Data denoising method and device, computer device, storage medium and program product

Abstract

This application relates to a data denoising method, apparatus, computer equipment, storage medium, and program product. The method includes: acquiring a data matrix generated by scanning a target area on a chip surface at various frequencies; determining the electromagnetic image to be processed from the electromagnetic image corresponding to the data matrix at each frequency; dividing each pixel of the electromagnetic image to be processed into a preset number of pixel value intervals; determining a histogram noise variance estimate of the electromagnetic image to be processed based on the median and frequency corresponding to each pixel value interval; and performing transform-domain collaborative denoising processing on the electromagnetic image to be processed based on the histogram noise variance estimate to obtain the target electromagnetic image. This application calculates the histogram noise variance estimate using the median and frequency corresponding to multiple pixel value intervals and reduces noise across the entire frequency distribution through transform-domain collaborative denoising processing. This method can better extract weaker electromagnetic signals and improve the quality of the electromagnetic image.

Description

Technical Field

[0001] This application relates to the field of image processing, and in particular to a data noise reduction method, apparatus, computer equipment, storage medium, and program product. Background Technology

[0002] With the development of electronic manufacturing technologies such as integrated circuit processes, chips and circuit boards are moving towards higher integration and higher speeds, raising concerns about electromagnetic reliability. Due to the high integration and large scale of chips and circuits, traditional manual troubleshooting is difficult once a fault occurs. Near-field scanning-based electromagnetic interference image reconstruction is currently the most effective method for addressing electromagnetic compatibility issues. However, variations in the sensitivity and accuracy of near-field scanning probes result in the acquisition of data containing significant noise, posing considerable challenges to subsequent information processing and analysis. Since most of the near-field electromagnetic data on the chip surface consists of noise signals, noise reduction algorithms are typically used in electromagnetic data processing to separate the electromagnetic signals from the noise.

[0003] Traditional noise reduction algorithms typically operate in the spatial domain, performing data operations directly on the original image and processing the grayscale values ​​of pixels. Examples include mean denoising, median denoising, and blind denoising methods based on artificial intelligence.

[0004] However, noise is distributed across the entire frequency range. Traditional noise reduction algorithms only work well for noise within a specific frequency range. When the noise signal and the electromagnetic signal are of similar magnitude or the scanned electromagnetic signal is weak, traditional noise reduction algorithms cannot effectively separate the noise signal from the electromagnetic signal, resulting in poor electromagnetic image quality. Summary of the Invention

[0005] Therefore, it is necessary to provide a data noise reduction method, apparatus, computer equipment, storage medium, and program product that can improve the electromagnetic image effect in response to the above-mentioned technical problems.

[0006] In a first aspect, this application provides a data noise reduction method, the method comprising:

[0007] The data matrix generated after scanning the target area on the chip surface at various frequency points is obtained;

[0008] The electromagnetic image to be processed is determined from the electromagnetic image corresponding to the data matrix at each frequency point;

[0009] The pixels of the electromagnetic image to be processed are divided into a preset number of pixel value ranges;

[0010] The histogram noise variance estimate of the electromagnetic image to be processed is determined based on the median and frequency corresponding to each pixel value interval; the frequency is determined based on the number of pixels in the electromagnetic image to be processed within the corresponding pixel value interval.

[0011] Based on the noise variance estimation of the histogram, transform domain collaborative noise reduction is performed on the electromagnetic image to be processed to obtain the target electromagnetic image.

[0012] In one embodiment, determining the histogram noise variance estimate of the electromagnetic image to be processed based on the median and frequency corresponding to each pixel value interval includes:

[0013] The histogram mean of the electromagnetic image to be processed is determined based on the median and frequency corresponding to each pixel value interval;

[0014] The histogram noise variance estimate is determined based on the median corresponding to each pixel value interval, the histogram mean, and the frequency.

[0015] In one embodiment, determining the histogram mean of the electromagnetic image to be processed based on the median and frequency corresponding to each pixel value interval includes:

[0016] The mean of the histogram is the sum of the first product of the median and the corresponding frequency for each range of pixel values.

[0017] In one embodiment, determining the histogram noise variance estimate based on the median corresponding to each pixel value interval, the histogram mean, and the frequency includes:

[0018] Determine the squared value of the difference between the median of each pixel value interval and the mean of the histogram;

[0019] Determine the second product of the square value corresponding to each pixel value interval and the corresponding frequency;

[0020] The summation result obtained by summing the results of each of the second products is used as the histogram noise variance estimate.

[0021] In one embodiment, determining the electromagnetic image to be processed from the electromagnetic images corresponding to the data matrix at each frequency point includes:

[0022] Based on the mean of each row element in the data matrix at each frequency point, determine the target row variance of each row element in the data matrix at each frequency point.

[0023] Based on the mean of each column element in the data matrix at each frequency point, determine the target column variance of each column element in the data matrix at each frequency point.

[0024] The target variance is determined based on the target row variance and target column variance corresponding to the data matrix at each frequency point.

[0025] The data matrix corresponding to the target variance that is greater than or equal to the preset variance threshold is taken as the target data matrix, and the electromagnetic image corresponding to the target data matrix is ​​taken as the electromagnetic image to be processed.

[0026] In one embodiment, the process of performing transform-domain collaborative denoising on the electromagnetic image to be processed based on the histogram noise variance estimation to obtain the target electromagnetic image includes:

[0027] Identify the similar blocks corresponding to each block in the electromagnetic image to be processed;

[0028] Based on each block in the electromagnetic image to be processed and its corresponding similar blocks, determine the basic estimated value of the electromagnetic image to be processed.

[0029] Based on the basic estimate and the histogram noise variance estimate, transform domain collaborative noise reduction is performed on the electromagnetic image to be processed to obtain the target electromagnetic image.

