CN109727295B - Electromagnetic image extraction method, electromagnetic image extraction device, computer equipment and storage medium - Google Patents
Electromagnetic image extraction method, electromagnetic image extraction device, computer equipment and storage medium Download PDFInfo
- Publication number
- CN109727295B CN109727295B CN201811542900.3A CN201811542900A CN109727295B CN 109727295 B CN109727295 B CN 109727295B CN 201811542900 A CN201811542900 A CN 201811542900A CN 109727295 B CN109727295 B CN 109727295B
- Authority
- CN
- China
- Prior art keywords
- frequency point
- emission value
- electromagnetic
- value matrix
- matrix
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
Images
Landscapes
- Image Processing (AREA)
- Investigating, Analyzing Materials By Fluorescence Or Luminescence (AREA)
Abstract
The application relates to an electromagnetic image extraction method, an electromagnetic image extraction device, computer equipment and a storage medium. The method comprises the following steps: acquiring electromagnetic field original image data obtained by scanning the surface of the object to be measured through the near-field surface, wherein the electromagnetic field original image data are emission values of the surface of the object to be measured scanned through the near-field surface under each frequency point; respectively storing the emission value of each frequency point in the electromagnetic field original image data into an emission value matrix corresponding to each frequency point according to the frequency point; respectively solving the standard deviation of the emission value matrix of each frequency point; screening the emission value matrix of each frequency point according to a preset standard deviation threshold value, and extracting the emission value matrix of the object to be detected at a specific frequency point; and generating an electromagnetic image of the object to be detected according to the emission value matrix of the object to be detected at the specific frequency point. By adopting the method, the original image data obtained by scanning can be automatically extracted without manually checking a specific frequency point for extraction.
Description
Technical Field
The present application relates to the field of near-field surface scanning technologies, and in particular, to an electromagnetic image extraction method and apparatus, a computer device, and a storage medium.
Background
Near field surface scanning is a common technique currently used for performing electromagnetic compatibility testing on plate products and device-level products, and the testing method is to obtain electromagnetic field distribution in a near field region of the surface of an object to be tested through an electromagnetic field probe, and can be used for electromagnetic interference positioning, evaluation and the like.
Generally, near field surface scanning tests collect electromagnetic emission signals of an object to be tested at different frequency points through a frequency spectrograph at a fixed point in a certain area range until all scanning points go through, but the collected signals include background noise besides intentional electromagnetic emission signals of the object to be tested, so that the intentional electromagnetic emission signals of the object to be tested need to be extracted.
Because the electromagnetic emission of the object to be detected only occurs on the specific frequency points, and other frequency points are all background noise signals, the conventional method finds the specific frequency points in a manual inspection mode and extracts data under the specific frequency points.
Disclosure of Invention
In view of the above, it is necessary to provide an electromagnetic image extraction method, an electromagnetic image extraction apparatus, a computer device, and a storage medium, which address the above technical problems.
An electromagnetic image extraction method, the method comprising:
acquiring electromagnetic field original image data obtained by scanning the surface of the object to be measured through the near-field surface, wherein the electromagnetic field original image data are emission values of the surface of the object to be measured scanned through the near-field surface under each frequency point;
respectively storing the emission value of each frequency point in the electromagnetic field original image data into an emission value matrix corresponding to each frequency point according to the frequency point;
respectively solving the standard deviation of the emission value matrix of each frequency point;
screening the emission value matrix of each frequency point according to a preset standard deviation threshold value, and extracting the emission value matrix of the object to be detected at a specific frequency point;
and generating an electromagnetic image of the object to be detected according to the emission value matrix of the object to be detected at the specific frequency point.
In one embodiment, the screening the emission value matrix of each frequency point according to the preset standard deviation threshold value comprises:
and screening out an emission value matrix with the standard deviation larger than a standard deviation threshold value from the emission value matrix of each frequency point.
In one embodiment, the electromagnetic image extracting method further includes:
respectively carrying out dimension reduction treatment on the emission value matrix of the object to be detected at each specific frequency point;
and clustering the emission value matrix of the object to be measured at each specific frequency point after dimension reduction to obtain a clustering result of the electromagnetic image of the object to be measured.
