Method and device for detecting position of ice layer in nuclear fusion target pellet, medium and equipment

Abstract

The application provides a method, device, medium and equipment for detecting the position of an ice layer in a nuclear fusion target pellet, the method comprising: acquiring an X-ray image of the nuclear fusion target pellet; obtaining a binary image through noise reduction processing and adaptive binarization processing according to the X-ray image; obtaining the spherical shell position of the nuclear fusion target pellet according to the binary image; obtaining a refined skeleton graph of the binary image through an inflation algorithm and an image skeleton extraction algorithm according to the spherical shell position; removing the skeleton branches in the refined skeleton graph to obtain a skeleton trunk graph; and obtaining the position of the ice layer in the nuclear fusion target pellet according to the skeleton trunk graph. The method for detecting the position of the ice layer in the nuclear fusion target pellet can significantly reduce the error rate of ice layer position detection, obtain accurate ice layer position information, and further determine the thickness of the ice layer.

Description

Technical Field

[0001] This invention relates to the field of nuclear fusion detection, and more specifically, to a method, apparatus, medium, and equipment for detecting the location of ice layers in a nuclear fusion target pellet. Background Technology

[0002] With the depletion of traditional fossil fuels and their severe environmental pollution, people are constantly searching for new energy sources to solve the current energy crisis. Among them, controlled nuclear fusion, as a new energy source with many advantages such as being clean, safe, and having abundant raw material sources, is gradually gaining popularity. Compared with nuclear fission, controlled nuclear fusion does not produce radioactive materials that pollute the environment, and it can operate continuously and stably in the thin atmosphere, making it cleaner and safer. Furthermore, the amount of nuclear fusion energy on Earth is far more abundant than that of nuclear fission energy.

[0003] To achieve nuclear fusion, the hydrogen isotopes deuterium and tritium must be heated to millions of degrees Celsius, causing deuterium-tritium (DT) to overcome their electrostatic repulsion and combine to form helium nuclei, releasing energy as high-energy neutrons. Laser fusion is currently considered a feasible method for achieving controlled nuclear fusion. A key step involves irradiating a target pellet with extremely high-power laser pulses, causing the thermonuclear fuel to ionize rapidly. This places extremely high demands on the uniformity of the laser beam and the quality of the target pellet. How DT is filled into the target pellet directly affects its quality, and the filling tube method is currently one of the mainstream filling methods.

[0004] First, the temperature of the target pellet is controlled at around 19K. Then, gaseous tritium (DT) is filled into the pellet. When the gaseous DT enters the pellet, it liquefies. Once the required liquid level is reached, any flow of DT into or out of the pellet is stopped. Next, the pellet is frozen. The freezing process takes up to 15 hours to form single crystals and layers. When radioactive tritium releases high-energy particles, laminar radiation occurs. Thicker solid regions receive more energy than thinner regions, causing DT to sublimate in the thicker solid regions and deposit in the thinner regions. Ultimately, a uniform layer of solid DT, i.e., an ice layer, forms on the inner surface of the pellet.

[0005] In controlled nuclear fusion, the thickness of the ice layer must be uniformly distributed within the range of 1 μm-rms. Since the ice layer's position within the target pellet can be used to calculate its thickness, detecting the location of the ice layer within the target pellet is crucial. To ensure target pellet quality, a highly accurate, stable, and noise-resistant ice layer detection method is needed to accurately detect the ice layer's position within X-ray images of controlled nuclear fusion target pellets. However, current technology lacks documentation of such a detection method. Summary of the Invention

[0006] To address the shortcomings of existing technologies, this invention provides a method, apparatus, medium, and equipment for detecting the location of ice layers in a nuclear fusion target pellet.

[0007] In a first aspect, the present invention provides a method for detecting the location of ice layers in a nuclear fusion target pellet, used to detect the location of ice layers from X-ray images of a controlled nuclear fusion target pellet, the method comprising:

[0008] Obtain X-ray images of nuclear fusion target pellets;

[0009] Based on the X-ray image, a binarized image is obtained through noise reduction and adaptive binarization.

[0010] The position of the spherical shell of the nuclear fusion target pellet is obtained based on the binarized image;

[0011] Based on the position of the spherical shell, a thinned skeleton map of the binarized image is obtained through a dilation algorithm and an image skeleton extraction algorithm.

[0012] Remove the skeleton branches from the refined skeleton diagram to obtain the skeleton trunk diagram;

[0013] Based on the main skeleton diagram, the location of the ice layer in the nuclear fusion target pellet is obtained.

[0014] Optionally, obtaining the spherical shell position of the nuclear fusion target pellet based on the binarized image includes:

[0015] Based on the binarized image, edge extraction is performed on the binarized image using a binarization function and the Sobel operator to obtain a binarized image containing edge information;

[0016] Perform a Hough transform on the binarized image containing edge information to obtain the preliminary position of the spherical shell;

[0017] Based on the initial position of the spherical shell, a first annular region within ±5% of the spherical shell radius is obtained. The pixels within the first annular region are then fitted using the least squares method to obtain the precise position of the spherical shell.

[0018] Optionally, performing a Hough transform on the binarized image containing edge information to obtain the preliminary position of the spherical shell includes:

[0019] A first Hough transform is performed on the binarized image containing edge information to obtain the center coordinates and radius of the spherical shell;

[0020] A preset error range for the center coordinates and the radius is obtained. Based on the preset error range, a second Hough transform is performed on the binarized image containing edge information to obtain the center coordinate range and the radius range, thereby obtaining the preliminary position of the spherical shell.

