Surface image acquisition method and surface image acquisition device for conical component

Abstract

The invention discloses a surface image acquisition method and a surface image acquisition device for a conical component, and the method comprises the steps: carrying out the line scanning of the conical component in rotary motion at a fixed line frequency through a line-scan digital camera in an image acquisition unit in cooperation with a telecentric lens, and obtaining an original image with regular distortion; establishing a coordinate mapping relation between pixel coordinates of the original image and surface geometric coordinates of the conical component according to the fixed line frequency and the rotary motion of the conical component; and correcting the original image according to the coordinate mapping relation to obtain a target image. According to the invention, a rotary scanning image with a constant amplification factor and no perspective distortion can be obtained, pixel-level inverse mapping correction based on the cone geometric model is realized, the scheme is easy to realize, low in cost and high in reliability, and submillimeter-level precision detection can be realized.

Description

Technical Field

[0001] This application relates to the field of surface defect detection technology, specifically to a surface image acquisition method and surface image acquisition device for tapered components. Background Technology

[0002] Conical or frustum-shaped components are commonly used in aerospace, energy, and power industries. For example, components such as the intake pipe, exhaust nozzle, and combustion chamber liner of turbine engines typically employ conical or frustum-shaped structures to achieve optimal flow rectification. These conical components are generally made of alloys or advanced composite materials and have extremely high requirements for online reliability. However, due to their unique physical, chemical, and mechanical properties, their surfaces are prone to various typical defects during manufacturing. Examples include scratches, cracks, and pits on metal surfaces, and bubbles, detachment, inclusions, wrinkles, and scratches on carbon fiber composite surfaces. These defects can have serious consequences in practical applications. Therefore, surface defect inspection of conical components is necessary before shipment. Currently, surface defect inspection of these conical components is generally performed manually by visual inspection, which suffers from low efficiency, worker fatigue, high rates of missed detections, and high rates of false detections.

[0003] Some manufacturers also use image recognition solutions for automated inspection. This solution requires first acquiring images of the conical surface of the conical component to obtain the image to be inspected. Currently, area scan cameras are generally used to inspect conical components. The area scan camera takes pictures from multiple different angles to obtain an image sequence covering the entire conical surface. This method has low detection efficiency, large data volume, and is difficult to stitch together with high accuracy. If a line scan camera is used to acquire images of the conical component, the linear velocity of the generatrix on the conical surface varies at different positions during the rotation of the conical component, with the larger end moving faster and the smaller end moving slower. This results in the acquired image having the problem of the larger end (large radius, high linear velocity) being stretched and the smaller end (small radius, low linear velocity) being compressed, causing significant image distortion. To address this, existing solutions use complex mechanical devices (such as synchronous speed change mechanisms) or software (dynamically adjusting the line frequency) to try to eliminate image distortion during the image acquisition process of the line scan camera. However, in practical applications, such systems are complex to design, costly, and have low reliability. Summary of the Invention

[0004] This application aims to address one of the technical problems in related technologies to a certain extent. To this end, this application provides a method and apparatus for acquiring surface images of conical components.

[0005] To achieve the above objectives, this application adopts the following technical solution: a surface image acquisition method for a tapered component, the surface image acquisition method comprising:

[0006] The linear scan camera in the image acquisition unit, in conjunction with the telecentric lens, performs line scanning on the rotating conical component at a fixed line frequency to acquire the original image with regular distortion.

[0007] Based on the fixed line frequency and the rotational motion of the conical component, a coordinate mapping relationship is established between the pixel coordinates of the original image and the surface geometric coordinates of the conical component;

[0008] The original image is corrected according to the coordinate mapping relationship to obtain the target image.

[0009] The application of this application has the following beneficial effects: By using a linear scan camera in the image acquisition unit in conjunction with a telecentric lens to perform line scanning on a rotating conical component at a fixed line frequency, a rotational scan image with constant magnification and no perspective distortion can be obtained. That is, although the acquired original image still has distortion problems, its distortion is regular, which facilitates subsequent algorithm correction. By establishing a coordinate mapping relationship and correcting the acquired original image according to the coordinate mapping relationship, pixel-level inverse mapping correction based on the conical geometric model can be achieved to obtain the target image, which can be used in image recognition schemes for surface defect detection. Compared with the existing technology that uses mechanical devices (such as synchronous speed change mechanisms) or software (dynamically adjusting the line frequency) to eliminate image distortion during the image acquisition process of the linear scan camera, the solution of this application is easier to implement, lower in cost, and more reliable, and can achieve sub-millimeter level precision detection.