[0030] Secondly, this application provides a data noise reduction apparatus, the apparatus comprising:

[0031] The acquisition module is used to acquire the data matrix generated after scanning the target area on the chip surface at various frequency points;

[0032] The first determining module is used to determine the electromagnetic image to be processed from the electromagnetic image corresponding to the data matrix at each frequency point;

[0033] The segmentation module is used to divide each pixel of the electromagnetic image to be processed into a preset number of pixel value ranges;

[0034] The second determining module is used to determine the histogram noise variance estimate of the electromagnetic image to be processed based on the median and frequency corresponding to each pixel value interval; the frequency is determined based on the number of pixels of the electromagnetic image to be processed in the corresponding pixel value interval.

[0035] The processing module is used to perform transform domain collaborative noise reduction processing on the electromagnetic image to be processed based on the noise variance estimation of the histogram, so as to obtain the target electromagnetic image.

[0036] Thirdly, this application provides a computer device including a memory and a processor, the memory storing a computer program, and the processor executing the computer program to implement the steps of the above-described method.

[0037] Fourthly, this application provides a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the steps of the above-described method.

[0038] Fifthly, this application provides a computer program product comprising a computer program that, when executed by a processor, implements the steps of the above-described method.

[0039] The aforementioned data denoising method, apparatus, computer equipment, storage medium, and program product first involve the computer equipment acquiring a data matrix generated by scanning the target area on the chip surface at various frequency points. Then, the electromagnetic image to be processed is determined from the electromagnetic image corresponding to the data matrix at each frequency point. Each pixel of the electromagnetic image to be processed is divided into a preset number of pixel value intervals. Based on the median and frequency corresponding to each pixel value interval, a histogram noise variance estimate of the electromagnetic image to be processed is determined. Finally, based on the histogram noise variance estimate, transform-domain collaborative denoising processing is performed on the electromagnetic image to be processed to obtain the target electromagnetic image. Traditional techniques perform data denoising processing in the spatial domain. Since this application can calculate the histogram noise variance estimate using the median and frequency corresponding to multiple pixel value intervals, and transforms the signal from the spatial domain to the frequency domain for denoising through transform-domain collaborative denoising processing, it can denoise the noise across the entire frequency distribution. Therefore, this method can better extract weaker electromagnetic signals and improve the quality of the electromagnetic image. Attached Figure Description

[0040] Figure 1 This is a schematic flowchart of a data noise reduction method provided in an embodiment of this application;

[0041] Figure 2 One of the flowcharts for a method to determine the variance of histogram noise provided in an embodiment of this application;

[0042] Figure 3 A second schematic flowchart illustrating a method for determining histogram noise variance provided in an embodiment of this application;

[0043] Figure 4 A flowchart illustrating a method for determining an electromagnetic image to be processed, provided in an embodiment of this application;

[0044] Figure 5 A flowchart illustrating a target electromagnetic image determination method provided in an embodiment of this application;

[0045] Figure 6 This is a structural block diagram of a data noise reduction device in one embodiment;

[0046] Figure 7 This is an internal structural diagram of a computer device in one embodiment. Detailed Implementation

[0047] To make the objectives, technical solutions, and advantages of this application clearer, the following detailed description is provided in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the scope of this application.

[0048] This application provides a data noise reduction method, referring to... Figure 1 , Figure 1 This is a flowchart illustrating a data noise reduction method provided in an embodiment of this application. The method includes the following steps:

[0049] S101. Obtain the data matrix generated after scanning the target area on the chip surface at various frequency points.

[0050] For example, the target region is a region of size m×n on the surface of the chip.

[0051] Specifically, a spectrum analyzer combined with an electromagnetic field probe is used to perform a serpentine scan of an m×n region on the chip surface. The computer then acquires the data matrix generated after scanning the m×n region at various frequency points. Assuming there are N frequency points, each frequency point corresponds to such a data matrix. The data matrix for a single frequency point in this region can be represented as:

[0052]

[0053] Among them, a mn It is an element in the data matrix of that region.

[0054] S102. Determine the electromagnetic image to be processed from the electromagnetic images corresponding to the data matrix at each frequency point.

[0055] Specifically, the computer equipment calculates the target variance of the data matrix at each frequency point and compares it with a preset variance threshold. The data matrix corresponding to the target variance greater than or equal to the preset variance threshold is taken as the target data matrix. Finally, the target data matrix is ​​used as the pixel values ​​of a grayscale image, and MATLAB or other graphics libraries are called for visualization. After visualization, an electromagnetic image is obtained, which is the electromagnetic image to be processed.

[0056] S103. Divide each pixel of the electromagnetic image to be processed into a preset number of pixel value ranges.

[0057] Specifically, the electromagnetic image to be processed can be treated as a grayscale image. If it uses 8 bits for storage, then there are 256 grayscale levels, or 256 pixel values. Assuming there are L pixel value intervals, each pixel within a given interval is counted as one. The computer device can then divide the pixels of the electromagnetic image to be processed into these L pixel value intervals. For example, if there are four pixel value intervals: [0-20], [20-100], [100-200], and [200-255], pixels with a value of 15 can be assigned to [0-20], pixels with a value of 32 can be assigned to [20-100], and so on.

[0058] S104. Determine the histogram noise variance estimate of the electromagnetic image to be processed based on the median and frequency corresponding to each pixel value interval; the frequency is determined based on the number of pixels in the electromagnetic image to be processed within the corresponding pixel value interval.

[0059] Specifically, computer equipment can calculate the histogram mean of the image to be processed based on the median and frequency corresponding to each pixel value interval. The histogram mean of the electromagnetic image to be processed can then be used to determine the histogram noise variance estimate. The frequency is the number of pixels in the electromagnetic image to be processed that fall within a corresponding pixel value interval. For example, there are four pixel value intervals: [0-20], [20-100], etc.

[0060] [100-200], [200-255], assuming there are pixels with values ​​of 4, 5, 10, 15, and 18 in the pixel value range [0-20], then the pixel with a value of 10 is the median of the pixel value range [0-20], and the frequency of the pixel value range is 5. The histogram mean of the image to be processed can be determined by the histogram mean expression, and the histogram noise variance estimate of the electromagnetic image to be processed can be determined based on the calculated histogram mean.