[0021] Optionally, obtaining the refined skeleton map of the binarized image based on the position of the spherical shell using a dilation algorithm and an image skeleton extraction algorithm includes:

[0022] Obtain the second annular region within a range of 0.7-0.8 times the radius of the spherical shell at the specified location;

[0023] Based on the second annular region, the ice layer boundary is expanded outward by an expansion algorithm, and then a thinned skeleton map of the binarized image is obtained by an image skeleton extraction algorithm.

[0024] Optionally, obtaining the thinned skeleton map of the binarized image through an image skeleton extraction algorithm includes:

[0025] The first screening is performed on all bright spots in the binarized image, and pixels that meet the condition of equation (1) are deleted:

[0026]

[0027] The remaining bright spots in the binarized image are filtered a second time, and the pixels that meet the equation (2) are deleted to obtain the thinned skeleton map of the binarized image:

[0028]

[0029] In equations (1) and (2), a pixel with a gray value of 0 is defined as the background, and a pixel with a gray value of 1 is defined as the foreground; N(P1) represents the number of foreground pixels among the 8 pixels adjacent to P1; S(P1) represents the cumulative number of times the gray value of adjacent pixels clockwise from P1 changes from 0 to 1; P1 represents the current pixel; and P2, P4, P6 and P8 represent the pixels directly above, below, to the left and to the right of P1, respectively.

[0030] Optionally, removing skeleton branches from the refined skeleton diagram to obtain the skeleton trunk diagram includes:

[0031] Based on the refined skeleton diagram, obtain all endpoints of a single skeleton, wherein each endpoint needs to satisfy either having only one pixel value of 1 in its eight neighborhoods, or having only two pixel values ​​of 1 in its eight neighborhoods and the two pixels being adjacent.

[0032] The two endpoints with the longest distance are selected and used as the beginning and end endpoints of the main skeleton.

[0033] Based on the first and last endpoints, obtain the main skeleton;

[0034] Set the first endpoint other than the beginning and end endpoints to zero, and then recursively proceed to the first bright spot in the eight domains of the first endpoint. If the first bright spot is not a node in the skeleton trunk, set the first bright spot to zero, and continue until a node in the skeleton trunk is encountered, at which point the recursion stops and the skeleton trunk graph is obtained.

[0035] Optionally, obtaining a binarized image from the X-ray image through noise reduction and adaptive binarization includes:

[0036] The X-ray image is processed sequentially by Gaussian low-pass filtering, grayscale stretching, and image sharpening to obtain a noise-reduced X-ray image;

[0037] The denoised X-ray image is processed by adaptive binarization to obtain a binarized image.

[0038] Secondly, the present invention provides a device for detecting the position of ice layers in a nuclear fusion target pellet, the device comprising:

[0039] The image acquisition module is used to acquire X-ray images of the nuclear fusion target pellet;

[0040] The binarization processing module is used to obtain a binarized image based on the X-ray image through noise reduction processing and adaptive binarization processing.

[0041] The shell position acquisition module is used to acquire the shell position of the nuclear fusion target pellet based on the processed binarized image.

[0042] The thinned skeleton acquisition module is used to process the binarized image according to the position of the spherical shell using a dilation algorithm and an image skeleton extraction algorithm to obtain a thinned skeleton map of the binarized image;

[0043] The skeleton trunk acquisition module is used to remove skeleton branches from the refined skeleton diagram and obtain the skeleton trunk diagram.

[0044] The ice layer location acquisition module is used to acquire the location of the middle ice layer of the X-ray image core fusion target pellet based on the skeleton backbone diagram.

[0045] Thirdly, the present invention provides a computer-readable storage medium storing a computer program for executing the method for detecting the location of ice layers in a nuclear fusion target pellet as described above.

[0046] Fourthly, the present invention provides a computer device, including a processor, a memory, and a computer program stored in the memory and executable on the processor, wherein the processor executes the computer program to implement the above-described method for detecting the position of ice layers in a nuclear fusion target pellet.

[0047] The present invention provides a method, apparatus, medium, and equipment for detecting the ice layer position in a nuclear fusion target pellet. By enhancing the effective information in X-ray images through noise reduction and binarization processing, and suppressing invalid noise, a good foundation is provided for subsequent processing. The position of the spherical shell of the nuclear fusion target pellet is then obtained. Based on this, an expansion process is performed to connect the broken ice layers. A refined skeleton map of the image is obtained through an image skeleton extraction algorithm. Since noise is easily generated around the ice layer area after expansion processing, the skeleton branches in the refined skeleton map are further removed, retaining only the main skeleton. This significantly reduces the error rate of ice layer position detection, obtains accurate ice layer position information, and thus further determines the thickness of the ice layer. Attached Figure Description

[0048] Figure 1 This is an application environment diagram of the method for detecting the position of ice layer in a nuclear fusion target pellet according to an embodiment of the present invention;

[0049] Figure 2 This is a flowchart illustrating the method for detecting the location of ice layers in a nuclear fusion target pellet according to an embodiment of the present invention.

[0050] Figure 3 This is a schematic diagram illustrating the principle of image space, i.e., parameter space, in the fast downsampling Hough transform spherical shell detection used in this embodiment of the invention.