[0010] Optionally, the fixed line frequency is set based on the linear velocity of the outer edge of the large end of the conical member during the rotational motion of the conical member.

[0011] Optionally, the fixed line frequency F satisfies the following formula:

[0012] ;

[0013] Where ω is the rotational angular velocity of the conical component, L is the generatrix length of the conical component, α is the half-apex angle of the conical component, and σ is the object space resolution of the image acquisition unit.

[0014] Optionally, the line scan camera is configured to continuously acquire N rows of images at a fixed line frequency to form the original image, wherein the row index u = 0, 1, 2, ..., N-1 and the column index v = 0, 1, 2, ..., P-1 in the pixel coordinates (u, v) of the original image; where P is the resolution of the line scan camera.

[0015] The coordinate mapping relationship is configured as follows:

[0016] ;

[0017] ;

[0018] in, Let θ be the rotation angle corresponding to the u-th row of the original image, and ω be the rotational angular velocity of the conical component. The acquisition time from the original image to the u-th row of the image;

[0019] The relative position of a pixel in the v-th column of the original image projected onto the generatrix of the tapered component. Let σ be the coordinate component of the pixel in the v-th column of the original image along the axial direction of the conical component, σ be the object space resolution of the image acquisition unit, L be the generatrix length of the conical component, and α be the half-apex angle of the conical component.

[0020] Optionally, the correction of the original image based on the coordinate mapping relationship includes:

[0021] The scaling factor λ is determined based on the coordinate mapping relationship. The scaling factor λ is used to characterize the local geometric scaling ratio of each row of pixels distributed along the axial direction of the conical component in the original image due to the difference in linear velocity during circumferential imaging.

[0022] A rectangular unfolding strategy based on arc length-generic coordinates is used to establish the correspondence between the coordinate system (l, s) of the target image and the coordinate system (θ, s) of the original image;

[0023] Based on the correspondence and the representations of the pixel coordinates (u, v) of the original image and the pixel coordinates (i, j) of the target image in their respective coordinate systems (θ, s) and (l, s), the coordinate transformation between the original image and the target image is completed, and the target image is generated.

[0024] Optionally, define the normalized bus position s∈[S min [,1], where s= S min For the cone apex s=1 corresponding to the large end, the scaling factor λ satisfies the following formula:

[0025] ;

[0026] Where s is the normalized bus position, Let be the scaling factor at position s on the generatrix of the tapered member. Let V(s) be the scaling factor at the reference datum, and V(s) be the linear velocity at position s on the generatrix of the tapered member. Let be the linear velocity of the outer edge of the large end of the conical component during its rotational motion. Let ω be the normalized position parameter value of the large end of the conical component, ω be the rotational angular velocity of the conical component, L be the generatrix length of the conical component, and α be the half-apex angle of the conical component. and All are 1.

[0027] Optionally, the correspondence is as follows:

[0028] ;

[0029] Where θ is the rotation angle, x(s) is the x-axis coordinate of a point on the generatrix, and α is the half-apex angle of the conical component;

[0030] Optionally, the size of the target image is set to H*W, where H and W are determined according to the pixel precision requirements of the target image, H is the number of rows, and W is the number of columns;

[0031] The relationship between the pixel coordinates (i, j) of the target image and the coordinate system (l, s) is expressed as follows:

[0032] ;

[0033] ;

[0034] Where i = 0, 1, 2, ..., H-1; j = 0, 1, 2, ..., W-1, representing the scaling factor at different positions in each row direction. L is the generatrix length of the conical member, and α is the semi-apex angle of the conical member;

[0035] Arc length range Normalized busbar position .