[0061] The expression for estimating the histogram noise variance of the electromagnetic image to be processed is as follows:

[0062]

[0063] Among them, z i p(z) represents the pixel corresponding to the median of each pixel value interval. i ) represents the frequency corresponding to each pixel value range.

[0064] S105. Based on the histogram noise variance estimation, perform transform domain collaborative noise reduction on the electromagnetic image to be processed to obtain the target electromagnetic image.

[0065] The basic idea of ​​collaborative noise reduction is to find similar blocks in the image and then perform a series of noise reduction processes.

[0066] Specifically, firstly, the electromagnetic image to be processed is input into the computer device, and a reference block P and a search window are selected. Based on this search window, similar blocks Q that are similar to the reference block P can be found. Assuming n reference blocks P are selected, each of these n reference blocks P will obtain a matching similar block Q using the above search method. Secondly, each reference block P and its corresponding similar block Q in the electromagnetic image to be processed are stacked, and a discrete cosine transform is performed. Then, a hard thresholding operation is performed on the coefficients obtained after the transform, followed by an inverse transform. The blocks obtained after the inverse transform are then placed back into their previously searched positions. Finally, a weighted average is calculated for all obtained overlapping block estimates to determine the basic estimate of the electromagnetic image to be processed. Finally, based on the basic estimate of the electromagnetic image to be processed and the histogram noise variance estimate obtained in S104, Wiener denoising is performed on the electromagnetic image to obtain a local estimate. A weighted average is then used to aggregate all obtained local estimates to calculate the final estimate of the electromagnetic image to be processed, thus obtaining the denoised target electromagnetic image.

[0067] The data denoising method provided in this application first involves a computer device acquiring a data matrix generated by scanning a target area on the chip surface at various frequency points. Then, the electromagnetic image to be processed is determined from the electromagnetic images corresponding to the data matrices at each frequency point. Each pixel of the electromagnetic image to be processed is divided into a preset number of pixel value intervals. The histogram noise variance estimate of the electromagnetic image to be processed is determined based on the median and frequency corresponding to each pixel value interval. Finally, based on the histogram noise variance estimate, transform-domain collaborative denoising processing is performed on the electromagnetic image to be processed to obtain the target electromagnetic image. Traditional techniques perform data denoising processing in the spatial domain. Since this application can calculate the histogram noise variance estimate using the median and frequency corresponding to multiple pixel value intervals, and transforms the signal from the spatial domain to the frequency domain for denoising, it can denoise the noise across the entire frequency distribution. Therefore, this method can better extract weaker electromagnetic signals and improve the quality of the electromagnetic image.

[0068] Figure 2 This is one of the flowcharts illustrating a method for determining histogram noise variance estimation according to an embodiment of this application. This embodiment relates to a possible implementation of how to determine the histogram noise variance estimation of an electromagnetic image to be processed based on the median and frequency corresponding to each pixel value interval. Based on the above embodiment, S104 includes:

[0069] S201. Determine the histogram mean of the electromagnetic image to be processed based on the median and frequency corresponding to each pixel value interval.

[0070] The frequency is the number of pixels in the electromagnetic image to be processed within the corresponding pixel interval.

[0071] Specifically, there are L pixel value intervals. The computer device can calculate the histogram mean of the electromagnetic image to be processed based on the median of each of the L pixel value intervals and the number of pixels of the electromagnetic image to be processed within the corresponding pixel intervals. The expression for calculating the histogram mean of the image to be processed is as follows:

[0072]

[0073] z in the expression i p(z) represents the pixel corresponding to the median of each pixel value interval. i This represents the frequency corresponding to each pixel value range. This represents the histogram mean of the image to be processed.

[0074] S202. Determine the histogram noise variance estimate based on the median, histogram mean, and frequency corresponding to each pixel value interval.

[0075] Specifically, the computer device calculates the histogram noise variance estimate based on the median of each of the L pixel value intervals, the histogram mean of the image to be processed calculated in S201, and the number of pixels in the electromagnetic image to be processed within the corresponding pixel intervals, using the following formula:

[0076]

[0077] z in the expression i p(z) represents the pixel corresponding to the median of each pixel value interval. i ) represents the frequency corresponding to each pixel value interval, and m represents the histogram mean of the image to be processed.

[0078] The histogram noise variance estimation method provided in this application first determines the histogram mean of the electromagnetic image to be processed based on the median and frequency corresponding to each pixel value interval. Then, it determines the histogram noise variance estimate based on the median, histogram mean, and frequency corresponding to each pixel value interval. Since the purpose of electromagnetic image denoising is to extract electromagnetic signals as much as possible and reduce the influence of noise signals on electromagnetic signals, and effective denoising requires prior information about the noise generation mechanism and distribution characteristics, such as the noise mean and variance, reasonable and scientific assumptions are made about the noise based on this prior information. Therefore, by calculating the histogram noise variance estimate using the median and frequency corresponding to multiple pixel value intervals, electromagnetic signals and noise signals can be better separated, improving the electromagnetic image quality.

[0079] In one embodiment, this embodiment relates to a possible implementation of how to determine the histogram mean of the electromagnetic image to be processed based on the median and frequency corresponding to each pixel value interval. Based on the above embodiment, S201 includes:

[0080] The mean of the histogram is the sum of the first product of the median and the corresponding frequency for each pixel value interval.

[0081] Specifically, the computer device multiplies the median of each of the L pixel value intervals by the frequency of that interval, using the result as the first product. Then, the sum of these first products across all pixel value intervals yields the histogram mean of the electromagnetic image to be processed. It should be noted that alternatively, the histogram mean of the electromagnetic image can be obtained by multiplying the first product of each pixel value interval by its corresponding scaling factor and then summing the results.

[0082] The method for determining the histogram mean of an electromagnetic image to be processed provided in this application embodiment is to take the histogram mean of the sum of the first product of the median and the corresponding frequency of each pixel value interval. Since the histogram noise variance estimate can be determined through the histogram mean, and the histogram noise variance estimate serves as prior information for making reasonable and scientific assumptions about the noise, it is possible to better separate electromagnetic signals and noise signals, thereby improving the quality of the electromagnetic image.

