High-resolution angular displacement absolute measurement method and device based on color coding imaging

Abstract

The invention discloses a high-resolution angular displacement absolute measurement method and device based on color coding imaging, belongs to the technical field of photoelectric angular displacement measurement, and aims to solve the problems in the prior art. A stepping motor of the device is connected with the center of the back of a circular coded disc, and a bearing of the stepping motor rotates to drive the circular coded disc to rotate; the laser light source and the high-speed CCD camera are placed right in front of the circular coded disc, and the high-speed CCD camera, the stepping motor and the center of the circular coded disc are coaxially placed and are connected with the PC terminal; laser emitted by the laser light source is vertically incident on a coded disc outer ring RGB color block on the front surface of the circular coded disc to form a laser circular light spot as a target detection point; when the stepping motor rotates, the high-speed CCD camera shoots the circular coded disc in real time to obtain a coded disc image signal with a target detection point; and the high-speed CCD camera shoots and stores the image signal in real time and uploads the image signal to the PC terminal for specific demodulation of rotation angle information. The device and the measurement method thereof have the measurement advantages of higher precision and high resolution.

Description

Technical Field

[0001] The invention discloses a high-resolution angular displacement absolute measurement method and device based on color coding imaging, belonging to the technical field of photoelectric angular displacement measurement. Background Art

[0002] Image-based angular displacement measurement technology plays a vital role in the field of photoelectric measurement. The rapid development of industrial technology fields such as aerospace military fields and precision machinery requires not only high precision and high resolution, but also higher requirements for its cost, ease of equipment operation, and anti-interference ability. The core component of the current image-based angular displacement measurement device is the grating code disk, which is usually composed of a light source, a grating code disk, an image sensor, and a signal processing part. During operation, the light source shines on the grating code disk engraved with a certain pattern. When the system rotates, the code disk pattern changes in a certain regular manner accordingly. The image sensor is placed close to the code disk, and the angle is decoded by capturing the local coding pattern that changes back and forth in the field of view.

[0003] The grating code disks in current image-based grating angular displacement measurement devices mostly use black and white engraving, that is, the encoding of two variable factors of 0-1. Image sensors mostly use CCD / CMOS. Because of its small imaging surface, the code disk usually uses a single circle engraving, which restricts the further improvement of the code disk encoding resolution. In addition, the grating disks are mostly based on glass materials for light-transmitting and opaque circumferential engraving, which increases the production cost. The light source also mostly uses a point or parallel light source. Affected by its diffuse reflection and light refraction factors, the black and white pattern of the grating code disk obtained by the image sensor produces errors. At the same time, the device requires a small spacing between the sensor and the code disk. During the operation of the device, dust and dirt on the grating disk, vibration and deflection of the rotating system, etc., will also affect the imaging, which will lead to increased measurement errors and weak anti-interference ability. Summary of the invention

[0004] In order to solve the problems that the grating code disk of the existing image-type grating angular displacement measurement device only adopts the black and white engraved line mode, the encoding is only a single single-circle encoding due to the restriction of the small photosensitive imaging surface of the image sensor, and the code disk and the image sensor in this structure are close, which makes the installation and adjustment difficult and the anti-environmental interference ability is weak, making it difficult to further improve the resolution, the present invention provides a high-resolution angular displacement absolute measurement method and device based on color coding imaging.

[0005] The technical solution adopted by the present invention is as follows:

[0006] A high-resolution absolute angular displacement measurement device based on color-coded imaging, which consists of a stepper motor, a circular code disc, a laser light source, a high-speed CCD camera and a PC terminal;

[0007] The stepper motor is connected to the back center of the circular code disk, and the rotation of the bearing of the stepper motor drives the rotation of the circular code disk; the laser light source and the high-speed CCD camera are placed in front of the circular code disk, wherein the high-speed CCD camera is coaxially placed with the stepper motor and the center of the circular code disk, and connected to the PC end; the laser light source emits a laser that is vertically incident on the RGB color block on the outer ring of the circular code disk on the front of the circular code disk, forming a laser circular spot, which serves as a target detection point; when the stepper motor rotates, the high-speed CCD camera takes a real-time shot of the circular code disk to obtain a code disk image signal with a target detection point; the high-speed CCD camera takes a real-time shot of the image signal, saves it and uploads it to the PC end, and the PC end performs a specific demodulation of the rotation angle information.