[0036] Optionally, generating the target image based on the representation of the pixel coordinates (i, j) of the target image in coordinate system (l, s) includes:

[0037] The pixel coordinates of each target image are traversed using an inverse mapping method. The pixel coordinates (i, j) of the target image are calculated using the inverse transformation formula and then inversely mapped to the floating-point pixel coordinates of the original image. ;

[0038] The inverse transformation formula can be obtained by solving the following relation:

[0039] ;

[0040] ;

[0041] Then the coordinates are inversely mapped to the original image. It can be represented as:

[0042] ;

[0043] ;

[0044] Based on the principle of region interpolation, each pixel floating-point coordinate In the original image, for a region of a certain size, calculate the weighted average of all pixels in that region to obtain the interpolated pixel value;

[0045] The interpolated pixel values ​​are assigned to the corresponding pixel (i, j) in the target image to generate the target image.

[0046] Furthermore, this application also provides a surface image acquisition device for a tapered component, the surface image acquisition device comprising:

[0047] A rotary drive unit is used to drive a tapered component to rotate about its own axis.

[0048] The image acquisition unit includes a line scan camera and a telecentric lens. The line scan camera is used in conjunction with the telecentric lens to perform line scanning on a rotating conical component at a fixed line frequency to acquire an original image with regular distortion.

[0049] The data processing unit is used to establish a coordinate mapping relationship between the pixel coordinates of the original image and the pixel coordinates of the surface unfolded image of the conical component based on the fixed line frequency and the rotational motion of the conical component, and to correct the original image based on the coordinate mapping relationship and the transformation algorithm to obtain the target image.

[0050] The reasoning process for the beneficial effects of the surface image acquisition device provided in this application are similar to those of the aforementioned surface image acquisition method, and will not be repeated here.

[0051] These features and advantages of this application will be disclosed in detail in the following specific embodiments and accompanying drawings. The best embodiments or means of this application will be shown in detail in conjunction with the accompanying drawings, but are not intended to limit the technical solutions of this application. In addition, each of these features, elements and components appearing in the following text and drawings is multiple and is labeled with different symbols or numbers for convenience, but all represent parts with the same or similar structure or function. Attached Figure Description

[0052] The following description, in conjunction with the accompanying drawings, further illustrates this application:

[0053] Figure 1 A flowchart of a surface image acquisition method for a tapered component is provided for the first aspect of this application;

[0054] Figure 2A schematic diagram of a surface image acquisition device for a tapered component is provided for the second aspect of this application;

[0055] Figure 3 A schematic diagram illustrating the determination of a three-dimensional spatial coordinate system based on a conical component;

[0056] Figure 4 A schematic diagram for determining a three-dimensional spatial coordinate system based on a frustum-shaped conical component;

[0057] Figure 5(a) is a schematic diagram of the rectangular unfolding of the conical surface (with a checkerboard pattern) of the conical component in an ideal state; Figure 5(b) is a schematic diagram of the original image acquired based on the conical component; Figure 5(c) is a schematic diagram of the target image finally obtained using the surface image acquisition method provided in this application.

[0058] The components include: 1. Rotation drive unit; 2. Image acquisition unit; 20. Linear array camera; 21. Telecentric lens; 22. Illumination module; 3. Data processing unit; and 4. Conical component. Detailed Implementation

[0059] The embodiments of this application are described in detail below. Examples of these embodiments are shown in the accompanying drawings, wherein the same or similar reference numerals denote the same or similar elements or elements having the same or similar functions throughout. The embodiments described are intended to explain this application and should not be construed as limiting it.

[0060] The terms "an embodiment," "example," or "example" used in this specification refer to a particular feature, structure, or characteristic described in connection with the embodiment itself that may be included in at least one embodiment disclosed in this application. The phrase "in an embodiment" appearing in various places throughout the specification does not necessarily refer to the same embodiment.

[0061] In the description of this application, it should be understood that the terms "upper," "lower," "front," "rear," "vertical," "horizontal," "top," "bottom," "inner," and "outer," etc., indicate the orientation or positional relationship based on the orientation or positional relationship shown in the accompanying drawings, and are only for the convenience of describing this application 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, and therefore should not be construed as a limitation on this application. In the description of this application, "a plurality of" means two or more, unless otherwise precisely specified.