[0083] Figure 3 This is a second flowchart illustrating a method for determining histogram noise variance estimation provided in this application embodiment. This embodiment relates to a possible implementation of how to determine histogram noise variance estimation based on the median, histogram mean, and frequency corresponding to each pixel value interval. Based on the above embodiment, S202 includes:

[0084] S301. Determine the squared value of the difference between the median and the mean of the histogram for each pixel value interval.

[0085] Specifically, referring to the formula in S202, the computer device determines the median of each pixel value interval in the L pixel value intervals, calculates the difference between the median and the mean of the histogram, and then calculates the square of the difference corresponding to each pixel value interval.

[0086] S302. Determine the second product of the square value and the corresponding frequency for each pixel value interval.

[0087] Specifically, referring to the formula in S202, the computer device multiplies the square of the difference between each pixel value interval obtained in the above steps with the frequency corresponding to each pixel value interval to obtain the second product result corresponding to each pixel value interval.

[0088] S303. The summation result obtained by summing the results of each second product is used as the histogram noise variance estimate.

[0089] Specifically, referring to the formula in S202, the computer device sums the second product results corresponding to each pixel value interval obtained in the above steps, and the summation result is used as the histogram noise variance estimate.

[0090] The method for determining histogram noise variance estimation provided in this application first involves a computer device determining the squared difference between the median and the histogram mean for each pixel value interval. Then, it determines the second product of the squared difference and the corresponding frequency for each pixel value interval. Finally, the summation of these second products is used as the histogram noise variance estimate. Calculating the histogram noise variance estimate using the median and histogram mean across multiple pixel value intervals allows for better extraction of electromagnetic signals and improves the quality of electromagnetic images.

[0091] Figure 4 This is a flowchart illustrating a method for determining an electromagnetic image to be processed, provided in an embodiment of this application. This embodiment relates to a possible implementation of how to determine the electromagnetic image to be processed from the electromagnetic images corresponding to the data matrix at each frequency point. Based on the above embodiment, S102 includes:

[0092] S401. Based on the mean of each row element in the data matrix at each frequency point, determine the target row variance of each row element in the data matrix at each frequency point.

[0093] Specifically, since the electromagnetic data obtained from near-field scanning of the chip surface is a two-dimensional data matrix with the same number of points as the length and width of the scanning area, and this data matrix contains spatial information, traditional variance calculation algorithms cannot be used directly; spatial information must also be considered. This algorithm considers information from both the row and column scanning directions. The computer device calculates the variance of the data matrix along the row direction based on the mean of each row element in the data matrix at each frequency point. The specific formula is as follows:

[0094]

[0095] Where, θ i It is the mean of the i-th row, n is the number of columns in the data matrix, and a ik Let be the element in the i-th row and k-th column of the matrix. The variance calculated using the above formula is then used as a one-dimensional vector to further calculate the variance, yielding the target row variance, as shown in the following formula:

[0096]

[0097]

[0098] Where, θ row It is all υ row (i) is the mean of the data matrix, where m is the number of rows in the data matrix.

[0099] S402. Based on the mean values ​​of each column element in the data matrix at each frequency point, determine the target column variance corresponding to each column element in the data matrix at each frequency point.

[0100] Specifically, similar to the method for calculating the target row variance described above, the computer device calculates the variance of the data matrix along the column direction based on the mean of each column element in the data matrix at each frequency point. The method for calculating the target column variance is expressed as follows:

[0101]

[0102] Where θ j It is the mean of the j-th column, m is the number of rows in the data matrix, and a kj Let be the element in the k-th row and j-th column of the matrix. The variance calculated using the above formula is then used as a one-dimensional vector to further calculate the variance of the target column, as shown in the following formula:

[0103]

[0104]

[0105] Where θ col It is all υ col The mean of (k), where n is the number of columns in the data matrix.

[0106] S403. Determine the target variance based on the target row variance and target column variance corresponding to the data matrix at each frequency point.

[0107] Specifically, the computer equipment calculates the root mean square of the target row variance and target column variance based on the target row variance and target column variance corresponding to the data matrix at each frequency point. The result of the root mean square is the target variance, and the specific formula is as follows:

[0108]

[0109] Where, δ i υ represents the target variance. row υ represents the target row variance. col This represents the target column variance. It should be noted that you can also multiply the target row variance by the corresponding preset coefficient, multiply the target column variance by the corresponding preset coefficient, and then calculate the root mean square of the target row and column variances. The result of the root mean square is the target variance.

[0110] S404. Take the data matrix corresponding to the target variance that is greater than or equal to the preset variance threshold as the target data matrix, and take the electromagnetic image corresponding to the target data matrix as the electromagnetic image to be processed.

[0111] Specifically, each data matrix corresponds to a target variance. A preset variance threshold is set manually. The target variance obtained in the above steps is compared with the preset variance threshold. The data matrix corresponding to the target variance that is greater than or equal to the preset variance threshold is taken as the target data matrix. The electromagnetic image obtained by visualizing the target data matrix is ​​taken as the electromagnetic image to be processed.

[0112] The method for determining the electromagnetic image to be processed provided in this application first determines the target row variance based on the mean of each row element in the data matrix at each frequency point. Then, it determines the target column variance based on the mean of each column element in the data matrix at each frequency point. Finally, it determines the target variance based on the target row and column variances of the data matrix at each frequency point. The data matrix with a target variance greater than or equal to a preset variance threshold is taken as the target data matrix, and the electromagnetic image corresponding to the target data matrix is ​​taken as the electromagnetic image to be processed. This application determines the target data matrix by calculating the target variance and visualizes the target data matrix to obtain the electromagnetic image to be processed. This allows for the initial filtering of noise signals from the original electromagnetic data, thus facilitating further noise reduction processing.

[0113] Figure 5 This is a flowchart illustrating a target electromagnetic image determination method provided in an embodiment of this application. This embodiment relates to a possible implementation of how to perform transform-domain collaborative denoising processing on the electromagnetic image to be processed based on histogram noise variance estimation to obtain the target electromagnetic image. Based on the above embodiment, step S105 includes:

[0114] S501. Determine the similar blocks corresponding to each block in the electromagnetic image to be processed.