In one embodiment, the performing the dimension reduction processing on the emission value matrix of the object to be measured at each specific frequency point respectively comprises:
recombining the emission value matrixes of the object to be detected at all the specific frequency points into an input matrix;
performing morphological segmentation processing on the input matrix to obtain a binarization matrix;
and performing dimension reduction processing on the binarization matrix.
In one embodiment, the dimension reduction processing on the emission value matrix of the specific frequency point of the object to be measured comprises the following steps:
and (3) performing dimensionality reduction on the emission value matrix of the object to be detected at each specific frequency point by adopting a principal component analysis method.
In one embodiment, the clustering the emission value matrix of the specific frequency point of the object to be measured after the dimension reduction comprises:
and clustering the emission value matrix of the object to be measured at each specific frequency point by adopting a K-Means clustering method.
In one embodiment, the electromagnetic image extracting method further includes:
and establishing an electromagnetic test model by using the clustering result of the electromagnetic image of the object to be tested.
An electromagnetic image extracting apparatus, characterized by comprising:
the device comprises an original image data acquisition module, a data acquisition module and a data acquisition module, wherein the original image data acquisition module is used for acquiring electromagnetic field original image data obtained by scanning the surface of an object to be measured through a near-field surface, and the electromagnetic field original image data are emission values of the surface of the object to be measured scanned through the near-field surface under each frequency point;
the data dividing module is used for storing the emission value of each frequency point in the electromagnetic field original image data as an emission value matrix of the corresponding frequency point according to the frequency point;
the standard deviation processing module is used for respectively solving the standard deviation of the emission value matrix of each frequency point;
the screening module is used for screening the emission value matrix of each frequency point according to a preset standard deviation threshold value and extracting the emission value matrix of the object to be tested at a specific frequency point;
and the image generation module is used for generating an electromagnetic image of the object to be detected according to the emission value matrix of the object to be detected at the specific frequency point.
A computer device comprising a memory and a processor, the memory storing a computer program, the processor implementing the following steps when executing the computer program:
acquiring electromagnetic field original image data obtained by scanning the surface of the object to be measured through the near-field surface, wherein the electromagnetic field original image data are emission values of the surface of the object to be measured scanned through the near-field surface under each frequency point;
respectively storing the emission value of each frequency point in the electromagnetic field original image data into an emission value matrix corresponding to each frequency point according to the frequency point;
respectively solving the standard deviation of the emission value matrix of each frequency point;
screening the emission value matrix of each frequency point according to a preset standard deviation threshold value, and extracting the emission value matrix of the object to be detected at a specific frequency point;
and generating an electromagnetic image of the object to be detected according to the emission value matrix of the object to be detected at the specific frequency point.
A computer-readable storage medium, on which a computer program is stored which, when executed by a processor, carries out the steps of:
acquiring electromagnetic field original image data obtained by scanning the surface of the object to be measured through the near-field surface, wherein the electromagnetic field original image data are emission values of the surface of the object to be measured scanned through the near-field surface under each frequency point;
respectively storing the emission value of each frequency point in the electromagnetic field original image data into an emission value matrix corresponding to each frequency point according to the frequency point;
respectively solving the standard deviation of the emission value matrix of each frequency point;
screening the emission value matrix of each frequency point according to a preset standard deviation threshold value, and extracting the emission value matrix of the object to be detected at a specific frequency point;
and generating an electromagnetic image of the object to be detected according to the emission value matrix of the object to be detected at the specific frequency point.
According to the electromagnetic image extraction method, the electromagnetic image extraction device, the computer equipment and the storage medium, the electromagnetic field original image data obtained by scanning the surface of the object to be detected on the near field surface are recombined according to the frequency, the standard deviation of the emission value matrix of each frequency point in a repeated data set is respectively solved, as the intentional electromagnetic emission signal is an electromagnetic image with a regular pattern, the emission value distribution of pixels is scattered, the standard deviation is large, the background noise is an electromagnetic image without a pattern, the emission value distribution of pixels is concentrated, and the standard deviation is small, the emission value matrix of each frequency point is screened according to the preset standard deviation threshold, the background noise can be removed, the intentional electromagnetic emission signal image data of the object to be detected at the specific frequency point is extracted, the original image data obtained by scanning can be automatically extracted, and the specific frequency point does not need to be manually checked for extraction.