[0062] In the description of this application, it should be noted that, unless otherwise expressly specified and limited, the terms "connected," "linked," and "connected" should be interpreted broadly. For example, they can refer to a fixed connection, a connection through an intermediary, or a connection within two elements or an interaction between two elements. Those skilled in the art can understand the specific meaning of the above terms in this application according to the specific circumstances.

[0063] like Figure 1 As shown, the first aspect of this application provides a method for acquiring surface images of a tapered member 4, such as... Figure 2 As shown, a second aspect of this application provides a surface image acquisition device for a tapered member 4, and the surface image acquisition method can be implemented using this surface image acquisition device. The surface image acquisition method includes the following steps:

[0064] S100: The linear scan camera 20 in the image acquisition unit 2, in conjunction with the telecentric lens 21, performs a line scan on the rotating conical component 4 at a fixed line frequency to acquire an original image with regular distortion.

[0065] S200: Establish the coordinate mapping relationship between the pixel coordinates of the original image and the surface geometric coordinates of the cone component 4 based on the fixed line frequency and the rotational motion of the cone component 4;

[0066] S300: Correct the original image according to the coordinate mapping relationship to obtain the target image.

[0067] The linear scan camera 20 in image acquisition unit 2, in conjunction with the telecentric lens 21, performs a line scan of the rotating conical component 4 at a fixed line frequency. This allows for the acquisition of a rotating scan image with constant magnification and no perspective distortion. In other words, while the acquired original image still exhibits distortion, the distortion follows a regular pattern, facilitating subsequent algorithmic correction. A coordinate mapping relationship is established, and the acquired original image is corrected based on this relationship, achieving pixel-level inverse mapping correction based on the conical geometric model. This results in the target image, which can be used in image recognition schemes for surface defect detection.

[0068] The surface image acquisition method provided in this application abandons the existing technology of eliminating image distortion during the image acquisition process of the line array camera 20 by using mechanical devices (such as synchronous speed change mechanisms) or software (dynamically adjusting line frequency). Instead, it actively acquires distortion with regularity (constant magnification, no perspective distortion), and then uses the established coordinate mapping relationship to correct the original image to obtain the desired target image. Compared with the existing technology, it eliminates the need for complex mechanical devices (such as synchronous speed change mechanisms) or software (dynamically adjusting line frequency), making it easier to implement, with lower cost and higher reliability, and enabling sub-millimeter level precision detection.

[0069] The second aspect of this application provides a surface image acquisition device comprising a rotation drive unit 1, an image acquisition unit 2, and a data processing unit 3. The rotation drive unit 1 drives a conical member 4 to rotate around its own axis. The image acquisition unit 2 includes a line scan camera 20 and a telecentric lens 21. The line scan camera 20, in conjunction with the telecentric lens 21, performs line scanning on the rotating conical member 4 at a fixed line frequency to acquire an original image with regular distortion. The data processing unit 3 establishes a coordinate mapping relationship between the pixel coordinates of the original image and the pixel coordinates of the unfolded surface image of the conical member 4 based on the fixed line frequency and the rotational motion of the conical member 4. It then corrects the original image according to the coordinate mapping relationship and a transformation algorithm to obtain a target image. In other words, this surface image acquisition device can be used to implement the aforementioned surface image acquisition method.

[0070] It should be noted that, as Figure 2 As shown, the image acquisition unit 2 also includes an illumination module 22. The line scan camera 20 and telecentric lens 21 are combined and installed to ensure that the optical axis of the telecentric lens 21 is perpendicular to the axis of the conical component 4, and that the target surface of the image sensor of the line scan camera 20 is parallel to the axis of the conical component 4. The illumination module 22 can use a high-brightness LED line light source to uniformly illuminate the conical surface of the conical component 4 from the side to obtain clear image contrast. The data processing unit 3 is typically an industrial computer or an embedded image processing system, communicatively connected to the rotation drive unit 1 and the image acquisition unit 2, and used to control the entire image acquisition, image transmission, image correction, and image output operations.