[0115] Specifically, first, the electromagnetic image to be processed is input into the computer device, and a reference block P and a search window are selected. The size of the search window can be assumed to be k×k. The reference block P and the search window are selected based on the experimental results, and the size of the search window is usually an odd number. Based on this search window, similar blocks Q that are similar to the reference block P can be found. Assuming that n reference blocks P are selected, then each of these n reference blocks P will obtain a matching similar block Q according to the above search method. The condition for selecting similar blocks Q can be expressed as:

[0116]

[0117] In the expression, t is a preset threshold. Let P denote the square of the 2-norm, P denote the specified reference block, and Q denote the similar block that is similar to it.

[0118] S502. Determine the basic estimated value of the electromagnetic image to be processed based on each block in the electromagnetic image to be processed and the corresponding similar block.

[0119] Specifically, each reference block P and its corresponding similar block Q in the electromagnetic image to be processed are stacked into a three-dimensional array. A discrete cosine transform is applied to the formed three-dimensional array, and the coefficients after the discrete cosine transform are hard-thresholded. Then, an inverse three-dimensional transform is performed to obtain estimates of all grouped blocks, which are then returned to their initial positions. Finally, a weighted average is calculated for all obtained overlapping blocks to obtain the basic estimate of the electromagnetic image to be processed. The specific formula can be expressed as:

[0120]

[0121] Where, when pixel x∈Q, χQ(x)=1, otherwise χQ(x)=0. The reference block is P. This represents the hard thresholding operation, and for pixel x∈Q, the estimated value of the pixel after collaborative filtering and inverse 3D transformation. Because similar blocks may overlap during acquisition, the base estimate u... basic (x) is to perform a weighted average of each pixel of all similar blocks according to the above formula to obtain the basic estimate of the electromagnetic image to be processed.

[0122] S503. Based on the basic estimated value and histogram noise variance estimate, perform transform domain collaborative noise reduction on the electromagnetic image to be processed to obtain the target electromagnetic image.

[0123] Specifically, based on the baseline estimate and histogram noise variance estimate of the electromagnetic image to be processed, Wiener denoising is performed to obtain local estimates. Then, a weighted average is used to aggregate all obtained local estimates to calculate the final estimate of the electromagnetic image to be processed, thus obtaining the denoised target electromagnetic image. The formula for Wiener denoising can be expressed as:

[0124]

[0125] in, After the 3D transformation is completed, Wiener filtering is performed on the coefficients of the 3D transformation, ρ basic (P) is the set of similar blocks. This is a histogram noise variance estimation. For each reference block P, Wiener denoising is used, and then a weighted average is used to aggregate all the obtained local estimates to calculate the final estimate of the electromagnetic image to be processed. The formula for the final estimate can be expressed as:

[0126]

[0127] Similar to the baseline estimate, where The final estimate of u is the square of the inverse of the 2-norm of the Wiener noise reduction coefficients. final (x) is to perform a weighted average of each pixel in all similar blocks according to the above formula, thereby obtaining the denoised target electromagnetic image.

[0128] The method for determining a target electromagnetic image provided in this application first determines the similar blocks corresponding to each block in the electromagnetic image to be processed. Then, based on each block in the electromagnetic image to be processed and its corresponding similar blocks, a basic estimate of the electromagnetic image to be processed is determined. Finally, based on the basic estimate and histogram noise variance estimation, transform domain collaborative denoising processing is performed on the electromagnetic image to be processed to obtain the target electromagnetic image. Since sensor probe measurement devices inevitably introduce noise signals, the transform domain collaborative denoising processing operation based on histogram noise variance estimation in this method can better extract weaker electromagnetic signals, thereby improving the quality of the electromagnetic image.

[0129] It should be understood that although the steps in the flowcharts of the embodiments described above are shown sequentially according to the arrows, these steps are not necessarily executed in the order indicated by the arrows. Unless explicitly stated herein, there is no strict order restriction on the execution of these steps, and they can be executed in other orders. Moreover, at least some steps in the flowcharts of the embodiments described above may include multiple steps or multiple stages. These steps or stages are not necessarily completed at the same time, but can be executed at different times. The execution order of these steps or stages is not necessarily sequential, but can be performed alternately or in turn with other steps or at least some of the steps or stages of other steps.

[0130] Based on the same inventive concept, this application also provides a data denoising apparatus for implementing the data denoising method described above. The solution provided by this apparatus is similar to the implementation scheme described in the above method; therefore, the specific limitations in one or more data denoising apparatus embodiments provided below can be found in the limitations of the data denoising method described above, and will not be repeated here.

[0131] In one embodiment, such as Figure 6As shown, a data noise reduction device 600 is provided, comprising: an acquisition module 601, a first determination module 602, a division module 603, a second determination module 604, and a processing module 605, wherein:

[0132] The acquisition module 601 is used to acquire the data matrix generated after scanning the target area on the chip surface at various frequency points.

[0133] The first determining module 602 is used to determine the electromagnetic image to be processed from the electromagnetic image corresponding to the data matrix at each frequency point.

[0134] The segmentation module 603 is used to divide each pixel of the electromagnetic image to be processed into a preset number of pixel value ranges.

[0135] The second determining module 604 is used to determine the histogram noise variance estimate of the electromagnetic image to be processed based on the median and frequency corresponding to each pixel value interval; the frequency is determined based on the number of pixels of the electromagnetic image to be processed in the corresponding pixel value interval.

[0136] The processing module 605 is used to perform transform domain collaborative noise reduction processing on the electromagnetic image to be processed based on the histogram noise variance estimation to obtain the target electromagnetic image.

[0137] In one embodiment, the second determining module 604 includes:

[0138] The first determining unit is used to determine the histogram mean of the electromagnetic image to be processed based on the median and frequency corresponding to each pixel value interval.

[0139] The second determining unit is used to determine the histogram noise variance estimate based on the median, histogram mean, and frequency corresponding to each pixel value interval.

