Sub-pixel edge extraction method based on interpolation method

Abstract

The invention relates to a sub-pixel edge extraction method based on an interpolation method, and belongs to the technical field of image processing and computer vision, and the method comprises the steps: obtaining a gray level image, and calculating the gradient of the gray level image in at least one direction based on a preset operator, so as to obtain the gradient amplitude of the gray level image; compressing the gradient magnitude to a preset threshold value, and calculating the gradient direction of each pixel in the gray level image; based on a preset interpolation model, in a preset period, repeatedly fitting the edge intensity in the gradient direction until a complete sub-pixel contour point set is obtained; and deploying the trained interpolation model as a sub-pixel edge extraction model so as to improve the image processing precision in edge detection. By means of the method, the problem that a traditional pixel-level edge detection method is insufficient in precision can be solved, and the detection precision and the measurement repeatability can be effectively improved in the field of high-precision measurement.

Description

Technical Field

[0001] The invention relates to the technical field of image processing and computer vision, and in particular to a sub-pixel edge extraction method based on an interpolation method. Background Art

[0002] Edge detection is one of the basic technologies in the field of image processing and computer vision, and is widely used in scenarios such as target detection, feature extraction, image segmentation, and industrial inspection. Its goal is to identify the boundary information of the target object through the grayscale, texture, or color changes of the image, providing basic support for subsequent high-level visual tasks.

[0003] However, in high-precision measurement and detection tasks, traditional pixel-level edge detection methods (such as Canny algorithm, Sobel operator, Prewitt operator, etc.) are limited by image resolution and cannot meet the requirements of sub-pixel accuracy.

[0004] Currently, there is a lack of a detection method that can meet the requirements of sub-pixel level precision edge detection technology.

[0005] The information disclosed in this background technology section is only intended to enhance the understanding of the overall background of the invention and should not be regarded as an acknowledgment or any form of suggestion that the information constitutes the prior art already known to a person skilled in the art. Summary of the invention

[0006] The technical problem to be solved by the present invention is to overcome the above-mentioned defects and provide a sub-pixel edge extraction method based on interpolation method.

[0007] In order to solve the above technical problems, the technical solution provided by the present invention is:

[0008] A sub-pixel edge extraction method based on interpolation method comprises: obtaining a grayscale image, and calculating the gradient of at least one direction of the grayscale image based on a preset operator to obtain the gradient amplitude of the grayscale image; compressing the gradient amplitude to a preset threshold, and calculating the gradient direction of each pixel in the grayscale image; based on a preset interpolation model, repeatedly fitting the edge strength in the gradient direction within a preset period until a complete set of sub-pixel contour points is obtained; and deploying the trained interpolation model as a sub-pixel edge extraction model to improve the accuracy of image processing in edge detection.

[0009] Optionally, acquiring a grayscale image and calculating the gradient of the grayscale image in at least one direction based on a preset operator to obtain the gradient amplitude of the grayscale image includes: using a Sobel and / or Canny operator to respectively calculate the gradients of the grayscale image in the x direction and the y direction.

[0010] Optionally, compressing the gradient amplitude to within a preset threshold includes: compressing the gradient amplitude of the grayscale image to between 0-255.

[0011] Optionally, based on a preset interpolation model, the edge strength in the gradient direction is repeatedly fitted within a preset period until a complete set of sub-pixel contour points is obtained, including: fitting the edge strength in the gradient direction based on an interpolation method; constructing a Cartesian coordinate system based on the edge strength; fitting a quadratic function using Lagrange interpolation and / or parabolic interpolation and / or least squares method, and finding the coordinates of the extreme points of the function; mapping the extreme point coordinates found from the Cartesian coordinate system to the original image coordinate system to obtain the final sub-pixel coordinates of the position.

[0012] Optionally, after mapping the extreme point coordinates found from the Cartesian coordinate system to the original image coordinate system to obtain the final sub-pixel coordinates of the position, the sub-pixel edge extraction method further includes: repeating the previous steps until a complete set of sub-pixel contour points is obtained. Optionally, the interpolation model is specifically a gradient amplitude estimation model based on the interpolation gradient direction.