Drawings
FIG. 1 is a schematic flow chart diagram of an electromagnetic image extraction method in one embodiment;
FIG. 2 is a schematic flow chart of an electromagnetic image extraction method according to another embodiment;
FIG. 3 is a flowchart illustrating a dimension reduction processing method of the electromagnetic image extraction method according to an embodiment;
FIG. 4 is a block diagram showing the structure of an electromagnetic image extracting apparatus according to an embodiment;
FIG. 5 is a diagram illustrating an internal structure of a computer device according to an embodiment.
Detailed Description
In order to make the objects, technical solutions and advantages of the present application more apparent, the present application is described in further detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative of the present application and are not intended to limit the present application.
In one embodiment, as shown in FIG. 1, an electromagnetic image extraction method is provided, which comprises steps 110-150:
and 110, acquiring electromagnetic field original image data obtained by scanning the surface of the object to be measured through the near-field surface, wherein the electromagnetic field original image data are emission values of the surface of the object to be measured through the near-field surface scanning at each frequency point.
The near-field surface scanning is to perform electromagnetic field near-field scanning on a PCB or an electronic component to obtain the electromagnetic field distribution condition of the object to be measured.
Specifically, a near-field surface scanning instrument is used for scanning the object to be detected, and electromagnetic signal emission values of all frequency points of the surface of the object to be detected in a certain frequency point range, which are obtained by scanning of the near-field surface scanning instrument, are obtained.
And 120, storing the emission value of each frequency point in the electromagnetic field original image data into an emission value matrix corresponding to each frequency point according to the frequency point.
Because the emission values directly obtained by scanning are values of a single pixel point, in order to facilitate processing and finally obtain electromagnetic images distributed according to frequency points, all the obtained emission values are divided according to the frequency points, and all the emission values corresponding to each frequency point are stored as an emission value matrix of the frequency point.
Storing the emission value of each single frequency point as a matrix Xm,
Wherein, XmRefers to a matrix of emission values, x, scanned at the mth frequency pointijIs the electromagnetic signal emission value of the ith row and the jth column pixel point.
And step 120, respectively solving the standard deviation of the emission value matrix of each frequency point.
The object to be measured is in the intentional electromagnetic emission signal distribution under specific frequency point more in disorder, and the discrete degree is high, therefore the standard deviation is great, and the emission signal distribution of background noise is comparatively concentrated, and the discrete degree is lower, therefore the standard deviation is less, can filter through predetermined standard deviation threshold value, make the data set when extracting the processing screening diminish, only need filter the standard deviation of every emission value matrix, need not to filter the data of every emission value matrix itself, shorten the time of extracting.
Specifically, let the standard deviation of the emission value of the mth frequency point be σm，
Wherein, mumThe average value of the emission values of all the pixel points of the mth frequency point is obtained, and n is the total number of the pixel points, that is, n is i × j.
And 140, screening the emission value matrix of each frequency point according to a preset standard deviation threshold value, and extracting the emission value matrix of the object to be detected at the specific frequency point.
And comparing the standard deviation of the emission values under all frequencies with a preset standard deviation threshold, wherein the emission value matrix with the standard deviation smaller than the threshold belongs to background noise, and the emission value matrix with the standard deviation reaching the threshold is the emission value matrix of the specific frequency point of the object to be detected, namely the data corresponding to the intentional electromagnetic emission signal of the object to be detected.
And 150, generating an electromagnetic image of the object to be detected according to the emission value matrix of the object to be detected at the specific frequency point.
And generating an electromagnetic image of the object to be detected according to the extracted emission value matrix, wherein the electromagnetic image of the object to be detected can visually observe the electromagnetic field distribution condition of the object to be detected.