[0140] In one embodiment, the first determining unit is specifically used to calculate the mean of the histogram of the sum of the first product of the median and the corresponding frequency for each pixel value interval.

[0141] In one embodiment, the second determining unit is specifically used to determine the squared value of the difference between the median and the histogram mean corresponding to each pixel value interval; determine the second product result of the squared value and the corresponding frequency corresponding to each pixel value interval; and use the summation result obtained by summing each second product result as the histogram noise variance estimate.

[0142] In one embodiment, the first determining module 602 includes:

[0143] The third determining unit is used to determine the target row variance corresponding to each row element in the data matrix at each frequency point based on the mean value corresponding to each row element in the data matrix at each frequency point.

[0144] The fourth determining unit is used to determine the target column variance corresponding to each column element in the data matrix at each frequency point based on the mean value corresponding to each column element in the data matrix at each frequency point.

[0145] The fifth determining unit is used to determine the target variance based on the target row variance and target column variance corresponding to the data matrix at each frequency point.

[0146] The sixth determining unit is used to take the data matrix corresponding to the target variance that is greater than or equal to the preset variance threshold as the target data matrix, and take the electromagnetic image corresponding to the target data matrix as the electromagnetic image to be processed.

[0147] In one embodiment, the processing module 605 includes:

[0148] The seventh determining unit is used to determine the similar blocks corresponding to each block in the electromagnetic image to be processed;

[0149] The eighth determining unit is used to determine the basic estimated value of the electromagnetic image to be processed based on each block in the electromagnetic image to be processed and the corresponding similar block.

[0150] The processing unit is used to perform transform domain collaborative noise reduction processing on the electromagnetic image to be processed based on the basic estimate and the histogram noise variance estimate to obtain the target electromagnetic image.

[0151] Each module in the aforementioned data noise reduction device can be implemented entirely or partially through software, hardware, or a combination thereof. These modules can be embedded in the processor of a computer device in hardware form or independent of it, or stored in the memory of a computer device in software form, so that the processor can call and execute the operations corresponding to each module.

[0152] In one embodiment, a computer device is provided, which may be a server, and its internal structure diagram may be as follows: Figure 7 As shown, the computer device includes a processor, memory, and a network interface connected via a system bus. The processor provides computing and control capabilities. The memory includes non-volatile storage media and internal memory. The non-volatile storage media stores the operating system, computer programs, and a database. The internal memory provides an environment for the operation of the operating system and computer programs stored in the non-volatile storage media. The database stores data. The network interface communicates with external terminals via a network connection. When executed by the processor, the computer program implements a data noise reduction method.

[0153] Those skilled in the art will understand that Figure 7The structure shown is merely a block diagram of a portion of the structure related to the present application and does not constitute a limitation on the computer device to which the present application is applied. Specific computer devices may include more or fewer components than those shown in the figure, or combine certain components, or have different component arrangements.

[0154] In one embodiment, a computer device is provided, including a memory and a processor, wherein the memory stores a computer program, and the processor executes the computer program to perform the following steps:

[0155] The data matrix generated after scanning the target area on the chip surface at various frequency points is obtained;

[0156] The electromagnetic image to be processed is determined from the electromagnetic image corresponding to the data matrix at each frequency point;

[0157] Divide each pixel of the electromagnetic image to be processed into a preset number of pixel value ranges;

[0158] The histogram noise variance estimate of the electromagnetic image to be processed is determined based on the median and frequency corresponding to each pixel value interval; the frequency is determined based on the number of pixels in the electromagnetic image to be processed within the corresponding pixel value interval.

[0159] Based on histogram noise variance estimation, transform domain collaborative noise reduction is performed on the electromagnetic image to be processed to obtain the target electromagnetic image.

[0160] In one embodiment, the processor, when executing a computer program, also performs the following steps:

[0161] The histogram mean of the electromagnetic image to be processed is determined based on the median and frequency corresponding to each pixel value interval;

[0162] The histogram noise variance estimate is determined based on the median, histogram mean, and frequency corresponding to each pixel value interval.

[0163] In one embodiment, the processor, when executing a computer program, also performs the following steps:

[0164] The mean of the histogram is the sum of the first product of the median and the corresponding frequency for each pixel value interval.

[0165] In one embodiment, the processor, when executing a computer program, also performs the following steps:

[0166] Determine the squared value of the difference between the median and the mean of the histogram for each pixel value interval;

[0167] Determine the second product of the square value corresponding to each pixel value interval and the corresponding frequency;

[0168] The summation result obtained by summing the results of each second product is used as the histogram noise variance estimate.

[0169] In one embodiment, the processor, when executing a computer program, also performs the following steps:

[0170] Based on the mean of each row element in the data matrix at each frequency point, determine the target row variance of each row element in the data matrix at each frequency point;

[0171] Based on the mean of each column element in the data matrix at each frequency point, determine the target column variance of each column element in the data matrix at each frequency point;

[0172] The target variance is determined based on the target row variance and target column variance corresponding to the data matrix at each frequency point;

[0173] The data matrix corresponding to the target variance that is greater than or equal to the preset variance threshold is taken as the target data matrix, and the electromagnetic image corresponding to the target data matrix is ​​taken as the electromagnetic image to be processed.

[0174] In one embodiment, the processor, when executing a computer program, also performs the following steps:

[0175] Identify the similar blocks corresponding to each block in the electromagnetic image to be processed;

[0176] Based on each block in the electromagnetic image to be processed and its corresponding similar blocks, determine the basic estimated value of the electromagnetic image to be processed;

[0177] Based on the baseline estimates and histogram noise variance estimates, transform domain collaborative noise reduction is performed on the electromagnetic image to be processed to obtain the target electromagnetic image.

[0178] 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:

[0179] The data matrix generated after scanning the target area on the chip surface at various frequency points is obtained;

[0180] The electromagnetic image to be processed is determined from the electromagnetic image corresponding to the data matrix at each frequency point;

[0181] Divide each pixel of the electromagnetic image to be processed into a preset number of pixel value ranges;

[0182] The histogram noise variance estimate of the electromagnetic image to be processed is determined based on the median and frequency corresponding to each pixel value interval; the frequency is determined based on the number of pixels in the electromagnetic image to be processed within the corresponding pixel value interval.