[0013] On the other hand, the present invention also provides a sub-pixel edge extraction system based on interpolation method, the sub-pixel edge extraction system includes a hardware component and a software component, the hardware component provides a computing basis for the software component to run the sub-pixel edge extraction method.

[0014] Optionally, the hardware components include: a high-resolution industrial camera, a dual telecentric lens, a light source control module, a computing unit, and a display device.

[0015] On the other hand, the present invention further provides a control device, which includes: a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the computer program to implement the sub-pixel edge extraction method.

[0016] On the other hand, the present invention further provides a machine-readable storage medium, on which instructions are stored, and the instructions enable a machine to execute the sub-pixel edge extraction method.

[0017] Compared with the traditional technology, the present invention has at least the following advantages:

[0018] (1) The method described in the present invention improves the accuracy of edge extraction and can reach the sub-pixel level.

[0019] (2) The method described in the present invention enhances the robustness against noise and effectively reduces errors by optimizing the interpolation algorithm.

[0020] (3) The method algorithm described in the present invention has low computational complexity and is suitable for real-time application scenarios. BRIEF DESCRIPTION OF THE DRAWINGS

[0021] Figure 1 is a flow chart of a sub-pixel edge extraction method based on interpolation method provided by an embodiment of the present invention;

[0022] Figure 2 is the original image of the grayscale image calculated by the present invention using a preset operator;

[0023] Figure 3 It is a schematic diagram of the present invention using a preset operator to calculate the gradient of a grayscale image in the X direction;

[0024] Figure 4 It is a schematic diagram of the present invention using a preset operator to calculate the gradient of a grayscale image in the Y direction;

[0025] Figure 5 It is a schematic diagram of the present invention after compressing the gradient amplitude to between 0 and 255;

[0026] Figure 6 is a schematic diagram of pixel gradient amplitude estimation provided by one of the examples of the present invention;

[0027] Figure 7 It is a schematic diagram of constructing a Cartesian coordinate system provided by one of the examples of the present invention;

[0028] Figure 8 It is a schematic diagram of fitting a quadratic function using Lagrange interpolation method and finding the coordinates of the extreme points of the function provided by one of the examples of the present invention;

[0029] Fig. 9 It is one of the schematic diagrams provided in one of the examples of the present invention for mapping the coordinates of the found extreme points from the Cartesian coordinate system to the original image coordinate system;

[0030] Fig.10 This is a second schematic diagram of mapping the coordinates of the found extreme point from the Cartesian coordinate system to the original image coordinate system provided by one of the examples of the present invention;

[0031] Fig.11 It is an original image of a comparative example provided in one of the examples of the present invention;

[0032] Fig.12 is a schematic diagram of the gradient amplitude of a comparative example provided in one of the examples of the present invention;

[0033] Fig.13 It is a schematic diagram of the edge contour of the Canny operator provided in one of the examples of the present invention;

[0034] Fig.14 This is a comparison chart of edge extraction algorithms provided in one of the examples of the present invention. DETAILED DESCRIPTION

[0035] Exemplary embodiments will be described in detail herein, examples of which are shown in the accompanying drawings. When the following description refers to the drawings, the same numbers in different drawings represent the same or similar elements unless otherwise indicated. The embodiments described in the following exemplary embodiments do not represent all embodiments consistent with the present disclosure. Instead, they are merely examples of devices consistent with some aspects of the present disclosure as detailed in the appended claims.

[0036] As mentioned above, in the high-precision measurement and detection tasks of the existing technology, traditional pixel-level edge detection methods (such as Canny algorithm, Sobel operator, Prewitt operator, etc.) are difficult to meet the requirements of sub-pixel accuracy due to the limitations of image resolution.

[0037] In view of this, the present invention provides a sub-pixel edge extraction method based on interpolation method, which can take into account the gradient of adjacent pixels of the pixel-level edge and finally calculate the precise edge coordinates by fitting the curve to improve the detection accuracy.