[0051] Figure 4 This is a schematic diagram showing the positional relationship between the initial pixel and the eight neighborhoods corresponding to that pixel in an embodiment of the present invention;

[0052] Figure 5 This is a schematic diagram of the skeleton endpoints in an embodiment of the present invention;

[0053] Figure 6 This is a schematic diagram of the main trunk and branches of the image skeleton in an embodiment of the present invention;

[0054] Figure 7 This is a schematic diagram illustrating the removal of skeleton branches using a recursive algorithm in an embodiment of the present invention;

[0055] Figure 8 This is a comparison image before and after removing the skeleton branches in an embodiment of the present invention;

[0056] Figure 9 This is a comparison image of the spherical shell and ice layer before and after detection in an embodiment of the present invention;

[0057] Figure 10 This is a structural block diagram of the device for detecting the position of the ice layer in a nuclear fusion target pellet according to an embodiment of the present invention;

[0058] Figure 11 This is an internal structural diagram of a computer device in an embodiment of the present invention. Detailed Implementation

[0059] To make the above-mentioned objects, features, and advantages of the present invention more apparent and understandable, specific embodiments of the present invention will be described in detail below with reference to the accompanying drawings. Although some embodiments of the present invention are shown in the drawings, it should be understood that the present invention can be implemented in various forms and should not be construed as limited to the embodiments set forth herein. Rather, these embodiments are provided to provide a more thorough and complete understanding of the present invention. It should be understood that the accompanying drawings and embodiments of the present invention are for illustrative purposes only and are not intended to limit the scope of protection of the present invention.

[0060] It should be understood that the various steps described in the method embodiments of the present invention may be performed in different orders and / or in parallel. Furthermore, the method embodiments may include additional steps and / or omit the steps shown. The scope of the present invention is not limited in this respect.

[0061] The term "comprising" and its variations as used in this invention are open-ended, meaning "including but not limited to". The term "based on" means "at least partially based on". The term "one embodiment" means "at least one embodiment"; the term "another embodiment" means "at least one additional embodiment"; the term "some embodiments" means "at least some embodiments"; the term "optionally" means "optional embodiments". Definitions of other terms will be given in the following description. It should be noted that the concepts of "first", "second", etc., mentioned in this invention are only used to distinguish different devices, modules, or units, and are not intended to limit the order of functions performed by these devices, modules, or units or their interdependencies.

[0062] It should be noted that the terms "a" and "a plurality of" used in this invention are illustrative rather than restrictive. Those skilled in the art should understand that, unless otherwise expressly indicated in the context, they should be understood as "one or more".

[0063] The names of the messages or information exchanged between the multiple devices in the embodiments of the present invention are for illustrative purposes only and are not intended to limit the scope of these messages or information.

[0064] Figure 1 This is an application environment diagram of a method for detecting the location of ice layers in a nuclear fusion target pellet, as illustrated in one embodiment. (Refer to...) Figure 1 This method for detecting the location of ice layers in a nuclear fusion target pellet is applied to a system for detecting the location of ice layers in a nuclear fusion target pellet. The system includes a terminal 110 and a server 120. The terminal 110 and server 120 are connected via a network. The terminal 110 can be a desktop terminal or a mobile terminal; the mobile terminal can be at least one of a mobile phone, tablet computer, or laptop computer. The server 120 can be a standalone server or a server cluster consisting of multiple servers.

[0065] like Figure 2 As shown, this embodiment of the invention provides a method for detecting the location of ice layers in a nuclear fusion target pellet, used to detect the location of ice layers from X-ray images of a controlled nuclear fusion target pellet. The method includes:

[0066] Step 210: Obtain X-ray images of the nuclear fusion target pellet;

[0067] Step 220: Based on the X-ray image, obtain a binarized image through noise reduction and adaptive binarization processing;

[0068] Step 230: Obtain the position of the spherical shell of the nuclear fusion target pellet based on the binarized image;

[0069] Step 240: Based on the position of the spherical shell, obtain the thinned skeleton map of the binarized image using a dilation algorithm and an image skeleton extraction algorithm;

[0070] Step 250: Remove the skeleton branches in the refined skeleton diagram to obtain the skeleton trunk diagram;

[0071] Step 260: Based on the skeleton backbone diagram, obtain the position of the ice layer in the nuclear fusion target pellet.

[0072] Figure 2 This is a flowchart illustrating a method for detecting the location of ice layers in a nuclear fusion target pellet, as described in one embodiment. It should be understood that, although... Figure 2 The steps in the flowchart are shown sequentially as indicated by the arrows, but these steps are not necessarily executed in the order indicated by the arrows. Unless otherwise explicitly stated herein, there is no strict order in which these steps are executed, and they can be performed in other orders. Figure 2 At least some of the steps in the process may include multiple sub-steps or multiple stages. These sub-steps or stages are not necessarily executed at the same time, but can be executed at different times. The execution order of these sub-steps or stages is not necessarily sequential, but can be executed in turn or alternately with other steps or at least some of the sub-steps or stages of other steps.

[0073] In step 220, the X-ray image is processed sequentially by Gaussian low-pass filtering, grayscale stretching, and image sharpening to obtain a noise-reduced X-ray image;

[0074] The denoised X-ray image is processed by adaptive binarization to obtain a binarized image.