[0008] The circular code disk is arranged according to the matrix code, and its surface is two circles of color blocks distributed in a certain order; the inner circle is three 120° RGB sectors distributed in sequence; the outer circle is a 30° RGB sector with the inner circle red color block as the starting point, and is arranged clockwise in the order of red, green, blue, green, blue, red, blue, green, red, green, red, and blue.

[0009] The wavelength band of the laser light source is visible light or infrared wavelength band, the high-speed CCD camera can receive the visible light wavelength band or extend the visible light to the infrared wavelength band, and the wavelength band of the high-speed CCD camera matches the wavelength band of the laser light source.

[0010] The measuring method of the high-resolution angular displacement absolute measuring device based on color-coded imaging comprises the following steps:

[0011] Step 1: Build the measuring device;

[0012] The bearing of the stepper motor is connected to the center of the circular code disk, and the two perform synchronous rotation. The laser light source and the high-speed CCD camera are placed in front of the circular code disk, wherein the high-speed CCD camera is coaxially placed with the stepper motor and the center of the circular code disk, and connected to the PC end. The laser light source is aligned with the outermost color block of the circular code disk, and the emitted light is vertically incident on the outer color block of the circular code disk, forming a laser circular spot, which is used as the target detection point.

[0013] Step 2: Image preprocessing;

[0014] A high-speed CCD camera is set in front of the circular code disk to take pictures of it, and the images of the circular code disk with target detection points are uploaded to the PC side for frame-by-frame processing. First, image preprocessing is performed to reduce background interference information. The image preprocessing removes the useless noise in the background of the code disk image with target detection points by using the adaptive median filtering method. Then, the filtered image is binarized, and the threshold is selected by using the iterative method to realize the binary segmentation of the image. Finally, the speckle noise in the binary-processed image with target detection points is removed.

[0015] Step 3: Detection of the center coordinates of the target point;

[0016] Measure the rotation angle, that is, it is converted to求取 the rotation angle of the target detection point, which is the detection of the rotational angular displacement of the target detection point under the dynamic rotation background of the circular code disk. First, use the Hough transform method to determine the center of the light source target detection point. Identify the circle from the edge information of the image and estimate its center and radius. Use the Hough gradient method to determine the center coordinates of the target detection point.

[0017] Step 4: Solve the rough code angle;

[0018] After determining the center coordinates of the target detection point, track the target under the dynamic background to obtain the color information of the color block where the detection point is located on the outer circle of the circular code disk at this time. Then, explore from the detection point to the centroid position of the circular code disk to obtain the color information of the corresponding inner circle color block. Then, fit a circle with the line connecting the centroid of the circular code disk and the center of the detection point as the radius as the exploration trajectory, and the detection point explores to both sides of the outer circle color block of the circular code disk to obtain the color information of the two adjacent sides of the detection point. Finally, determine the absolute position through the color information of the outer circle color block, inner circle color block, and adjacent two-side color blocks where the detection point is located. At this time, the absolute angle obtained by looking up the coding table is the rough code angle;

[0019] The calculation formula for the rough code angle θ1 is:

[0020] θ1 = n × 30° (one)

[0021] where n is the number of color blocks passed by the detection point;

[0022] Step 5: Solve the fine code angle and perform angle subdivision;

[0023] Calculate the angular displacement of the device rotation according to the included angle between the coordinates of the center of each target point and the rotation center coordinates. Use the pixel information between two frames collected by the camera for the target detection point, and during the working process of the device rotating and exploring and tracking the initial and final trajectories of the target detection point, use the frame rate number l between the initial and final positions and the image information of each frame collected by the camera to calculate and obtain the fine code angle θ2;

[0024] The calculation formula for the fine code angle θ2 is:

[0025] θ2=l*Ψ (II)

[0026] Ψ=δ / f (Three)

[0027] Among them, Ψ is the resolution angle between two frames of the high-speed CCD camera, δ is the CCD pixel size, and f is the focal length of the CCD lens;

[0028] By using mathematical methods such as interpolation and fitting to perform sub-pixel segmentation on the collected image information, the segmentation angle θ3 can be further obtained;

[0029] Step 6: Solve the final angle;

[0030] The final angle θ obtained by combining steps 4 and 5 is output as:

[0031] θ=θ1+θ2+θ3(four)

[0032] Step 7: Angle correction;

[0033] When the system rotates, the ideal motion path of the center coordinates of the target detection point should be a standard circular trajectory. The ellipse fitting technology based on the least squares method is used to perform least squares fitting on each target detection point and coordinates calculated using the actual collected image information. By mapping discrete points one by one with the standard circular trajectory, the trajectory of each angle point can be corrected.