[0183] Based on histogram noise variance estimation, transform domain collaborative noise reduction is performed on the electromagnetic image to be processed to obtain the target electromagnetic image.

[0184] In one embodiment, when the computer program is executed by a processor, it also performs the following steps:

[0185] The histogram mean of the electromagnetic image to be processed is determined based on the median and frequency corresponding to each pixel value interval;

[0186] The histogram noise variance estimate is determined based on the median, histogram mean, and frequency corresponding to each pixel value interval.

[0187] In one embodiment, when the computer program is executed by a processor, it also performs the following steps:

[0188] The mean of the histogram is the sum of the first product of the median and the corresponding frequency for each pixel value interval.

[0189] In one embodiment, when the computer program is executed by a processor, it also performs the following steps:

[0190] Determine the squared value of the difference between the median and the mean of the histogram for each pixel value interval;

[0191] Determine the second product of the square value corresponding to each pixel value interval and the corresponding frequency;

[0192] The summation result obtained by summing the results of each second product is used as the histogram noise variance estimate.

[0193] In one embodiment, when the computer program is executed by a processor, it also performs the following steps:

[0194] Based on the mean of each row element in the data matrix at each frequency point, determine the target row variance of each row element in the data matrix at each frequency point;

[0195] Based on the mean of each column element in the data matrix at each frequency point, determine the target column variance of each column element in the data matrix at each frequency point;

[0196] The target variance is determined based on the target row variance and target column variance corresponding to the data matrix at each frequency point;

[0197] The data matrix corresponding to the target variance that is greater than or equal to the preset variance threshold is taken as the target data matrix, and the electromagnetic image corresponding to the target data matrix is ​​taken as the electromagnetic image to be processed.

[0198] In one embodiment, when the computer program is executed by a processor, it also performs the following steps:

[0199] Identify the similar blocks corresponding to each block in the electromagnetic image to be processed;

[0200] Based on each block in the electromagnetic image to be processed and its corresponding similar blocks, determine the basic estimated value of the electromagnetic image to be processed;

[0201] Based on the baseline estimates and histogram noise variance estimates, transform domain collaborative noise reduction is performed on the electromagnetic image to be processed to obtain the target electromagnetic image.

[0202] In one embodiment, a computer program product is provided, including a computer program that, when executed by a processor, performs the following steps:

[0203] The data matrix generated after scanning the target area on the chip surface at various frequency points is obtained;

[0204] The electromagnetic image to be processed is determined from the electromagnetic image corresponding to the data matrix at each frequency point;

[0205] Divide each pixel of the electromagnetic image to be processed into a preset number of pixel value ranges;

[0206] The histogram noise variance estimate of the electromagnetic image to be processed is determined based on the median and frequency corresponding to each pixel value interval; the frequency is determined based on the number of pixels in the electromagnetic image to be processed within the corresponding pixel value interval.

[0207] Based on histogram noise variance estimation, transform domain collaborative noise reduction is performed on the electromagnetic image to be processed to obtain the target electromagnetic image.

[0208] In one embodiment, when the computer program is executed by a processor, it also performs the following steps:

[0209] The histogram mean of the electromagnetic image to be processed is determined based on the median and frequency corresponding to each pixel value interval;

[0210] The histogram noise variance estimate is determined based on the median, histogram mean, and frequency corresponding to each pixel value interval.

[0211] In one embodiment, when the computer program is executed by a processor, it also performs the following steps:

[0212] The mean of the histogram is the sum of the first product of the median and the corresponding frequency for each pixel value interval.

[0213] In one embodiment, when the computer program is executed by a processor, it also performs the following steps:

[0214] Determine the squared value of the difference between the median and the mean of the histogram for each pixel value interval;

[0215] Determine the second product of the square value corresponding to each pixel value interval and the corresponding frequency;

[0216] The summation result obtained by summing the results of each second product is used as the histogram noise variance estimate.

[0217] In one embodiment, when the computer program is executed by a processor, it also performs the following steps:

[0218] Based on the mean of each row element in the data matrix at each frequency point, determine the target row variance of each row element in the data matrix at each frequency point;

[0219] Based on the mean of each column element in the data matrix at each frequency point, determine the target column variance of each column element in the data matrix at each frequency point;

[0220] The target variance is determined based on the target row variance and target column variance corresponding to the data matrix at each frequency point;

[0221] The data matrix corresponding to the target variance that is greater than or equal to the preset variance threshold is taken as the target data matrix, and the electromagnetic image corresponding to the target data matrix is ​​taken as the electromagnetic image to be processed.

[0222] In one embodiment, when the computer program is executed by a processor, it also performs the following steps:

[0223] Identify the similar blocks corresponding to each block in the electromagnetic image to be processed;

[0224] Based on each block in the electromagnetic image to be processed and its corresponding similar blocks, determine the basic estimated value of the electromagnetic image to be processed;

[0225] Based on the baseline estimates and histogram noise variance estimates, transform domain collaborative noise reduction is performed on the electromagnetic image to be processed to obtain the target electromagnetic image.

[0226] It should be noted that the user information (including but not limited to user device information, user personal information, etc.) and data (including but not limited to data used for analysis, data stored, data displayed, etc.) involved in this application are all information and data authorized by the user or fully authorized by all parties.