[0075] Gaussian low-pass filtering suppresses noise in X-ray images. When performing a Fourier transform, the grayscale values ​​of edges and noise pixels in the original image change drastically, corresponding to significant variations in the high-frequency components of their spectrogram. Gaussian low-pass filtering effectively removes noise by suppressing these high-frequency components. Gray-scale stretching enhances the contrast of grayscale values, and image sharpening further enhances this contrast. Finally, adaptive binarization is used to binarize the X-ray image, enabling the segmentation and extraction of effective information. In other words, before binarization, Gaussian low-pass filtering, gray-scale stretching, and image sharpening reduce noise, thereby enhancing the effective information and reducing noise interference to obtain a more accurate binarized image.

[0076] Specifically, the X-ray image is subjected to Gaussian low-pass filtering, including:

[0077] Obtain an X-ray image f0(x,y) with pixel size M*N, and define the scaling parameters as P = 2M and Q = 2N; padded the image f0(x,y) with 0s to obtain a padded image f of size P*Q. p (x,y); use (-1) x+y Multiply by f p (x,y) is used to move the image to the transformation center; then a Fourier transform is performed to obtain F(u,v); a low-pass filter H(u,v) is selected, with the filter size also being P*Q, centered at (P / 2,Q / 2), and multiplied by the previously obtained Fourier image F(u,v) to obtain G(u,v) = H(u,v)*F(u,v). Furthermore, in this embodiment of the invention, a Gaussian filter is selected, and its expression is:

[0078]

[0079] Finally, the image g is obtained through inverse Fourier transform and inverse translation operations. p (x,y), take the top 1 / 4 of the image from the left corner to obtain the filtered image f1(x,y);

[0080] Grayscale stretching of X-ray images includes:

[0081] For the filtered image f1(x,y), its gray value at a certain location, without any prior conditions, can be considered as a random variable in the interval [0,L-1]. Let the probability density function PDF of the input image be p. r (r), an image of size M*N, with gray levels r k The probability of occurrence is approximately:

[0082]

[0083] Where, n k grayscale level r k The total number of pixels. Then the gray levels in the input image are r. k The pixels are mapped to gray levels s in the output image. k for:

[0084]

[0085] Finally, the image f2(x,y) after grayscale stretching is obtained;

[0086] Sharpening X-ray images includes:

[0087] The image f2(x,y) after grayscale stretching still has slight defects and needs to be further sharpened and enhanced using the Placus operator;

[0088] The Laplacian operator for a two-dimensional image function f(x,y) is defined as follows:

[0089]

[0090] For digital image processing, the second-order differential is defined in the following difference form:

[0091]

[0092] Similarly, in the y-direction:

[0093]

[0094] The discrete Laplace operator for the two variables is:

[0095]

[0096] The Laplacian operator will enhance regions with abrupt changes in grayscale in the image, ultimately resulting in the sharpened image f3(x,y);

[0097] Adaptive binarization processing of X-ray images includes:

[0098] For the image f3(x,y) after image sharpening, in order to achieve the segmentation and extraction of effective information in the image, a threshold T is selected to perform binarization processing on the image;

[0099] The binarized image contains only two grayscale information values, and the conversion relationship is as follows:

[0100]

[0101] Choosing an appropriate threshold T is a key factor in determining the image binarization effect. This embodiment of the invention uses the Otsu algorithm to calculate and determine the binarization threshold T based on the inter-class variance of global pixels. The specific calculation process is as follows:

[0102] Input an image of size M*N, where {0,1,2,...,L-1} represents L different gray levels in the image, and n... i Let i represent the number of pixels with gray level i. Then:

[0103]

[0104] In the formula, p i This represents the proportion of all pixels with gray level i.

[0105] Calculate the cumulative mean of grayscale value k in the image:

[0106]

[0107] The average grayscale value of the entire image is:

[0108]

[0109] For k = 0, 1, 2, ..., L-1, calculate the between-class variance:

[0110]

[0111] Calculation makes The gray level k that yields the maximum value * Then the binarization threshold T = k * .

[0112] In step 230, the position of the spherical shell of the nuclear fusion target pellet is obtained based on the binarized image;

[0113] Specifically, based on the binarized image, edge extraction is performed on the binarized image using a binarization function and the Sobel operator to obtain a binarized image containing edge information;

[0114] Perform a Hough transform on the binarized image containing edge information to obtain the preliminary position of the spherical shell;

[0115] Based on the initial position of the spherical shell, a first annular region within ±5% of the spherical shell radius is obtained. The pixels within the first annular region are then fitted using the least squares method to obtain the precise position of the spherical shell.

[0116] Let the amplitude of the image be w×h, and the range of the circle detection radius be unknown. If the maximum circle radius is the maximum amplitude of the image, i.e. r = 1 / 2×max(w,h), then the parameter space size is w×h×r. Performing a Hough transform on this parameter space will greatly occupy memory space and affect the algorithm speed.

[0117] Further, performing a Hough transform on the binarized image containing edge information to obtain the preliminary position of the spherical shell includes:

[0118] A first Hough transform is performed on the binarized image containing edge information to obtain the center coordinates and radius of the spherical shell;

[0119] A preset error range for the center coordinates and the radius is obtained. Based on the preset error range, a second Hough transform is performed on the binarized image containing edge information to obtain the center coordinate range and the radius range, thereby obtaining the preliminary position of the spherical shell.

