loop control and realize the tracking and temperature measurement of the stirring head.

[0063] The beneficial effects of this invention are that it can react instantly when the stirring head moves and control the motor to drive the rotary table so that the temperature sensor always faces the stirring head. The reaction is faster, the temperature measurement deviation is smaller, and the temperature of the stirring head can be measured without contact with the stirring head, thus avoiding the inability of the stirring head to work properly due to contact temperature measurement.

[0064] This invention has been described through embodiments. Those skilled in the art will understand that various changes or equivalent substitutions can be made to these features and embodiments without departing from the spirit and scope of the invention. Furthermore, under the teachings of this invention, these features and embodiments can be modified to adapt to specific situations and materials without departing from the spirit and scope of the invention. Therefore, this invention is not limited to the specific embodiments disclosed herein, and all embodiments falling within the scope of the claims of this application are within the protection scope of this invention.

Claims

1. A weld seam tracking and temperature measurement method based on image recognition, characterized in that, A weld seam tracking and temperature measurement device based on image recognition is used, including a camera (3), a temperature sensor (4), a controller (5), and a motor (6). The driving end of the motor (6) is fixedly connected to the camera (3) and the temperature sensor (4) to drive the camera (3) and the temperature sensor (4) to rotate in the horizontal direction. The controller (5) is electrically connected to the control motor (6), the temperature sensor (4), and the camera (3) respectively. The device also includes a bracket (1), and a rotating platform (2) is provided on the top of the bracket (1). The camera (3) and the temperature sensor (4) are fixed on the rotating platform (2). The driving end of the motor (6) is fixedly connected to the bottom of the rotating platform (2) to drive the rotating platform (2) to rotate in the horizontal direction. A power supply (7) is also provided in the bracket (1) to supply power to the camera (3), the temperature sensor (4), the controller (5), and the motor (6). The weld seam tracking and temperature measurement method using the above-mentioned image recognition-based weld seam tracking and temperature measurement device specifically includes the following steps: S1: Acquire images of the moving stirring head position using an image recognition-based weld seam tracking and temperature measurement device; S2: The obtained image of the stirring head position is enhanced by histogram equalization; S3: Train the YOLO object detection network using backpropagation; then feed the image enhanced by histogram equalization into the trained YOLO object detection network for object detection and output the coordinate information of the stirring head in the image; S4: During the movement of the stirring head, the coordinate information of the stirring head is obtained again by repeating steps S1-S3. The position information of the stirring head is compared with the position information of the stirring head in the previous two steps by using the position offset algorithm to calculate the distance moved by the stirring head. S5: By combining the motor PID control algorithm with the moving distance of the stirring head, the motor (6) is controlled in real time to drive the temperature sensor (4) to track the stirring head for temperature measurement.

2. The weld seam tracking and temperature measurement method based on image recognition according to claim 1, characterized in that, In step S2, the histogram equalization formula is: , =0,1,2,…， ； in, For the total number of pixels in the image, For the first Histogram distribution of images per pixel It is the number of gray levels in the image. The number of pixels in the image, For image grayscale levels, This is the grayscale value after equalization.

3. The weld seam tracking and temperature measurement method based on image recognition according to claim 2, characterized in that, The algorithm obtained after training the YOLO object detection network in step S3 is as follows: ； in, This is the YOLO object detection network. For the input vector, For neuron connection weights, For bias, For the number of network layers, , This is the activation function.

4. The weld seam tracking and temperature measurement method based on image recognition according to claim 3, characterized in that, The expression for outputting the coordinate information of the stirring head in the image in step S3 is: ； in For the enhanced image, , This indicates the pixel position of the stirring head in the image. This shows the coordinates of the stirring head in the image.

5. The weld seam tracking and temperature measurement method based on image recognition according to claim 4, characterized in that, The expression for calculating the moving distance of the stirring head in step S4 based on the position offset algorithm is as follows: ； in This is the ratio of pixel displacement in the image to actual machine displacement. This shows the coordinates of the stirring head in the image for the next stirring cycle. This shows the coordinates of the previous stirring head in the image. , This indicates the pixel position of the stirring head in the image. This represents the actual distance traveled.

6. The weld seam tracking and temperature measurement method based on image recognition according to claim 5, characterized in that, The expression for controlling the motor (6) according to the motor PID control algorithm and tracking the stirring head for temperature measurement in step S5 is as follows: ； in To control the speed of the rear motor, This represents the actual distance traveled.

7. The weld seam tracking and temperature measurement method based on image recognition according to claim 1, characterized in that, The controller (5) is an STM32 microcontroller.

8. The weld seam tracking and temperature measurement method based on image recognition according to claim 1, characterized in that, The bracket (1) is a hollow truncated cone, and the power supply (7) and controller (5) are fixed on the inner wall of the bottom surface of the hollow truncated cone.

Images