[0038] Please refer to the instruction manual Figure 1 ,like Figure 1 As shown, the present invention provides a sub-pixel edge extraction method based on interpolation method, the control subject of which can be a controller, which includes the following steps:

[0039] S100, acquiring a grayscale image, and calculating the gradient of the grayscale image in at least one direction based on a preset operator to obtain the gradient amplitude of the grayscale image.

[0040] S200, compressing the gradient amplitude to a preset threshold, and calculating the gradient direction of each pixel in the grayscale image.

[0041] S300 , based on a preset interpolation model, repeatedly fitting the edge strength in the gradient direction within a preset period until a complete set of sub-pixel contour points is obtained.

[0042] S400, deploying the trained interpolation model as a sub-pixel edge extraction model to improve the accuracy of image processing in edge detection.

[0043] Through the above method, the present invention can repeatedly fit the edge strength in the gradient direction of each pixel in the grayscale image within a preset period, and finally obtain a complete sub-pixel contour point set by continuously optimizing the preset model to improve the accuracy of image processing in edge detection.

[0044] In order to further explain the method of the present invention clearly, based on the above embodiment, the present invention also provides another preferred embodiment. In another preferred embodiment of the present invention, the S100 may also include: S110, using the Sobel and / or Canny operators to respectively calculate the gradients of the grayscale image in the X direction and the Y direction.

[0045] For example, the calculation result of S110 is shown in the attached Figure 2-Figure 4 As shown, Figure 2 is the original image of the grayscale image calculated by the preset operator, Figure 3 To calculate the gradient diagram of the grayscale image in the X direction, Figure 4 Schematic diagram for calculating the gradient of a grayscale image in the Y direction. Figure 2 After calculation using the Sobel and / or Canny operators, we finally get Figure 3 Figure 4 The gradient map in the X and Y directions is shown.

[0046] For example, the Sobel operator is used to calculate the gradients of the grayscale image in the X and Y directions. The Sobel operator is a discrete differential operator used for edge detection. It approximates the gradient of the image by calculating the grayscale change rate of each pixel in the image. The gradient can reflect the drastic degree of change of the pixel value in the image, thereby helping to identify the edge information in the image. The Sobel operator is mainly composed of two convolution kernels, one for calculating the horizontal (X direction) gradient and the other for calculating the vertical (Y direction) gradient. When calculating the gradient in the X direction, the Sobel operator X-direction convolution kernel is usually [-1,0,1; -2,0,2; -1,0,1] (here a semicolon is used to indicate the separation of rows). The design of this convolution kernel is based on the sensitivity to the change of the pixel grayscale value in the horizontal direction. When the convolution kernel is convolved with the image, it performs a weighted summation of the grayscale values ​​of each pixel and its left and right adjacent pixels to obtain the approximate gradient value of the pixel in the X direction.

[0047] For each pixel in the image, the Sobel operator X-direction convolution kernel is convolved with the pixel and its surrounding 3×3 neighborhood pixels. The specific calculation method is to multiply each element in the convolution kernel with the neighborhood pixel value at the corresponding position, and then add all the products. For example, for the pixel with coordinates (i, j) in the image, its X-direction gradient GX(i, j) is calculated as follows:

[0048] GX(i,j)=(-1)*f(i-1,j-1)+0*f(i-1,j)+1*f(i-1,j+1)+(-2)*f(i,j-1)+0*f(i,j)+2*f(i,j+1)+(-1)*f(i+1,j-1)+0*f(i+1,j)+1*f(i+1,j+1), where f(i,j) represents the grayscale value of the pixel in the image at coordinate (i,j).

[0049] Therefore, by repeating the above convolution calculation for each pixel in the image, we can get the gradient map of the entire image in the X direction. In this gradient map, the value of each pixel represents the degree of grayscale change of the pixel in the X direction. The places with larger gradient values ​​usually correspond to the edges or areas with obvious texture changes in the image.

[0050] When calculating the gradient in the Y direction, the Sobel operator Y direction convolution kernel is usually [-1,-2,-1; 0,0,0; 1,2,1]. This convolution kernel is mainly used to detect grayscale changes in the vertical direction. Its principle is similar to the X direction convolution kernel, but the weight distribution is designed for the vertical direction.