[0071] The rotary drive unit 1 can use existing devices. For example, the rotary drive unit 1 may include a high-precision servo motor and an electric gripper. A connecting structure (shaft or block structure) can be set on the large end face of the conical member 4. The driving force of the servo motor can be transmitted to the conical member 4 by using the electric gripper to hold the connecting structure, so as to drive the conical member 4 to rotate around its own axis. Generally, the conical member 4 is driven to rotate at a set speed at a uniform speed.

[0072] In this embodiment, when setting the fixed line frequency, the linear velocity of the outer edge of the large end of the conical member 4 during its rotational motion is used as a reference. The large end of the conical member 4 refers to the end with a relatively large diameter, and conversely, the small end refers to the end with a relatively small diameter. For a conical member 4 with a complete conical surface, the small end refers to the tip of the conical member 4; for a conical member 4 with a truncated conical surface, its overall shape is essentially a frustum, and the small end refers to the end with a relatively small diameter on the frustum-shaped conical member 4. Setting the fixed line frequency in this way ensures the integrity of the acquired original image. Specifically, the principle of image acquisition by the line scan camera 20 is to record the relative motion between the surface of the acquired object and the sensor in the line scan camera 20. If the line frequency is lower than the linear velocity of a point on the acquired object, image information will be lost (undersampling); if the line frequency is higher than the linear velocity of a point on the acquired object, image information will be redundant. By setting a fixed line frequency in the above manner, it can be ensured that the original image obtained by acquiring the conical component 4 will only have a "redundancy" problem (a manifestation of oversampling, but the image information can be ensured to be complete), and there will be no problem of image information loss.

[0073] Specifically, the fixed line frequency F satisfies the following formula:

[0074] ;

[0075] Where ω is the rotational angular velocity of the conical component, L is the generatrix length of the conical component, α is the half-apex angle of the conical component, and σ is the object space resolution of the image acquisition unit. The object space resolution of the image acquisition unit refers to the physical size of a pixel on the image sensor within the line scan camera on the actual surface of the object, within the entire image acquisition unit. The unit of object space resolution is generally millimeters per pixel (mm / pixel) or micrometers per pixel (µm / pixel). Specifically, in this application, the object space resolution is a fixed parameter related to the image sensor, telecentric lens, and working distance in the line scan camera.

[0076] The process of setting a fixed line frequency is further explained as follows: First, the three-dimensional spatial coordinate system is determined based on the image acquisition unit in the surface image detection device, combined with... Figure 3 As shown, the vertex of the conical component is set as the origin O (0,0,0), the Z-axis is the axis of the conical component itself, pointing from the smaller end to the larger end, the X-axis is the optical axis of the telecentric lens, pointing from the telecentric lens to the conical component, and the Y-axis is perpendicular to the plane formed by the X-axis and Z-axis. It should be noted that, combined with... Figure 4 As shown, when the conical component is a frustum with a truncated conical surface, the point where the extended generatrices of the frustum intersect is the origin O (0,0,0).

[0077] It's easy to understand that a conical surface can be considered as being formed by rotating a generatrix around the Z-axis. Let the semi-vertex angle of this generatrix be α (0° < α < 90°), the length of the generatrix be L, and the normalized generatrix position parameter s ∈ [0, 1] be set. When s = 0, it corresponds to the vertex of the conical component; when s = 1, it corresponds to the large end of the conical component. Then, the parametric equation of the conical surface in the XZ plane can be expressed as:

[0078] ;

[0079] ;

[0080] From this, the linear velocity of any point on the generatrix of the conical component can be obtained:

[0081] ;

[0082] In the formula, ω represents the rotational angular velocity of the conical component. For a conical component of a fixed size, given L and α, the linear velocity V(s) at each point on the generatrix of the conical component is proportional to the normalized generatrix position parameter s. That is, the linear velocity increases linearly from the small end to the large end of the conical component. Since the scaling ratio of the image acquired by the line scan camera is directly determined by the ratio of the linear velocity, the distortion of the original image also follows a linear law, which facilitates subsequent correction.