[0227] Those skilled in the art will understand that all or part of the processes in the methods of the above embodiments can be implemented by a computer program instructing related hardware. The computer program can be stored in a non-volatile computer-readable storage medium, and when executed, it can include the processes of the embodiments of the above methods. Any references to memory, databases, or other media used in the embodiments provided in this application can include at least one of non-volatile and volatile memory. Non-volatile memory can include read-only memory (ROM), magnetic tape, floppy disk, flash memory, optical memory, high-density embedded non-volatile memory, resistive random access memory (ReRAM), magnetic random access memory (MRAM), ferroelectric random access memory (FRAM), phase change memory (PCM), graphene memory, etc. Volatile memory can include random access memory (RAM) or external cache memory, etc. By way of illustration and not limitation, RAM can take many forms, such as Static Random Access Memory (SRAM) or Dynamic Random Access Memory (DRAM). The databases involved in the embodiments provided in this application may include at least one type of relational database and non-relational database. Non-relational databases may include, but are not limited to, blockchain-based distributed databases. The processors involved in the embodiments provided in this application may be general-purpose processors, central processing units, graphics processing units, digital signal processors, programmable logic devices, quantum computing-based data processing logic devices, etc., and are not limited to these.

[0228] The technical features of the above embodiments can be combined in any way. For the sake of brevity, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0229] The embodiments described above are merely illustrative of several implementation methods of this application, and while the descriptions are specific and detailed, they should not be construed as limiting the scope of this patent application. It should be noted that those skilled in the art can make various modifications and improvements without departing from the concept of this application, and these all fall within the protection scope of this application. Therefore, the protection scope of this application should be determined by the appended claims.

Claims

1. A data noise reduction method, characterized in that, The method includes: The data matrix generated after scanning the target area on the chip surface at various frequency points is obtained; The electromagnetic image to be processed is determined from the electromagnetic image corresponding to the data matrix at each frequency point; The pixels of the electromagnetic image to be processed are divided into a preset number of pixel value ranges; The histogram noise variance estimate of the electromagnetic image to be processed is determined based on the median and frequency corresponding to each pixel value interval; the frequency is determined based on the number of pixels in the electromagnetic image to be processed within the corresponding pixel value interval. Based on the basic estimate of the electromagnetic image to be processed and the histogram noise variance estimate, transform domain collaborative noise reduction is performed on the electromagnetic image to be processed to obtain the target electromagnetic image; the basic estimate is obtained by stacking each reference block and the corresponding similar block in the electromagnetic image to be processed, performing discrete cosine transform, then performing hard thresholding on the coefficients obtained after transformation, and performing inverse transform, and putting the block returned by inverse transform back to the position of the previously searched block, and performing a weighted average of all obtained overlapping block estimates by block to determine the target electromagnetic image; The step of determining the histogram noise variance estimate of the electromagnetic image to be processed based on the median and frequency corresponding to each of the pixel value intervals includes: The sum of the first product of the median and the corresponding frequency for each pixel value interval is used as the mean of the histogram. Determine the squared value of the difference between the median and the mean of the histogram for each of the pixel value intervals; Determine the second product of the square value corresponding to each pixel value interval and the corresponding frequency; The summation result obtained by summing the results of each second product is used as the histogram noise variance estimate.

2. The method according to claim 1, characterized in that, Determining the electromagnetic image to be processed from the electromagnetic images corresponding to the data matrices at each of the aforementioned frequency points includes: Based on the mean value of each row element in the data matrix at each frequency point, determine the target row variance of each row element in the data matrix at each frequency point. Based on the mean of each column element in the data matrix at each frequency point, determine the target column variance of each column element in the data matrix at each frequency point. The target variance is determined based on the target row variance and the target column variance corresponding to the data matrix at each frequency point; The data matrix corresponding to the target variance that is greater than or equal to the preset variance threshold is taken as the target data matrix, and the electromagnetic image corresponding to the target data matrix is ​​taken as the electromagnetic image to be processed.

3. The method according to claim 1, characterized in that, The step of performing transform-domain collaborative noise reduction on the electromagnetic image to be processed based on the basic estimated value of the electromagnetic image to be processed and the histogram noise variance estimate to obtain the target electromagnetic image includes: Identify the similar blocks corresponding to each block in the electromagnetic image to be processed; Based on each block in the electromagnetic image to be processed and its corresponding similar blocks, determine the basic estimated value of the electromagnetic image to be processed; Based on the basic estimate and the histogram noise variance estimate, transform domain collaborative noise reduction is performed on the electromagnetic image to be processed to obtain the target electromagnetic image.

4. A data noise reduction device, characterized in that, The device includes: The acquisition module is used to acquire the data matrix generated after scanning the target area on the chip surface at various frequency points; The first determining module is used to determine the electromagnetic image to be processed from the electromagnetic images corresponding to the data matrix at each frequency point; The segmentation module is used to divide each pixel of the electromagnetic image to be processed into a preset number of pixel value ranges; The second determining module is used to determine the histogram noise variance estimate of the electromagnetic image to be processed based on the median and frequency corresponding to each of the pixel value intervals; the frequency is determined based on the number of pixels of the electromagnetic image to be processed in the corresponding pixel value interval. The processing module is used to perform transform domain collaborative noise reduction processing on the electromagnetic image to be processed based on the basic estimate value and the histogram noise variance estimate to obtain the target electromagnetic image. The basic estimate value is obtained by stacking each reference block and the corresponding similar block in the electromagnetic image to be processed, performing discrete cosine transform, then performing hard thresholding on the coefficients obtained after transformation, and performing inverse transform. The block returned by inverse transform is placed back into the position of the previously searched block, and the weighted average of all obtained overlapping block estimates is performed to determine the target electromagnetic image. The step of determining the histogram noise variance estimate of the electromagnetic image to be processed based on the median and frequency corresponding to each of the pixel value intervals includes: The sum of the first product of the median and the corresponding frequency for each pixel value interval is used as the mean of the histogram. Determine the squared value of the difference between the median and the mean of the histogram for each of the pixel value intervals; Determine the second product of the square value corresponding to each pixel value interval and the corresponding frequency; The summation result obtained by summing the results of each second product is used as the histogram noise variance estimate.

5. A computer device comprising a memory and a processor, wherein the memory stores a computer program, characterized in that, When the processor executes the computer program, it implements the steps of the method according to any one of claims 1 to 3.

6. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by a processor, it implements the steps of the method according to any one of claims 1 to 3.

7. A computer program product, comprising a computer program, characterized in that, When the computer program is executed by a processor, it implements the steps of the method according to any one of claims 1 to 3.

Images