[0051] Similarly, for each pixel in the image, the Sobel operator Y-direction convolution kernel is convolved with the pixel and its 3×3 neighboring pixels. For example, for a pixel with coordinates (i, j), its Y-direction gradient GY(i, j) is calculated as follows:

[0052] GY(i,j)=(-1)*f(i-1,j-1)+(-2)*f(i,j-1)+(-1)*f(i+1,j-1)+0*f(i-1 ,j)+0*f(i,j)+0*f(i+1,j)+1*f(i-1,j+1)+2*f(i,j+1)+1*f(i+1,j+1).

[0053] Therefore, by performing the above calculations on each pixel of the entire image, we can get the gradient map of the image in the Y direction. The pixel values ​​in the Y-direction gradient map represent the grayscale change degree of the image in the vertical direction. Similarly, there will be larger gradient values ​​at the edges and texture changes.

[0054] In another preferred embodiment of the present invention, the S200 may further include: S210, compressing the gradient amplitude of the grayscale image to between 0 and 255. For example, and continuing the above example, the compressed gradient amplitude image is as follows: Figure 5As shown, the linear scaling method can be used to compress the gradient amplitude of the grayscale image. For example: the minimum value of the original gradient amplitude is minVal, and the maximum value is maxVal. For any gradient amplitude G, its compressed amplitude G' can be calculated by the following formula: G' = ((G-minVal) / (maxVal-minVal))*255. In this way, the original gradient amplitude is mapped to the range of 0-255. For example, if a gradient amplitude is 1000, minVal = 0, maxVal = 2000, then the compressed amplitude G' = ((1000-0) / (2000-0))*255 = 127.5 (usually rounding operations such as rounding or truncation are performed).

[0055] In another preferred embodiment of the present invention, the S300 may further include: S310, fitting the edge strength in the gradient direction based on the interpolation method. Figure 6 As shown in the attached Figure 6 Among them, point A and point B are located between two pixels, and their own gradient amplitude cannot be directly obtained; therefore, it is necessary to estimate the coordinates of these two points and their corresponding gradient assignments. The gradient amplitude of A Same reason

[0056]

[0057] In another preferred embodiment of the present invention, the step S310 may further include: S320, constructing a Cartesian coordinate system based on the edge strength. Figure 7 As shown in the attached Figure 7 Among them, a new Cartesian coordinate system is constructed with the black line direction as the x-axis, the center points of A and B as the origin, and the gradient amplitude as the y-axis.

[0058] In another preferred embodiment of the present invention, the step S320 may further include: S330, fitting a quadratic function using Lagrange interpolation method and / or parabola interpolation method and / or least square method, and finding the coordinates of the extreme points of the function. For example, and continuing the above example, the Lagrange interpolation method is used to fit a quadratic function, and the result is as follows: Figure 8 shown.

[0059] In another preferred embodiment of the present invention, the step S330 may further include: S340, mapping the extreme point coordinates found from the Cartesian coordinate system to the original image coordinate system to obtain the final sub-pixel coordinates of the position. For example, and continuing the above example, according to the method described in S340, the x-axis coordinate Xc corresponding to the extreme point can be first found, and then the xc position can be mapped to the image coordinate system, that is, the sub-pixel coordinates of the contour point can be obtained (such as Figure 9-10 As shown, the image coordinate system is of float type).

[0060] S350 , repeat S310 - S340 until a complete sub-pixel contour point set is obtained.

[0061] Finally, the complete sub-pixel contour point set obtained above is verified, as shown in the following example: Fig.11 is the original grayscale image, Fig.13 is the edge image extracted by the Canny operator in the prior art. Fig.12 This is the sub-pixel edge contour extracted by the embodiment of the present invention. The extraction algorithm is compared with Fig.14 As shown in the edge extraction algorithm comparison diagram, it can be seen that the sub-pixel edge extraction algorithm based on interpolation proposed in the present invention extracts smoother edges. The main reason is that the sub-pixel algorithm proposed in the present invention can accurately locate the edge of the image to the position between two pixels, making the granularity of the edge change more delicate, so the edge transition is smoother. Compared with the traditional Canny algorithm, it can only locate the edge to the pixel level unit, and the coordinate system of the image is integer data, so its performance in the oblique line is stepped and the fluctuation range is large.