[0120] To address the issues of high memory consumption and computational complexity in the classic Hough transform algorithm, this invention proposes a fast Hough transform circle detection method for real-time measurement to obtain the preliminary position of a spherical shell, focusing on reducing parameter space and information redundancy. Specifically:

[0121] The original image is downsampled by a factor of 4 to obtain a reduced image. The magnitude of the 4x downsampled image is w / 4 × h / 4. A first Hough transform is then performed on it. Since the width, height, and radius of the circle are all reduced to 1 / 4 of the original, the maximum value of each component in the parameter space is also reduced to 1 / 4. The overall size of the parameter space is 1 / 64 times that of the original image, significantly reducing memory space and improving algorithm speed. Simultaneously, because the image size is only 1 / 16 of the original, the number of pixels involved in the calculation is also 1 / 16 times smaller, which greatly reduces the computational cost of the Hough transform.

[0122] Suppose that the center coordinates and radius of the circle detected after the first Hough transform are (x1, y1, r1); then, by magnifying the detected center coordinates and radius by 4 times, we can obtain the approximate values ​​of the center coordinates and radius of the original image, namely (4x1, 4y1, 4r1).

[0123] Then, considering a certain error range and appropriately relaxing the detection scale, the preset error ranges for the center coordinates and radius are obtained. A new parameter space is then established for the second Hough transform. Let the preset error ranges for the center coordinates and radius be δ1 and δ2, respectively. Then the ranges of the three quantization parameters in the parameter space are as follows:

[0124]

[0125] Where (x2, y2) is the range of center coordinates in the image space of the circle, and r2 is the range of the circle's radius.

[0126] Figure 3 Image (a) is a schematic diagram of image space transformation. Figure 3 Figure (b) is a schematic diagram of parameter space transformation, as shown below. Figure 3 As shown, before the second Hough transform, although it is necessary to restore the original image size by a 4x downsampling, making the image size in the second Hough transform 16 times the image space of the first Hough transform, the first Hough transform provides a rough coordinate of the center and radius, and considers a preset error range, thus limiting the parameter space to within 2δ1 and 2δ2. That is, the image range that needs to be traversed is only... Figure 3 In the rightmost figure (a), the circular position is shown, and in the new parameter space, 2δ1 << x1, y1, 2δ2 << r1. Therefore, the second Hough transform avoids the computational burden caused by a large amount of information redundancy, and the computation time is faster than the first Hough transform. In addition, the Hough transform method provided in this embodiment of the invention does not require a pre-set detection radius, reducing the factor of manual intervention.

[0127] The initial position of the spherical shell of the nuclear fusion target can be obtained through the first and second Hough transforms. In order to obtain the precise position of the shell, precise circle detection needs to be performed on the image. The specific steps are as follows:

[0128] Based on the radius range obtained from the initial position of the spherical shell, a first annular region with a radius of ±5% is extracted. The pixels within the first annular region are then fitted with a circle using the least squares method to obtain the precise position of the spherical shell.

[0129] The basic idea of ​​least squares curve fitting is to minimize the error between all data points and the estimated points (fitting points), thus obtaining the least squares fitted curve. The specific mathematical principle is as follows: For a given set of data {(x... i ,y i If the fitted curve model is y = f(x), then the error distance at the i-th position is f(x). i )-y i The sum of the squared errors of all points is The f(x) is adjusted using certain methods (generally, the function model with undetermined parameters is estimated first, and then the function parameters are adjusted) to obtain the result. The function corresponding to the minimum value is obtained from the fitted curve f(x).

[0130] In step 240, based on the position of the spherical shell, a thinned skeleton map of the binarized image is obtained using a dilation algorithm and an image skeleton extraction algorithm.

[0131] Specifically, a second annular region within a range of 0.7-0.8 times the radius of the spherical shell is obtained at the location of the spherical shell;

[0132] Based on the second annular region, the ice layer boundary is expanded outward by an expansion algorithm, and then a thinned skeleton map of the binarized image is obtained by an image skeleton extraction algorithm.

[0133] In the dilation algorithm, dilation merges all background points in contact with the object into the object, causing the boundary to expand outward, which can fill the voids in the object. In the embodiment of the present invention, the dilation algorithm expands the ice layer boundary outward, while filling some voids in the ice layer area and eliminating small particle noise contained in the ice layer area.

[0134] The core of the image skeleton extraction algorithm is to perform parallel and iterative calculations on the thinning algorithm. It performs logical operations on the eight neighborhoods of the starting point (i.e., top, bottom, left, right, top-left, bottom-left, top-right, and bottom-right). Points that meet the criteria for non-skeleton points are marked as neighborhood points. After traversing all bright spots within connected regions of the image, the marked neighborhood points are deleted, and this process is repeated until the termination condition is met. This algorithm maintains the original straight line direction, intersections, inflection points, and other features while also maintaining computational speed, resulting in superior performance.

[0135] The step of obtaining the refined skeleton map of the binarized image through an image skeleton extraction algorithm includes:

[0136] The first screening is performed on all bright spots in the binarized image, and pixels that meet the condition of equation (1) are deleted:

[0137]

[0138] The remaining bright spots in the binarized image are filtered a second time, and the pixels that meet the equation (2) are deleted to obtain the thinned skeleton map of the binarized image:

[0139]

[0140] In equations (1) and (2), a pixel with a gray value of 0 is defined as the background, and a pixel with a gray value of 1 is defined as the foreground; N(P1) represents the number of foreground pixels among the 8 pixels adjacent to P1; S(P1) represents the cumulative number of times the gray value of adjacent pixels clockwise from P1 changes from 0 to 1; P1 represents the current pixel (starting point); and P2, P4, P6 and P8 represent the pixels directly above, below, to the left and to the right of P1, respectively.