In the electromagnetic image extraction method, electromagnetic field original image data obtained by scanning the surface of an object to be detected on a near field surface are respectively stored as emission value matrixes according to frequency points to respectively calculate the standard deviation of the emission value matrixes of the frequency points, and because an intentional electromagnetic emission signal is an electromagnetic image with a regular pattern, the emission value distribution of pixels is scattered, the standard deviation is large, background noise is an electromagnetic image without a pattern, the emission value distribution of pixels is concentrated, and the standard deviation is small, the emission value matrixes of the frequency points are screened according to a preset standard deviation threshold value, background noise can be removed, the intentional electromagnetic emission signal image data of the object to be detected at a specific frequency point is extracted, the original image data obtained by scanning can be automatically extracted, and the specific frequency point does not need to be manually checked for extraction.
In one embodiment, the step of screening the emission value matrix of each frequency point according to a preset standard deviation threshold includes:
and screening out an emission value matrix with the standard deviation larger than a standard deviation threshold value from the emission value matrix of each frequency point.
And comparing the standard deviation of the emission value matrix of each frequency point with a preset standard deviation threshold, wherein the standard deviation is larger than the emission value matrix of the standard deviation threshold, the corresponding frequency point is a specific frequency point, and the emission value is an intentional electromagnetic emission signal value of the object to be detected under the specific frequency point, so that the part of the emission value matrix is screened out, and the emission values in the emission value matrix of which the standard deviation is smaller than the standard deviation threshold are all electromagnetic emission signals of background noise.
In one embodiment, as shown in FIG. 2, the electromagnetic image extraction method further comprises steps 160-170:
and 160, respectively carrying out dimension reduction processing on the emission value matrix of the object to be detected at each specific frequency point.
The dimension reduction is carried out on the emission value matrix of each specific frequency point of the extracted object to be measured, the data amount required to be processed by clustering is reduced, the processing efficiency is improved, the interference can be reduced, the error is reduced, and the clustering accuracy is improved.
And 170, clustering the emission value matrix of the object to be measured at each specific frequency point after dimension reduction to obtain a clustering result of the electromagnetic image of the object to be measured.
At present, no processing method exists for analyzing whether the extracted electromagnetic images are related or not, if modeling test is required to be carried out by using the electromagnetic images, a relatively accurate model can be established only by using data of all the electromagnetic images for modeling, the workload is very large, and the time investment for carrying out the test is high.
The clustering can analyze the similarity between the electromagnetic images of the object to be tested, the electromagnetic images with higher similarity can be divided into one class through the clustering, and in the subsequent test or evaluation, only one or two electromagnetic images are extracted from one class to be analyzed, so that the data representing the characteristics of the electromagnetic images can be obtained to be evaluated or modeled.
In one embodiment, the clustering method may be K-Means clustering, mean shift clustering, density-based clustering (DBSCAN), maximum Expectation (EM) clustering using Gaussian Mixture Model (GMM), agglomerative hierarchical clustering, or Graph Community Detection (Graph Community Detection), and the like, and those skilled in the art may select the clustering method as needed.
In one embodiment, as shown in fig. 3, the performing dimension reduction on the emission value matrix of the object to be measured at each specific frequency point includes steps 161 to 163:
step 161, recombining the emission value matrix of the object to be measured at each specific frequency point into an input matrix;
specifically, the matrix of the emission value of the object to be measured at the mth frequency point is set as YmIs a reaction of YmSpread out as a vector AmAnd then:
wherein, yijAnd extracting the emission value of the electromagnetic image of the object to be detected on the ith row and the jth column pixel point.
And (3) transposing and recombining the vector obtained by expanding the extracted emission value matrix under each specific frequency point into an input matrix E, wherein the input matrix contains the emission value matrix under m frequency points, and then:
step 162, performing morphological segmentation processing on the matrix to be input to obtain a binary matrix;
and 163, performing dimension reduction processing on the binary matrix.
The morphological segmentation is to highlight the outline of the electromagnetic image of the object to be measured, specifically, the pattern part in the electromagnetic image is converted into full black through a morphological segmentation algorithm, the remaining part is full white, the black is represented by 1, the white is represented by 0, so that a binarization matrix can be obtained, the Euclidean distance of points of the same pattern in a feature space can be smaller by the image correspondingly generated by the binarization matrix, and the Euclidean distance of points of different patterns in the feature space is larger, so that the clustering process is more convenient.