[0034] The beneficial effects of the present invention are:

[0035] 1. The present invention relates to a circular code disk based on double-circle RGB three-variable color coding, which uses 3D printing technology to achieve low-cost production of the code disk, and combines machine vision technology with a high-speed CCD camera to achieve absolute measurement of visual color-coded angular displacement, breaking through the current situation where traditional image-based grating angular displacement measurement methods are limited to single-circle, 0-1 coding.

[0036] 2. The structure of the present invention breaks through the limitation of the close placement of the image sensor and the code disk of the traditional image-type grating angular displacement measurement device. The laser light source is set to align with the color-coded color block on the outermost circle of the circular code disk. The high-speed CCD camera is aligned with the center of the code disk and the center axis of the shaft system and is set directly opposite the code disk. The circular laser spot formed by the laser irradiation on the code disk is used as the target detection point to convert the measurement of the rotational angular displacement into the tracking measurement of the target detection point under the dynamic background. The laser light source can be a visible light band or an infrared band, and the high-speed camera can be matched with the band expansion of its light source. The design of this non-contact measurement device does not have complex optical path debugging and has strong flexibility, robustness and fault tolerance. When the system is working, factors such as deflection and vibration have little effect on the camera acquiring the code disk image, and the optical structure of the system is simple and easy to install and adjust.

[0037] 3. The present invention relates to a high-resolution absolute angular displacement measurement method based on color-coded imaging, which breaks through the angle demodulation idea of ​​traditional image-based grating angular displacement measurement that is limited to encoding and decoding, and converts it into a method of realizing angle demodulation by combining the tracking control of the light spot formed by a laser point source, thus avoiding the complexity of optical and mechanical alignment of traditional devices.

[0038] 4. The present invention solves the center coordinates of the target detection point, uses it to explore the left and right sides of the detection point and the center direction of the circular code disk, obtains the color block information, and determines the rough rotation angle by querying the coding table; then, the camera collects the pixel information of two frames of images to calculate the precise rotation angle, and uses the sub-pixel subdivision method to subdivide, and further obtains the subdivision angle information with higher resolution; finally, the rotation trajectory of the target detection point is fitted and mapped with the standard circular trajectory, and the measured angle information is corrected. Compared with the traditional angular displacement measurement method, this measurement method innovatively designs the structure between the laser light source and the code disk, and designs an angle demodulation algorithm based on the recognition and tracking of the target detection point image. This method has the advantages of higher precision and high resolution measurement.

[0039] 5. The structure and manufacturing process of the present invention are simple. Not only is the measurement principle simpler than that of traditional angular displacement, but the device is also simpler, the installation and adjustment are more convenient, and the visual effect is excellent. It solves the measurement problem that there is currently no visual absolute angular displacement encoding device based on color coding imaging in China. The device is more suitable for life-level application scenarios such as electronic roller rangefinders and image measurement. BRIEF DESCRIPTION OF THE DRAWINGS

[0040] Figure 1 It is a schematic diagram of a high-resolution absolute angular displacement measurement device based on color-coded imaging of the present invention;

[0041] Figure 2 is a schematic diagram of a circular code disk in the present invention;

[0042] Figure 3 It is a schematic diagram of the demodulation of the code disk image type angular displacement signal in the present invention;

[0043] Figure 4 It is a schematic diagram of the code disk image denoising preprocessing process in the present invention;

[0044] Figure 5 It is a schematic diagram of solving the center coordinates of the target detection point based on the Hough gradient method in the present invention;

[0045] Figure 6 It is a schematic diagram of angle accurate demodulation calculation in the present invention;

[0046] Figure 7 It is a schematic diagram of angle solution and correction in the present invention. DETAILED DESCRIPTION

[0047] The present invention is further described in detail below with reference to the accompanying drawings.