[0141] The positional relationship of pixels P1-P9 is as follows: Figure 4As shown, P1 represents the current pixel (starting point), and P2-P9 represent the 8 pixels adjacent to P1, respectively.

[0142] In step 250, the skeleton branches in the refined skeleton diagram are removed to obtain the skeleton trunk diagram.

[0143] As described in step 240, after the image is processed by the dilation algorithm, the broken ice layer regions will be connected. After processing by the image skeleton extraction algorithm, the single pixel position information of the ice layer can be obtained. However, after the dilation algorithm, there may be noise around the ice layer that is connected to the ice layer region, which will result in many branches on the ice layer path after processing by the image skeleton extraction algorithm, making the ice layer position detection inaccurate.

[0144] To further improve the accuracy and precision of recognition, this invention proposes a skeleton branch removal algorithm, which effectively removes the branches of the skeleton while retaining the main body of the skeleton.

[0145] Specifically, the skeleton branch removal algorithm includes:

[0146] Based on the refined skeleton diagram, obtain all endpoints of a single skeleton, wherein each endpoint needs to satisfy either having only one pixel value of 1 in its eight neighborhoods, or having only two pixel values ​​of 1 in its eight neighborhoods and the two pixels being adjacent.

[0147] The two endpoints with the longest distance are selected and used as the beginning and end endpoints of the main skeleton.

[0148] Based on the first and last endpoints, obtain the main skeleton;

[0149] Set the first endpoint other than the beginning and end endpoints to zero, and then recursively proceed to the first bright spot in the eight domains of the first endpoint. If the first bright spot is not a node in the skeleton trunk, set the first bright spot to zero, and continue until a node in the skeleton trunk is encountered, at which point the recursion stops and the skeleton trunk graph is obtained.

[0150] The principle of the skeleton branch removal algorithm is as follows:

[0151] First, obtain all endpoints in a single skeleton connected region. Each endpoint must satisfy the following condition: either only one pixel in its eight neighborhoods has a value of 1, or only two adjacent pixels in its eight neighborhoods have a value of 1. Figure 5 As shown, Figure 5 In the middle (a), there are skeleton endpoints with only one pixel value of 1 in each of the eight neighborhoods. Figure 5 In the middle (b), there are skeleton endpoints in the eight neighborhoods where only two pixels have a value of 1 and the two pixels are adjacent.

[0152] After obtaining all endpoints of a single skeleton, based on the fact that all skeleton branch endpoints are basically distributed on both sides of the ice layer, and the distance between the first and last endpoints on the main path of the ice layer is the longest, the first and last endpoints of the skeleton trunk are selected. The path of the first and last endpoints along the skeleton is the single-pixel curve of the ice layer region to be extracted. Figure 6 As shown, Figure 6 The black trajectory represents the ice layer (i.e., the main skeleton) that needs to be extracted, and the gray trajectory represents the skeleton branches that need to be removed.

[0153] After obtaining the start and end points of the main skeleton, the skeleton branches are removed using a recursive algorithm, specifically as follows: Figure 7 As shown, starting from the first endpoint other than the first and second endpoints of the non-skeleton backbone, the first endpoint is set to zero, and then the process is recursively repeated to the first bright spot in the eight neighborhoods of the first endpoint. If the first bright spot is not a node in the skeleton backbone, the first bright spot is set to zero. The recursion stops when a node in the skeleton backbone is encountered, and the skeleton backbone graph is obtained. Here, a node is defined as having two or more non-adjacent pixels within its eight neighborhoods.

[0154] It should be noted that in the actual process of removing skeleton branches, a single removal often cannot remove all branches completely, and it is necessary to process in a loop until only the beginning and end points of the skeleton remain.

[0155] like Figure 8 As shown, Figure 8 Image (a) is the refined skeleton diagram of the image before removing skeleton branches. Figure 8 (b) is the skeleton trunk diagram after removing skeleton branches.

[0156] Step 260: Based on the skeleton backbone diagram, obtain the position of the ice layer in the nuclear fusion target pellet.

[0157] Specifically, the skeleton backbone image is added to the binarized image, and all ice layer regions are connected to obtain the location of the ice layer in the nuclear fusion target pellet.

[0158] like Figure 9 As shown, Figure 9 (a) is an X-ray image before the location of the spherical shell and ice layer was detected. Figure 9 Image (b) shows the X-ray image after detecting the positions of the spherical shell and ice layer. The dark lines represent the detection results of the spherical shell, and the lighter lines inside the dark lines represent the detection results of the ice layer positions within the nuclear fusion target pellet. Figure 9 As can be seen, the detection of the ice layer position in the nuclear fusion target pellet provided by the embodiments of the present invention can detect the spherical shell and ice layer position of the nuclear fusion target pellet.

[0159] In order to perform the steps in the above embodiments and various optional embodiments, such as Figure 10As shown, another embodiment of the present invention provides a device for detecting the location of ice layers in a nuclear fusion target pellet, the device comprising:

[0160] Image acquisition module 1100 is used to acquire X-ray images of nuclear fusion target pellets;

[0161] The binarization processing module 1200 is used to obtain a binarized image based on the X-ray image through noise reduction processing and adaptive binarization processing.

[0162] The shell position acquisition module 1300 is used to acquire the shell position of the nuclear fusion target pellet based on the processed binarized image.