In one embodiment, the dimension reduction processing on the emission value matrix or the binarization matrix of the specific frequency point of the object to be measured can adopt methods such as a Principal Component Analysis (PCA), a t-distribution neighborhood embedding algorithm (t-SNE), a multidimensional scaling Method (MDS) or an SVD matrix decomposition method, and the like, and can be selected by a person skilled in the art according to needs.
Principal Component Analysis (PCA), among others, is a commonly used dimension-reducing statistical method that uses an orthogonal transformation to convert a set of possibly correlated variable input data into a set of linearly uncorrelated variable values called principal components. Since PCA simply transforms the input data, it can be used for both classification and regression problems. The kernel method kernelized PCA can be used for the non-linear case, but due to its good mathematical properties, the speed of finding the transformed feature space, and the ability to interconvert between the re-original and transformed features, PCA can satisfy most cases already at the time of dimensionality reduction or feature extraction. The skilled person can select as desired
Given the original space, PCA will find a linear mapping to a lower dimensional space. Since it is desirable to have the projections of all samples as separated as possible, it is desirable to maximize the variance of the projection points. The variance retained by PCA is the largest and the error of the final reconstruction (from transformed back to original) is the smallest.
The specific treatment method comprises the following steps:
to measure the degree of dispersion between each dimension of the feature vector (i.e., each pixel within the electromagnetic image), the covariance between the two dimensions A, B is calculated as follows:
let cpqCov (p, q) represents the p-th row and q-th column of the covariance matrix C, which is defined as follows:
according to the definition formula of the feature vector:
wherein
The n different vectors λ are the eigenvalues corresponding to the n eigenvectors, and the eigenvector matrix V formed by the n eigenvectors is as follows:
V＝(v1 v2 …vn)
for a single feature vector viI is an integer satisfying 1. ltoreq. i.ltoreq.n. Thus, the above definition can be converted into:
C×vi＝λvi
solving n such eigenvector equations can obtain vectors containing n different eigenvalues
And a matrix V composed of n eigenvectors. Since the eigenvectors are unitized (i.e. length is 1) and are orthogonal to each other, these eigenvectors can be used as the basis of the dimension data to be reduced (extracted matrix of emission values of specific frequency points of the object to be measured, i.e. electromagnetic image of the object to be measured). To obtain V and of C
And then, reducing the dimension of the extracted emission value matrix of the specific frequency point of the object to be detected. Feature vector V in Vi(i.e. each column) in accordance withCharacteristic value lambdaiArranging in descending order, and then taking the first k main components as the extracted emission value matrix set E of the specific frequency point of the object to be measured1The substrate matrix V' that can be projected is as follows:
V′＝(v′1 v′2… v′k)
wherein v'1、v′2……v′kThe first k eigenvectors are arranged in descending order of eigenvalues and are also the principal components of the original matrix C. In general, the principal component in V 'contains the main information of the original covariance matrix C, so V' can be used to refer to C.
The final step of PCA is to multiply the averaged input matrix E with the projected substrate matrix V', i.e. project the averaged input matrix onto a substrate that is the principal component and is selected, i.e. the projection matrix is as follows:
P＝V′T×Emean-adjusted T
i.e. it can be deduced that:
wherein f ism kThe index m of (a) indicates that the element is located in the mth column of the projection matrix P, and represents that the column is data of the mth frequency in the emission value matrix set; f. ofm kThe superscript k indicates that the element is located in the kth row of the projection matrix P, which represents that the row is a principal projection, i.e., a data dimension left after dimensionality reduction.
Through the processing, the data is reduced from the original dimensionality to the k dimensionality, too much information cannot be lost, and the minimum error is ensured during reconstruction.
After dimension reduction, each column of the projection matrix P is clustered, and a classification result with relatively high accuracy can be obtained.
In one embodiment, the emission value matrix of the specific frequency points of the object to be measured after the dimension reduction is clustered by adopting a K-Means clustering method.
In another embodiment, the binary matrix after dimension reduction is clustered by adopting a K-Means clustering method.
The K-Means clustering method has the characteristic of simple and convenient calculation and can effectively improve the calculation speed.
The specific clustering steps are as follows:
step 171, taking each row of the projection matrix P as a frequency point in the feature space, and initializing k centroids (i.e. central points) at random.