[0048] like Figure 1 As shown, a high-resolution angular displacement absolute measurement device based on color-coded imaging is composed of a stepper motor 1, a circular code disk 2, a laser light source 3, a high-speed CCD camera 4 and a PC terminal 5.

[0049] The stepper motor 1 is connected to the back center of the circular code disk 2, and the rotation of the bearing of the stepper motor 1 drives the rotation of the circular code disk 2. The laser light source 3 and the high-speed CCD camera 4 are placed in front of the circular code disk 2, wherein the high-speed CCD camera 4 is coaxially placed with the stepper motor 1 and the center of the circular code disk 2, and connected to the PC terminal 5; the laser light source 3 is aligned with the outermost color block of the circular code disk, and emits a laser vertically incident on the outer RGB color block of the code disk on the front of the circular code disk 2, forming a laser circular spot, which is used as the target detection point; when the stepper motor 1 rotates, the high-speed CCD camera 4 takes real-time photos of the circular code disk 2 to obtain the image signal of the code disk with the target detection point; the high-speed CCD camera 4 takes real-time photos and saves the image signal and uploads it to the PC terminal 5, and the PC terminal 5 performs specific demodulation of the rotation angle information.

[0050] like Figure 2 As shown, the circular code disk 2 is arranged according to the matrix code, and its surface is two circles of color blocks distributed in a certain order. The inner circle is three 120° RGB (red, green, blue) sectors distributed in sequence. The outer circle is a 30° RGB (red, green, blue) sector with the inner circle red block as the starting point, and is arranged clockwise in the order of red, green, blue, green, blue, red, blue, green, red, green, red, blue, and the color coding table is shown in Table 1. Use new materials for 3D printing to make low-cost code disks.

[0051] Table 1 Color coding table

[0052]

[0053] The laser light source 3 vertically irradiates the color block on the outer ring of the circular code disk 2 to form a dot-shaped laser spot D.

[0054] The high-speed CCD camera 4 can receive wavelengths in both the visible light and infrared wavelengths.

[0055] The laser light source 3 may have a wavelength in the visible light or infrared wavelength, and can be matched with the wavelength of the high-speed CCD camera 4 .

[0056] When the device is working, the high-speed CCD camera 4 shoots the circular code disk 2 to obtain the code disk image signal with the target detection point, and transmits the image signal to the PC 5; firstly, the image filtering preprocessing is performed to enhance the code disk image quality, and then the center coordinates of the target point on the color code disk are solved by the Hough gradient method; the coarse code angle of the current detection point rotation is obtained by the background color block information where the target point is located; the frame rate value before and after the camera sampling is used, combined with the center coordinates of the target point to solve the precise angle; finally, the subdivision is performed through the sub-pixel subdivision algorithm, and the actual angle information is calculated and output; finally, the target point fitting trajectory is mapped with the standard circle trajectory to correct the demodulated angle information.

[0057] The high-resolution absolute measurement method of angular displacement based on color-coded imaging is as follows:

[0058] Step 1: Build the measuring device;

[0059] The bearing of the stepper motor 1 is connected to the center of the circular code disk 2, and the two perform synchronous rotation. The laser light source 3 and the high-speed CCD camera 4 are placed in front of the circular code disk 2, wherein the high-speed CCD camera 4 is coaxially placed with the stepper motor 1 and the center of the circular code disk 2, and the high-speed CCD camera 4 is connected to the PC terminal 5 via a USB data cable. The laser light source 3 is placed at the outermost color block of the circular code disk, and the emitted light is vertically incident on the outer color block of the circular code disk 2, and the incident point is Figure 2 Go to point D.

[0060] Step 2: Image preprocessing;

[0061] A high-speed CCD camera 4 is arranged in front of the circular code disk 2 to shoot it, and the shot image of the circular code disk 2 with the target detection point D is uploaded to the PC terminal 5, and processed frame by frame at the PC terminal 5.