[0083] Furthermore, given that the magnification of the telecentric lens is M, the maximum image plane movement speed is:

[0084] ;

[0085] The minimum row frequency that satisfies the Nyquist sampling theorem is:

[0086] ;

[0087] Since the object space resolution σ = p / M, the above equation can be equivalently expressed as:

[0088] ;

[0089] In summary, the fixed line frequency setting is now complete.

[0090] Furthermore, the line scan camera is set to continuously acquire N rows of images at a fixed line frequency to form an original image. The row index u = 0, 1, 2, ..., N-1 and the column index v = 0, 1, 2, ..., P-1 in the pixel coordinates (u, v) of the original image; where P is the resolution of the line scan camera.

[0091] The coordinate mapping relationship is configured as follows:

[0092] ;

[0093] ;

[0094] in, Let θ be the rotation angle corresponding to the u-th row of the original image, and ω be the rotational angular velocity of the conical component. is the acquisition time from the original image to the u-th row of the image.

[0095] The relative position of the pixel in column v of the original image projected onto the generatrix of the tapered member. Let σ be the coordinate component of the pixel in the v-th column of the original image along the axis of the conical component, σ be the object space resolution of the image acquisition unit, L be the generatrix length of the conical component, and α be the half-apex angle of the conical component.

[0096] Furthermore, the correction of the original image based on coordinate mapping relationships includes:

[0097] The scaling factor λ is determined based on the coordinate mapping relationship. The scaling factor λ is used to characterize the local geometric scaling ratio of each row of pixels distributed along the axial direction of the conical component in the original image due to the difference in linear velocity during circumferential imaging.

[0098] A rectangular unfolding strategy based on arc length-generic coordinates is used to establish the correspondence between the coordinate system (l, s) of the target image and the coordinate system (θ, s) of the original image. It should be noted that although the ordinate of both coordinate systems is represented by the normalized generic position s, when actually generating the image, it is necessary to clarify that the row height of the original image is the projection of the row height of the target image. That is, the cone height obtained by the camera is actually the projection of its generic on the axial direction.

[0099] Based on the correspondence and the representations of the pixel coordinates (u, v) of the original image and the pixel coordinates (i, j) of the target image in their respective coordinate systems (θ, s) and (l, s), the coordinate transformation between the original image and the target image is completed, and the target image is generated.

[0100] Thus, the generation of the target image can be achieved.

[0101] Wherein, the normalized bus position s∈[S] is defined. min , 1], where s=S min For the cone apex, s=1 corresponds to the large end, then the scaling factor λ satisfies the following formula:

[0102] ;

[0103] Where s is the normalized bus position, Let be the scaling factor at position s on the generatrix of the tapered member. Let V(s) be the scaling factor at the reference datum, and V(s) be the linear velocity at position s on the generatrix of the tapered member. Let be the linear velocity of the outer edge of the large end of the conical component during its rotational motion. Let ω be the normalized position parameter value of the large end of the conical component, ω be the rotational angular velocity of the conical component, L be the generatrix length of the conical component, and α be the half-apex angle of the conical component. and All are 1.

[0104] Furthermore, referring to Figures 5(a), 5(b), and 5(c), considering the engineering implementation difficulty and image processing difficulty, as mentioned above, this application adopts a rectangular unfolding strategy based on arc length-generic coordinates to establish the correspondence between the coordinate system (l, s) of the target image and the coordinate system (θ, s) of the original image. The horizontal coordinate of the target image is the arc length l, and the vertical coordinate is the normalized generic position s. The horizontal coordinate l is modulated by the scaling factor λ. In the original image acquired by the linear array camera, the same angular interval Δθ corresponds to the same number of rows, while in the target image, the same angular interval Δθ corresponds to different arc lengths Δl at different normalized generic positions s. That is, in the original image, different arc lengths under different horizontal coordinates (column indices) occupy the same number of rows, resulting in image distortion. Therefore, it is easier to implement by linearly mapping the fan-shaped graphic unfolded from the cone to each row of the rectangle according to the arc length, and then adjusting the horizontal scale based on the scaling factor, so that the angular interval Δθ at each normalized generic position s can correspond to the actual arc length Δl. Intuitively, this involves forcibly stretching a curved fan shape (the actual unfolded shape of the cone surface of a conical component) into a straight rectangle (the shape of the output target image). This straight rectangle contains a valid content area of ​​triangles (when the conical component is a cone) or trapezoids (when the conical component is a frustum). The valid content area refers to the region targeted by image recognition during subsequent surface defect detection. It should also be noted that when defining the normalized generatrix position, for a frustum structure, its generatrix can be considered a segment of the cone's generatrix. Therefore, for a cone, its small end S... min =0; For a frustum structure, its small end 0 < S min <1, s=1 corresponds to the large end of the frustum structure.