And 172, calculating the distance from each frequency point to each centroid, and classifying each frequency point to obtain the centroid with the shortest distance.
And 173, calculating the mean values of all the frequency points classified into the centroid for each centroid respectively, and updating the obtained mean values into new centroids.
Repeating steps 172-173 until the intra-class variance converges to a minimum value.
The intra-class variance is:
wherein p isjIs a frequency point, ciFor the class represented by the corresponding centroid i, i e (1,2 … k), μiIs the center of mass.
In one embodiment, the electromagnetic image extracting method further includes:
and step 180, establishing an electromagnetic test model by using the clustering result of the electromagnetic image of the object to be tested.
And if z types exist in the clustering result, extracting a part of electromagnetic image data from each type for modeling the type of products, wherein the electromagnetic test model is to be established, and the work load is reduced.
It should be understood that although the various steps in the flow charts of fig. 1-3 are shown in order as indicated by the arrows, the steps are not necessarily performed in order as indicated by the arrows. The steps are not performed in the exact order shown and described, and may be performed in other orders, unless explicitly stated otherwise. Moreover, at least some of the steps in fig. 1-3 may include multiple sub-steps or multiple stages that are not necessarily performed at the same time, but may be performed at different times, and the order of performance of the sub-steps or stages is not necessarily sequential, but may be performed in turn or alternating with other steps or at least some of the sub-steps or stages of other steps.
In one embodiment, as shown in fig. 4, there is provided an electromagnetic image extracting apparatus including: an original image data obtaining module 210, a data dividing module 220, a standard deviation processing module 230, a screening module 240, and an image generating module 250, wherein:
an original image data obtaining module 210, configured to obtain electromagnetic field original image data obtained by scanning the surface of the object to be measured through the near-field surface, where the electromagnetic field original image data is an emission value of the surface of the object to be measured through near-field surface scanning at each frequency point;
the data dividing module 220 is configured to store the emission value of each frequency point in the electromagnetic field original image data as an emission value matrix of a corresponding frequency point according to the frequency point;
a standard deviation processing module 230, configured to separately calculate a standard deviation for the emission value matrix of each frequency point;
the screening module 240 is configured to screen the emission value matrix of each frequency point according to a preset standard deviation threshold, and extract the emission value matrix of the object to be tested at a specific frequency point;
and the image generating module 250 is configured to generate an electromagnetic image of the object to be measured according to the emission value matrix of the object to be measured at the specific frequency point.
In one embodiment, the electromagnetic image extracting apparatus further includes: the dimensionality reduction processing module and the clustering module are used for solving the problems that:
the dimension reduction module is used for respectively carrying out dimension reduction processing on the emission value matrix of the object to be measured at each specific frequency point;
and the clustering module is used for clustering the emission value matrix of the object to be measured at each specific frequency point after dimension reduction to obtain a clustering result of the electromagnetic image of the object to be measured.
In one embodiment, the dimension reduction processing module comprises: restructuring module, binarization module and dimension reduction operation module, wherein:
the recombination module is used for recombining the emission value matrix of the object to be detected at each specific frequency point into an input matrix;
the binarization module is used for performing morphological segmentation processing on an emission value matrix of a specific frequency point of an object to be detected to obtain a binarization matrix;
and the dimension reduction operation module is used for carrying out dimension reduction processing on the binarization matrix.
In one embodiment, the electromagnetic image extracting apparatus further includes:
and the modeling module is used for establishing an electromagnetic test model by utilizing the clustering result of the electromagnetic image of the object to be tested.
For specific limitations of the electromagnetic image extraction device, reference may be made to the above limitations of the electromagnetic image extraction method, which are not described herein again. The respective modules in the electromagnetic image extracting apparatus described above may be entirely or partially implemented by software, hardware, and a combination thereof. The modules can be embedded in a hardware form or independent from a processor in the computer device, and can also be stored in a memory in the computer device in a software form, so that the processor can call and execute operations corresponding to the modules.