[0062] like Figure 3 As shown, the overall processing flow of the PC terminal 5 is to first perform image preprocessing to remove the background, code disk material particle noise, etc.; then detect the center coordinates of the target point, and then use the images between adjacent sampling frame rates, combined with the center coordinates of the target point and the center coordinates of the code disk, to solve the rotation angle information; finally, use the fitted target detection point rotation trajectory and the standard circle trajectory to map and correct the angle.

[0063] like Figure 4As shown in the figure, the specific image preprocessing process is to remove the noise of the code disk image with target detection points by using the adaptive median filtering method, then binarize the filtered image, and use the iterative method to select the threshold to achieve the binary segmentation of the image; then, remove the speckle noise from the binarized image, and analyze the total number of pixels contained in each area, that is: when the number of pixels corresponding to a certain area does not reach the preset threshold, it can be reasonably inferred that the area is potential spots and is filled.

[0064] Step 3: Detect the center coordinates of the target point;

[0065] After the image preprocessing operation, in order to measure the angle of target movement, that is, it is necessary to measure the rotation angle of the target detection point under the dynamic background, that is, the rotation angle of the laser incident point D. The design uses the Hough gradient method to determine the center of the light source detection point D.

[0066] like Figure 5 As shown in the figure, firstly, the Canny algorithm is used to extract edge information from the original image to obtain a clear binary edge image. On this basis, the Sobel differential operator is used to calculate the local gradient value and direction θ of each pixel point of the original image.

[0067] For each edge point (x, y), estimate the possible center position (Xc, Yc) based on the gradient direction of the point:

[0068]

[0069] Where r is the radius of the circle and θ is the gradient direction.

[0070] Next, systematically traverse each valid (non-zero) pixel position in the obtained Canny edge binary image, and extend the straight line segment according to the gradient direction at these pixel points (i.e., the orthogonal direction of the tangent direction). For the estimated center position (Xc, Yc), the count value of the corresponding center position in the accumulator is incremented by 1. Count the possible center positions (Xc, Yc), and finally output the center (Xc, Yc) with the largest number in the accumulator, which is the center coordinate of the light source detection point D.

[0071] Step 4, solving the coarse code angle;

[0072] After obtaining the central coordinates of the light source detection point D, it is necessary to track the target under a dynamic background, obtain the color information of the color block where the detection point is located on the outer circle of the circular code disk at this time, and then explore from the detection point to the centroid position of the code disk to obtain the corresponding color information of the inner circle color block. Then, the target detection point D takes the circle with the line connecting the centroid of the circular code disk 2 and the center of the detection point as the radius as the exploration trajectory, and explores towards both sides of the color block on the outer circle of the code disk to obtain the color information of the color blocks on both sides of the detection point. Finally, by using the color of the outer circle color block where the detection point is located, the color of the inner circle color block, and the color of the outer circle color block adjacent to the detection point, the rough angle information θ1 of the absolute position is determined by looking up the color code table 1 of the code disk.

[0073] The angle obtained by looking up the code table is the rough code angle, and the rough code angle formula is:

[0074] θ1 = n × 30° (1)

[0075] where n is the number of color blocks passed by the detection point on the outer circle of the circular code disk.

[0076] Step 5: Solve the fine code angle and perform angle subdivision;

[0077] Calculate the angular displacement of the device rotation according to the included angle between the coordinates of the center of each target point and the coordinates of the rotation center; use the pixel information between two frames collected by the camera for the target detection point, and during the working process of the device rotating to explore and track the initial and final trajectories of the target detection point, use the frame rate number l between the initial and final positions and the image information of each frame collected by the camera to calculate and obtain the fine code angle θ2.

[0078] As Figure 6 shown, taking the target detection point turning from point S to point F as an example, with the center O of the code disk as the origin, the arc KG as the rotation trajectory of the target detection point D, and OA, OB, OC, OD representing the scale lines of the code disk for each frame of the camera, the angle between the two points can be obtained through the frame rate number l between the initial and final points S and F, that is, the fine code angle θ2;

[0079] The calculation formula for the fine code angle θ2 is:

[0080] θ2 = l * Ψ (2)

[0081] Ψ = δ / f (3)

[0082] where Ψ is the resolution angle between two frames of the high-speed CCD camera, δ is the CCD pixel size, and f is the CCD lens focal length.

[0083] Using mathematical means such as interpolation and fitting to perform sub-pixel subdivision on the collected image information, the subdivision angle θ3 can be further obtained.