[0105] The aforementioned correspondence is as follows:

[0106] ;

[0107] Where θ is the rotation angle, x(s) is the scaling factor, and α is the half-apex angle of the conical component;

[0108] The target image size is set to H*W, where H and W are determined based on the pixel precision requirements of the target image. H represents the number of rows, and W represents the number of columns.

[0109] The relationship between the pixel coordinates (i, j) of the target image and the coordinate system (l, s) is expressed as follows:

[0110] ;

[0111] ;

[0112] Where i = 0, 1, 2, ..., H-1; j = 0, 1, 2, ..., W-1, representing the scaling factor at different positions in each row direction. L is the generatrix length of the conical member, and α is the semi-apex angle of the conical member;

[0113] Arc length range Normalized busbar position .

[0114] Furthermore, to improve image quality, the aforementioned generation of the target image based on the representation of the pixel coordinates (i, j) of the target image in coordinate system (l, s) includes:

[0115] The pixel coordinates of each target image are traversed using an inverse mapping method. The pixel coordinates (i, j) of the target image are calculated using the inverse transformation formula and then inversely mapped to the floating-point pixel coordinates of the original image. ;

[0116] The inverse transformation formula can be obtained by solving the following relation:

[0117] ;

[0118] ;

[0119] Then the coordinates are inversely mapped to the original image. It can be represented as:

[0120] ;

[0121] ;

[0122] Based on the principle of region interpolation, each pixel's floating-point coordinate In the original image, for a region of a certain size, calculate the weighted average of all pixels in that region to obtain the interpolated pixel value;

[0123] The interpolated pixel values ​​are assigned to the corresponding pixel (i, j) in the target image to generate the target image.

[0124] The above are merely specific embodiments of this application, but the scope of protection of this application is not limited thereto. Those skilled in the art should understand that this application includes, but is not limited to, the contents described in the accompanying drawings and the specific embodiments above. Any modifications that do not depart from the functional and structural principles of this application will be included within the scope of the claims.

Claims

1. A method for acquiring surface images of conical components, characterized in that, The surface image acquisition method includes: The linear scan camera in the image acquisition unit, in conjunction with the telecentric lens, performs line scanning on the rotating conical component at a fixed line frequency to acquire the original image with regular distortion. Based on the fixed line frequency and the rotational motion of the conical component, a coordinate mapping relationship is established between the pixel coordinates of the original image and the surface geometric coordinates of the conical component; The original image is corrected according to the coordinate mapping relationship to obtain the target image.

2. The surface image acquisition method as described in claim 1, characterized in that, The fixed line frequency is set based on the linear velocity of the outer edge of the large end of the conical component when the conical component rotates.

3. The surface image acquisition method as described in claim 2, characterized in that, The fixed line frequency F satisfies the following formula: ; Where ω is the rotational angular velocity of the conical component, L is the generatrix length of the conical component, α is the half-apex angle of the conical component, and σ is the object space resolution of the image acquisition unit.

4. The surface image acquisition method as described in claim 1, characterized in that, The line scan camera is set to continuously acquire N rows of images at a fixed line frequency to form the original image. The row index u = 0, 1, 2, ..., N-1 and the column index v = 0, 1, 2, ..., P-1 in the pixel coordinates (u, v) of the original image; where P is the resolution of the line scan camera. The coordinate mapping relationship is configured as follows: ； ； in, Let θ be the rotation angle corresponding to the u-th row of the original image, and ω be the rotational angular velocity of the conical component. The acquisition time from the original image to the u-th row of the image; The relative position of a pixel in the v-th column of the original image projected onto the generatrix of the tapered component. Let σ be the coordinate component of the pixel in the v-th column of the original image along the axial direction of the conical component, σ be the object space resolution of the image acquisition unit, L be the generatrix length of the conical component, and α be the half-apex angle of the conical component.