[0163] The thinned skeleton acquisition module 1400 is used to process the binarized image according to the position of the spherical shell using a dilation algorithm and an image skeleton extraction algorithm to obtain a thinned skeleton map of the binarized image.

[0164] The skeleton trunk acquisition module 1500 is used to remove skeleton branches from the refined skeleton diagram and obtain the skeleton trunk diagram.

[0165] The ice layer location acquisition module 1600 is used to acquire the location of the middle ice layer of the X-ray image core fusion target pellet based on the skeleton backbone diagram.

[0166] In one embodiment, a computer device is provided, including 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 perform the following steps:

[0167] Obtain X-ray images of nuclear fusion target pellets;

[0168] Based on the X-ray image, a binarized image is obtained through noise reduction and adaptive binarization.

[0169] The position of the spherical shell of the nuclear fusion target pellet is obtained based on the binarized image;

[0170] Based on the position of the spherical shell, a thinned skeleton map of the binarized image is obtained through a dilation algorithm and an image skeleton extraction algorithm.

[0171] Remove the skeleton branches from the refined skeleton diagram to obtain the skeleton trunk diagram;

[0172] Based on the main skeleton diagram, the location of the ice layer in the nuclear fusion target pellet is obtained.

[0173] Figure 11 An internal structural diagram of a computer device in one embodiment is shown. Specifically, this computer device may be... Figure 1 Terminal 110 (or server 120) in the middle. For example... Figure 11As shown, the computer device includes a processor, memory, network interface, input device, and display screen connected via a system bus. The memory includes a non-volatile storage medium and internal memory. The non-volatile storage medium stores an operating system and may also store a computer program. When executed by the processor, this computer program enables the processor to implement a method for detecting the location of ice layers in a nuclear fusion target pellet. The internal memory may also store a computer program, which, when executed by the processor, enables the processor to implement the method for detecting the location of ice layers in a nuclear fusion target pellet. The display screen can be an LCD screen or an e-ink screen. The input device can be a touch layer covering the display screen, buttons, a trackball, or a touchpad mounted on the computer device's casing, or an external keyboard, touchpad, or mouse.

[0174] In one embodiment, the processor, when executing the computer program, also implements the steps of the method for detecting the location of ice layers in the nuclear fusion target pellet described above.

[0175] In one embodiment, a computer-readable storage medium is provided having a computer program stored thereon, the computer program performing the following steps when executed by a processor:

[0176] Obtain X-ray images of nuclear fusion target pellets;

[0177] Based on the X-ray image, a binarized image is obtained through noise reduction and adaptive binarization.

[0178] The position of the spherical shell of the nuclear fusion target pellet is obtained based on the binarized image;

[0179] Based on the position of the spherical shell, a thinned skeleton map of the binarized image is obtained through a dilation algorithm and an image skeleton extraction algorithm.

[0180] Remove the skeleton branches from the refined skeleton diagram to obtain the skeleton trunk diagram;

[0181] Based on the main skeleton diagram, the location of the ice layer in the nuclear fusion target pellet is obtained.

[0182] In one embodiment, when the computer program is executed by a processor, it also implements the steps of the method for detecting the location of ice layers in the nuclear fusion target pellet described above.

[0183] Those skilled in the art will understand that all or part of the processes in the above embodiments can be implemented by a computer program instructing related hardware. The program can be stored in a non-volatile computer-readable storage medium, and when executed, it can include the processes of the embodiments described above. Any references to memory, storage, databases, or other media used in the embodiments provided in this application can include non-volatile and / or volatile memory. Non-volatile memory can include read-only memory (ROM), programmable ROM (PROM), electrically programmable ROM (EPROM), electrically erasable programmable ROM (EEPROM), or flash memory. Volatile memory can include random access memory (RAM) or external cache memory. By way of illustration and not limitation, RAM is available in various forms, such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), dual data rate SDRAM (DDRSDRAM), enhanced SDRAM (ESDRAM), synchronous link DRAM (SLDRAM), Rambus direct RAM (RDRAM), direct memory bus dynamic RAM (DRDRAM), and memory bus dynamic RAM (RDRAM), etc.

[0184] The above description is merely a specific embodiment of the present invention, enabling those skilled in the art to understand or implement the invention. Various modifications to these embodiments will be readily apparent to those skilled in the art, and the general principles defined in this invention may be implemented in other embodiments without departing from the spirit or scope of the invention. Therefore, the present invention is not to be limited to the embodiments shown herein, but is to be accorded the widest scope consistent with the principles and novel features claimed in this invention.

Claims

1. A method for detecting the location of ice layers in a nuclear fusion target pellet, characterized in that, The method for detecting the location of ice layers from X-ray images of controlled nuclear fusion targets includes: Obtain X-ray images of nuclear fusion target pellets; Based on the X-ray image, a binarized image is obtained through noise reduction and adaptive binarization. Based on the binarized image, the location of the spherical shell of the nuclear fusion target pellet is obtained, including: Based on the binarized image, edge extraction is performed on the binarized image using a binarization function and the Sobel operator to obtain a binarized image containing edge information; Perform a Hough transform on the binarized image containing edge information to obtain the preliminary position of the spherical shell; Based on the initial position of the spherical shell, a first annular region within ±5% of the radius of the spherical shell is obtained. The pixels within the first annular region are then fitted with a circle using the least squares method to obtain the precise position of the spherical shell. Based on the position of the spherical shell, a thinned skeleton map of the binarized image is obtained using a dilation algorithm and an image skeleton extraction algorithm, including: Obtain the second annular region within a range of 0.7-0.8 times the radius of the spherical shell at the specified location; Based on the second annular region, the ice layer boundary is expanded outward by an expansion algorithm, and then a thinned skeleton map of the binarized image is obtained by an image skeleton extraction algorithm. Remove the skeleton branches from the refined skeleton diagram to obtain the skeleton trunk diagram; Based on the main skeleton diagram, the location of the ice layer in the nuclear fusion target pellet is obtained.