[0084] Step 6: Solve the final angle;

[0085] Combine the final angle θ obtained in steps 4 and 5 and output it.

[0086] θ=θ1+θ2+θ3(four)

[0087] Step 7: Angle correction;

[0088] When the system rotates, the ideal motion path of the center coordinates of the target detection point should be a standard circular trajectory. However, in practical applications, due to the influence of equipment installation deviation and image noise, the actual trajectory obtained is an ellipse. Therefore, this device adopts the ellipse fitting technology based on the least squares method to perform least squares fitting on each target detection point and coordinates calculated using the actual collected image information. By mapping discrete points one by one with the standard circular trajectory, the trajectory of each angle point is corrected, thereby improving the angle measurement accuracy.

[0089] Based on the above measurement method, a specific example is provided: the designed code disk diameter is 50mm, the sensing structure thickness is 5mm; to facilitate human observation, a helium-neon laser is used as the light source with a wavelength of 632.8nm; the high-speed CCD camera used has a frame rate of 120 frames per second. Figure 7 As shown, taking point M moving clockwise to point E as an example, after preprocessing operations such as filtering, binarization, and spot removal, the coordinates of the center point are obtained by the Hough gradient method as (45.5663, 42.4213); after dynamic detection, the color of the outer circle block where point E is located is blue, and point E explores the position of the centroid of the code disk, and the inner circle color block is blue. Then, with the fitting circle with OE as the radius as the exploration trajectory, point E explores the adjacent color blocks on both sides, and obtains red and red respectively. The coarse code angle θ1=1×30°=30° is determined by looking up the coding table;

[0090] A and B are the motion trajectory points of the edge line of the blue color block respectively. During the movement from point A to point E, the frame rate of the high-speed CCD camera is l=80, and the resolution angle Ψ of the high-speed CCD camera between two frames is 0.23°. It can be seen that the precise angle θ2=80×0.23°=18.4°; further through the sub-pixel subdivision method, the subdivision angle θ3=50" is obtained, and the demodulated angle θ=θ1+θ2+θ3=30°+18.4°+50"=48°24′50".

[0091] Finally, the ellipse fitting technology based on the least squares method is adopted to fit the 64 target detection points and coordinates actually collected. By mapping the actual ellipse trajectory to the theoretical circular trajectory where the green spot is located, the trajectory of each angle point is corrected. By correcting point E to point F, the final rotation angle is 48°24′47".

Claims

1. A high-resolution absolute angular displacement measurement device based on color-coded imaging, characterized in that: The device is composed of a stepping motor (1), a circular code disk (2), a laser light source (3), a high-speed CCD camera (4) and a PC terminal (5); The stepper motor (1) is connected to the back center of the circular code disk (2), and the rotation of the bearing of the stepper motor (1) drives the rotation of the circular code disk (2); the laser light source (3) and the high-speed CCD camera (4) are placed in front of the circular code disk (2), wherein the high-speed CCD camera (4) is coaxially placed with the stepper motor (1) and the center of the circular code disk (2), and is connected to the PC end (5); the laser light source (3) emits laser light vertically incident on the RGB color block on the outer ring of the circular code disk (2) on the front of the circular code disk (2), forming a laser circular spot, which serves as a target detection point; when the stepper motor (1) rotates, the high-speed CCD camera (4) takes real-time photos of the circular code disk (2) to obtain a code disk image signal with a target detection point; the high-speed CCD camera (4) takes real-time photos and saves the image signal and uploads it to the PC end (5), and the PC end (5) performs specific demodulation of the rotation angle information.

2. The high-resolution absolute angular displacement measurement device based on color-coded imaging according to claim 1, characterized in that: The circular code disk (2) is arranged according to the matrix code, and its surface is composed of two circles of color blocks distributed in a certain order; the inner circle is composed of three 120° RGB sectors distributed in sequence; the outer circle is composed of 30° RGB sectors, with the inner circle red color block as the starting point, and arranged clockwise in the order of red, green, blue, green, blue, red, blue, green, red, green, red, blue.