5. The surface image acquisition method as described in claim 4, characterized in that, The correction of the original image based on the coordinate mapping relationship includes: The scaling factor λ is determined based on the coordinate mapping relationship. The scaling factor λ is used to characterize the local geometric scaling ratio of each row of pixels distributed along the axial direction of the conical component in the original image due to the difference in linear velocity during circumferential imaging. A rectangular unfolding strategy based on arc length-generic coordinates is used to establish the correspondence between the coordinate system (l, s) of the target image and the coordinate system (θ, s) of the original image; Based on the correspondence and the representations of the pixel coordinates (u, v) of the original image and the pixel coordinates (i, j) of the target image in their respective coordinate systems (θ, s) and (l, s), the coordinate transformation between the original image and the target image is completed, and the target image is generated.

6. The surface image acquisition method as described in claim 5, characterized in that, Define the normalized bus position s∈[S min [,1], where s= S min For the cone apex, s=1 corresponds to the large end, then the scaling factor λ satisfies the following formula: ； Where s is the normalized bus position, Let be the scaling factor at position s on the generatrix of the tapered member. Let V(s) be the scaling factor at the reference datum, and V(s) be the linear velocity at position s on the generatrix of the tapered member. Let be the linear velocity of the outer edge of the large end of the conical component during its rotational motion. Let ω be the normalized position parameter value of the large end of the conical component, ω be the rotational angular velocity of the conical component, L be the generatrix length of the conical component, and α be the half-apex angle of the conical component. and All are 1.

7. The surface image acquisition method as described in claim 5, characterized in that, The correspondence is as follows: ； Where θ is the rotation angle, x(s) is the x-axis coordinate of a point on the generatrix, and α is the half-apex angle of the conical component.

8. The surface image acquisition method as described in claim 7, characterized in that, The size of the target image is set to H*W, where H and W are determined according to the pixel precision requirements of the target image, H is the number of rows, and W is the number of columns; The relationship between the pixel coordinates (i, j) of the target image and the coordinate system (l, s) is expressed as follows: ； ； Where i = 0, 1, 2, ..., H-1; j = 0, 1, 2, ..., W-1, representing the scaling factor at different positions in each row direction. L is the generatrix length of the conical member, and α is the semi-apex angle of the conical member; Arc length range Normalized busbar position .

9. The surface image acquisition method as described in claim 8, characterized in that, Generating the target image based on the representation of the pixel coordinates (i, j) of the target image in coordinate system (l, s) includes: The pixel coordinates of each target image are traversed using an inverse mapping method. The pixel coordinates (i, j) of the target image are calculated using the inverse transformation formula and then inversely mapped to the floating-point pixel coordinates of the original image. ; The inverse transformation formula can be obtained by solving the following relation: ； ； Then the coordinates are inversely mapped to the original image. It can be represented as: ； ； Based on the principle of region interpolation, each pixel floating-point coordinate In the original image, for a region of a certain size, calculate the weighted average of all pixels in that region to obtain the interpolated pixel value; The interpolated pixel values ​​are assigned to the corresponding pixel (i, j) in the target image to generate the target image.

10. A surface image acquisition device for a conical component, characterized in that, The surface image acquisition device includes: A rotary drive unit is used to drive a tapered component to rotate about its own axis. The image acquisition unit includes a line scan camera and a telecentric lens. The line scan camera is used in conjunction with the telecentric lens to perform line scanning on a rotating conical component at a fixed line frequency to acquire an original image with regular distortion. The data processing unit is used to establish a coordinate mapping relationship between the pixel coordinates of the original image and the pixel coordinates of the surface unfolded image of the conical component based on the fixed line frequency and the rotational motion of the conical component, and to correct the original image based on the coordinate mapping relationship and the transformation algorithm to obtain the target image.

Images