2. The method for detecting the location of ice layers in a nuclear fusion target pellet according to claim 1, characterized in that, The step of performing a Hough transform on the binarized image containing edge information to obtain the preliminary position of the spherical shell includes: A first Hough transform is performed on the binarized image containing edge information to obtain the center coordinates and radius of the spherical shell; A preset error range for the center coordinates and the radius is obtained. Based on the preset error range, a second Hough transform is performed on the binarized image containing edge information to obtain the center coordinate range and the radius range, thereby obtaining the preliminary position of the spherical shell.

3. The method for detecting the position of ice layer in a nuclear fusion target pellet according to claim 1, characterized in that, The step of obtaining the refined skeleton map of the binarized image through an image skeleton extraction algorithm includes: The first screening is performed on all bright spots in the binarized image, and pixels that meet the condition of equation (1) are deleted: （1）； The remaining bright spots in the binarized image are filtered a second time, and the pixels that meet the equation (2) are deleted to obtain the thinned skeleton map of the binarized image: （2）； In equations (1) and (2), a pixel with a gray value of 0 is defined as the background, and a pixel with a gray value of 1 is defined as the foreground. Indicates with The number of foreground pixels is found in a group of 8 adjacent pixels; Indicated by Centered on the target pixel, the cumulative number of times the grayscale value of adjacent pixels changes from 0 to 1 in a clockwise direction. P 1 represents the current pixel. P 2. P 4. P 6 and P 8 represents P 1. Pixels directly above, below, to the left, and to the right.

4. The method for detecting the position of ice layer in a nuclear fusion target pellet according to claim 1, characterized in that, The step of removing skeleton branches from the refined skeleton diagram to obtain the skeleton trunk diagram includes: Based on the refined skeleton diagram, all endpoints of a single skeleton are obtained, wherein each endpoint must satisfy either having only one pixel value of 1 in its eight neighborhoods, or having only two pixel values ​​of 1 in its eight neighborhoods and the two pixels being adjacent. The two endpoints with the longest distance are selected and used as the beginning and end endpoints of the main skeleton. Based on the first and last endpoints, obtain the main skeleton; Set the first endpoint other than the first and last endpoints to zero, and then recursively go to the first bright spot in the eight neighborhoods of the first endpoint. If the first bright spot is not a node in the skeleton trunk, set the first bright spot to zero, until a node in the skeleton trunk is encountered, the recursion stops, and the skeleton trunk graph is obtained.

5. The method for detecting the location of ice layers in a nuclear fusion target pellet according to claim 1, characterized in that, The step of obtaining a binarized image from the X-ray image through noise reduction and adaptive binarization includes: The X-ray image is processed sequentially by Gaussian low-pass filtering, grayscale stretching, and image sharpening to obtain a noise-reduced X-ray image; The denoised X-ray image is processed by adaptive binarization to obtain a binarized image.

6. A device for detecting the position of ice layer in a nuclear fusion target pellet, characterized in that, The device includes: The image acquisition module is used to acquire X-ray images of the nuclear fusion target pellet; The binarization processing module is used to obtain a binarized image based on the X-ray image through noise reduction processing and adaptive binarization processing. The shell position acquisition module is used to acquire the shell position of the nuclear fusion target pellet based on the processed binarized image, including: Based on the binarized image, edge extraction is performed on the binarized image using a binarization function and the Sobel operator to obtain a binarized image containing edge information; Perform a Hough transform on the binarized image containing edge information to obtain the preliminary position of the spherical shell; Based on the initial position of the spherical shell, a first annular region within ±5% of the radius of the spherical shell is obtained. The pixels within the first annular region are then fitted with a circle using the least squares method to obtain the precise position of the spherical shell. The thinned skeleton acquisition module is used to process the binarized image based on the position of the spherical shell using a dilation algorithm and an image skeleton extraction algorithm to obtain a thinned skeleton map of the binarized image, including: Obtain the second annular region within a range of 0.7-0.8 times the radius of the spherical shell at the specified location; Based on the second annular region, the ice layer boundary is expanded outward by an expansion algorithm, and then a thinned skeleton map of the binarized image is obtained by an image skeleton extraction algorithm. The skeleton trunk acquisition module is used to remove skeleton branches from the refined skeleton diagram and obtain the skeleton trunk diagram. The ice layer location acquisition module is used to acquire the location of the middle ice layer of the X-ray image kernel fusion target pellet based on the skeleton backbone diagram.

7. A computer-readable storage medium, characterized in that, The storage medium stores a computer program for executing the method for detecting the location of ice layers in a nuclear fusion target pellet according to any one of claims 1-5.

8. A computer device, characterized in that, The method includes a processor, a memory, and a computer program stored in the memory and executable on the processor. When the processor executes the computer program, it implements the method for detecting the location of ice layers in a nuclear fusion target pellet as described in any one of claims 1-5.

Images