3. The high-resolution absolute angular displacement measurement device based on color-coded imaging according to claim 1, characterized in that: The wavelength band of the laser light source (3) is the visible light band or the infrared band, the high-speed CCD camera (4) can receive the visible light band or extend the visible light band to the infrared band, and the wavelength band of the high-speed CCD camera (4) matches the wavelength band of the laser light source (3).

4. The measurement method of the high-resolution angular displacement absolute measurement device based on color-coded imaging according to claim 1 is characterized in that: The method comprises the following steps: Step 1: Build the measuring device; The bearing of the stepper motor (1) is connected to the center of the circular code disk (2), and the two perform synchronous rotation. The laser light source (3) and the high-speed CCD camera (4) are placed in front of the circular code disk (2), wherein the high-speed CCD camera (4) is coaxially placed with the stepper motor (1) and the center of the circular code disk (2), and is connected to the PC end (5). The laser light source (3) is placed to align with the outermost color block of the circular code disk, and the emitted light is vertically incident on the outer color block of the circular code disk (2), forming a laser circular spot, which is used as the target detection point. Step 2: Image preprocessing; A high-speed CCD camera (4) is arranged in front of a circular code disk (2) to shoot the circular code disk (2), and the image of the circular code disk (2) with the target detection point is uploaded to a PC (5), and processed frame by frame on the PC (5); first, image preprocessing is performed to reduce background interference information; the image preprocessing uses an adaptive median filtering method to remove useless noise from the background of the code disk image with the target detection point; then the filtered image is binarized, and a threshold is selected by an iterative method to achieve image binarization segmentation; finally, the binarized image with the target detection point is subjected to spot noise removal; Step 3: Detect the center coordinates of the target point; Measure the rotation angle, which is converted to obtaining the angle moved by the target detection point, that is, the detection of the rotational angular displacement of the target detection point under the dynamic rotation of the circular code disk (2); First, use the Hough transform method to determine the center of the light source target detection point; Identify a circle from the edge information of the image and estimate its center and radius; Use the Hough gradient method to determine the center coordinates of the target detection point; Step 4: Solve the coarse code angle; After determining the center coordinates of the target detection point, track the target under the dynamic background to obtain the color information of the color block where the detection point is located on the outer circle of the circular code disk (2) at this time; Then explore from the detection point to the centroid position of the circular code disk (2) to obtain the color information of the corresponding inner circle color block; Then, fit a circle with the line connecting the centroid of the circular code disk (2) and the center of the detection point as the radius as the exploration trajectory, and the detection point explores to both sides of the color block on the outer circle of the circular code disk (2) to obtain the color information of the two adjacent sides of the detection point; Finally, determine the absolute position through the color information of the outer circle color block, inner circle color block, and two adjacent side color blocks where the detection point is located. At this time, the absolute angle obtained by looking up the coding table is the coarse code angle; The calculation formula for the coarse code angle θ1 is: θ1 = n × 30° (One) Where, n is the number of color blocks passed by the detection point; Step 5: Solve the fine code angle and perform angle subdivision; Calculate the angular displacement of the device rotation based on the included angle between the coordinates of the center of each target point and the rotation center coordinates; Use the pixel information between two frames collected by the camera for the target detection point, and during the working process of the device rotating to explore and track the initial and final trajectories of the target detection point, use the frame rate number l between the initial and final positions and the image information of each frame collected by the camera to calculate and obtain the fine code angle θ2; The calculation formula for the fine code angle θ2 is: θ2 = l * Ψ (Two) Ψ = δ / f (Three) Where, Ψ is the resolution angle between two frames of the high-speed CCD camera, δ is the CCD pixel size, and f is the CCD lens focal length; Use mathematical means such as interpolation and fitting to perform sub-pixel subdivision on the collected image information, and the subdivision angle θ3 can be further obtained; Step 6: Solve the final angle; The final angle θ obtained by combining Steps 4 and 5 is output as: θ = θ1 + θ2 + θ3 (Four) Step 7: Angle correction; When the system rotates and works, the ideal movement path of the center coordinates of the target detection point should be a standard circular trajectory. Adopt the ellipse fitting technology based on the least squares method to perform least squares fitting on each target detection point and coordinates calculated using the actual collected image information. Through the one-to-one mapping of discrete points with the standard circular trajectory, the correction of the trajectory of each angle point is realized.

Images