[0062] Therefore, the sub-pixel edge extraction method based on interpolation method described in the present invention has at least the following advantages compared with the prior art:

[0063] (1) The method described in the present invention improves the accuracy of edge extraction and can reach the sub-pixel level.

[0064] (2) The method described in the present invention enhances the robustness against noise and effectively reduces errors by optimizing the interpolation algorithm.

[0065] (3) The method algorithm described in the present invention has low computational complexity and is suitable for real-time application scenarios.

[0066] Based on the same general inventive concept, the present invention also provides a sub-pixel edge extraction system based on interpolation method, the sub-pixel edge extraction system includes a hardware component and a software component, the hardware component provides a computing basis for the software component to run the sub-pixel edge extraction method.

[0067] Based on the same general inventive concept, the present invention also provides a control device, which includes: a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the computer program to implement the sub-pixel edge extraction method.

[0068] Based on the same general inventive concept, the present invention further provides a machine-readable storage medium, on which instructions are stored, and the instructions enable a machine to execute the sub-pixel edge extraction method.

[0069] It should be understood that in various embodiments of the present invention, the size of the serial numbers of the above-mentioned processes does not mean the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of the present invention.

[0070] In addition, the terms "system" and "network" are often used interchangeably in this article. The term "and / or" in this article is only a description of the association relationship of the associated objects, indicating that there can be three relationships. For example, A and / or B can mean: A exists alone, A and B exist at the same time, and B exists alone. In addition, the character " / " in this article generally indicates that the associated objects before and after are in an "or" relationship.

[0071] It should be understood that in the embodiment of the present invention, "B corresponding to A" means that B is associated with A, and B can be determined according to A. However, it should also be understood that determining B according to A does not mean determining B only according to A, and B can also be determined according to A and / or other information.

[0072] Those of ordinary skill in the art will appreciate that the units and algorithm steps of each example described in conjunction with the embodiments disclosed herein can be implemented with electronic hardware, computer software, or a combination of the two. In order to clearly illustrate the interchangeability of hardware and software, the composition and steps of each example have been generally described in terms of function in the above description. Whether these functions are performed in hardware or software depends on the specific application and design constraints of the technical solution. Professional and technical personnel can use different methods to implement the described functions for each specific application, but such implementation should not be considered to be beyond the scope of the present invention. Those skilled in the art can clearly understand that for the convenience and simplicity of description, the specific working process of the system, device, and unit described above can refer to the corresponding process in the aforementioned method embodiment, and will not be repeated here.

[0073] In the several embodiments provided in the present application, it should be understood that the disclosed systems, devices and methods can be implemented in other ways. For example, the device embodiments described above are only schematic. For example, the division of the units is only a logical function division. There may be other division methods in actual implementation, such as multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. In addition, the mutual coupling or direct coupling or communication connection shown or discussed can be an indirect coupling or communication connection through some interfaces, devices or units, or it can be an electrical, mechanical or other form of connection.

[0074] The units described as separate components may or may not be physically separated, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed on multiple network units. Some or all of the units may be selected according to actual needs to achieve the purpose of the embodiments of the present invention.

[0075] In addition, each functional unit in each embodiment of the present invention may be integrated into one processing unit, or each unit may exist physically separately, or two or more units may be integrated into one unit. The above-mentioned integrated unit may be implemented in the form of hardware or in the form of software functional units.

[0076] Through the description of the above embodiments, it is clear to those skilled in the art that the present invention can be implemented in hardware, firmware, or a combination thereof. When software is used for implementation, the above functions can be stored in a computer-readable medium or transmitted as one or more instructions or codes on a computer-readable medium. Computer-readable media include computer storage media and communication media, wherein the communication media include any medium that facilitates the transmission of a computer program from one place to another. The storage medium can be any available medium that a computer can access. Taking this as an example but not limited to: a computer-readable medium can include RAM, ROM, EEPROM, CD-ROM or other optical disk storage, disk storage medium or other magnetic storage device, or any other medium that can be used to carry or store a desired program code in the form of an instruction or data structure and can be accessed by a computer. In addition. Any connection can be appropriately a computer-readable medium. For example, if the software is transmitted from a website, a server or other remote source using a coaxial cable, an optical fiber cable, a twisted pair, a digital subscriber line (DSL), or wireless technologies such as infrared, radio, and microwaves, then the coaxial cable, optical fiber cable, twisted pair, DSL, or wireless technologies such as infrared, wireless, and microwaves are included in the fixing of the medium. As used in the present invention, disk and disc include compact disc (CD), laser disc, optical disc, digital versatile disc (DVD), floppy disk and blue-ray disc, wherein disk usually reproduces data magnetically, while disc uses laser to reproduce data optically. The above combination should also be included in the scope of protection of computer-readable media.

[0077] In short, the above is only a preferred embodiment of the technical solution of the present invention, and is not intended to limit the protection scope of the present invention. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present invention should be included in the protection scope of the present invention.

Claims

1. A sub-pixel edge extraction method based on interpolation method, characterized in that: include: Acquire a grayscale image, and calculate the gradient of the grayscale image in at least one direction based on a preset operator to obtain the gradient amplitude of the grayscale image; Compressing the gradient amplitude to a preset threshold, and calculating the gradient direction of each pixel in the grayscale image; Based on a preset interpolation model, repeatedly fitting the edge strength in the gradient direction within a preset period until a complete set of sub-pixel contour points is obtained; The trained interpolation model is deployed as a sub-pixel edge extraction model to improve the accuracy of image processing in edge detection.

2. The sub-pixel edge extraction method according to claim 1, characterized in that: The step of acquiring a grayscale image and calculating the gradient of the grayscale image in at least one direction based on a preset operator to obtain the gradient magnitude of the grayscale image includes: using a Sobel and / or Canny operator to respectively calculate the gradients of the grayscale image in the x direction and the y direction.

3. The sub-pixel edge extraction method according to claim 1, characterized in that: The compressing the gradient amplitude to a preset threshold value includes: compressing the gradient amplitude of the grayscale image to a value between 0 and 255.

4. The sub-pixel edge extraction method according to claim 1, characterized in that: The method of repeatedly fitting the edge strength in the gradient direction based on the preset interpolation model within a preset period until a complete set of sub-pixel contour points is obtained includes: Fit the edge strength in the gradient direction based on interpolation method; Based on the edge strength, construct a Cartesian coordinate system; Fitting a quadratic function using Lagrange interpolation method and / or parabolic interpolation method and / or least square method, and finding the coordinates of the extreme points of the function; The found extreme point coordinates are mapped from the Cartesian coordinate system to the original image coordinate system to obtain the final sub-pixel coordinates of the position.

5. The sub-pixel edge extraction method according to claim 4, characterized in that: After mapping the found extreme point coordinates from the Cartesian coordinate system to the original image coordinate system to obtain the final sub-pixel coordinates of the position, the sub-pixel edge extraction method further includes: Repeat the previous steps until a complete set of sub-pixel contour points is obtained.

6. The sub-pixel edge extraction method according to claim 1, characterized in that: The interpolation model is specifically a gradient amplitude estimation model based on the gradient direction of interpolation.

7. A sub-pixel edge extraction system based on interpolation method, characterized in that: The sub-pixel edge extraction system includes a hardware component and a software component, wherein the hardware component provides a computing basis for the software component to run the sub-pixel edge extraction method according to any one of claims 1-5.

8. The sub-pixel edge extraction system based on interpolation method according to claim 7, characterized in that: The hardware components include: a high-resolution industrial camera, a dual telecentric lens, a light source control module, a computing unit, and a display device.

9. A control device, characterized in that: The control device comprises: a memory, a processor, and a computer program stored in the memory and executable on the processor, and the processor executes the computer program to implement the sub-pixel edge extraction method according to any one of claims 1 to 6.

10. A machine-readable storage medium, characterized in that: The machine-readable storage medium stores instructions, which enable the machine to execute the sub-pixel edge extraction method according to any one of claims 1-6.

Images