In one embodiment, a computer device is provided, which may be a server, the internal structure of which may be as shown in fig. 5. The computer device includes a processor, a memory, a network interface, and a database connected by a system bus. Wherein the processor of the computer device is configured to provide computing and control capabilities. The memory of the computer device comprises a nonvolatile storage medium and an internal memory. The non-volatile storage medium stores an operating system, a computer program, and a database. The internal memory provides an environment for the operation of an operating system and computer programs in the non-volatile storage medium. The network interface of the computer device is used for communicating with an external terminal through a network connection. The computer program is executed by a processor to implement an electromagnetic image extraction method.
Those skilled in the art will appreciate that the architecture shown in fig. 5 is merely a block diagram of some of the structures associated with the disclosed aspects and is not intended to limit the computing devices to which the disclosed aspects apply, as particular computing devices may include more or less components than those shown, or may combine certain components, or have a different arrangement of components.
In one embodiment, a computer device is provided, comprising a memory and a processor, the memory having a computer program stored therein, the processor implementing the following steps when executing the computer program:
acquiring electromagnetic field original image data obtained by scanning the surface of the object to be measured through the near-field surface, wherein the electromagnetic field original image data are emission values of the surface of the object to be measured scanned through the near-field surface under each frequency point;
respectively storing the emission value of each frequency point in the electromagnetic field original image data into an emission value matrix corresponding to each frequency point according to the frequency point;
respectively solving the standard deviation of the emission value matrix of each frequency point;
screening the emission value matrix of each frequency point according to a preset standard deviation threshold value, and extracting the emission value matrix of the object to be detected at a specific frequency point;
and generating an electromagnetic image of the object to be detected according to the emission value matrix of the object to be detected at the specific frequency point.
In one embodiment, the processor, when executing the computer program, further performs the steps of:
respectively carrying out dimension reduction treatment on the emission value matrix of the object to be detected at each specific frequency point;
and clustering the emission value matrix of the object to be measured at each specific frequency point after dimension reduction to obtain a clustering result of the electromagnetic image of the object to be measured.
In one embodiment, the processor, when executing the computer program, further performs the steps of:
recombining the emission value matrix of the object to be detected at each specific frequency point into an input matrix;
performing morphological segmentation processing on the input matrix to obtain a binarization matrix;
and performing dimension reduction processing on the binarization matrix.
In one embodiment, the processor, when executing the computer program, further performs the steps of:
and establishing an electromagnetic test model by using the clustering result of the electromagnetic image of the object to be tested.
In one embodiment, a computer-readable storage medium is provided, having a computer program stored thereon, which when executed by a processor, performs the steps of:
acquiring electromagnetic field original image data obtained by scanning the surface of the object to be measured through the near-field surface, wherein the electromagnetic field original image data are emission values of the surface of the object to be measured scanned through the near-field surface under each frequency point;
respectively storing the emission value of each frequency point in the electromagnetic field original image data into an emission value matrix corresponding to each frequency point according to the frequency point;
respectively solving the standard deviation of the emission value matrix of each frequency point;
screening the emission value matrix of each frequency point according to a preset standard deviation threshold value, and extracting the emission value matrix of the object to be detected at a specific frequency point;
and generating an electromagnetic image of the object to be detected according to the emission value matrix of the object to be detected at the specific frequency point.
In one embodiment, the computer program when executed by the processor further performs the steps of:
respectively carrying out dimension reduction treatment on the emission value matrix of the object to be detected at each specific frequency point;
and clustering the emission value matrix of the object to be measured at each specific frequency point after dimension reduction to obtain a clustering result of the electromagnetic image of the object to be measured.
In one embodiment, the computer program when executed by the processor further performs the steps of:
recombining the emission value matrix of the object to be detected at each specific frequency point into an input matrix;
performing morphological segmentation processing on the input matrix to obtain a binarization matrix;
and performing dimension reduction processing on the binarization matrix.
In one embodiment, the computer program when executed by the processor further performs the steps of:
and establishing an electromagnetic test model by using the clustering result of the electromagnetic image of the object to be tested.
It will be understood by those skilled in the art that all or part of the processes of the methods of the embodiments described above can be implemented by hardware instructions of a computer program, which can be stored in a non-volatile computer-readable storage medium, and when executed, can include the processes of the embodiments of the methods described above. Any reference to memory, storage, database, or other medium used in the embodiments provided herein may include non-volatile and/or volatile memory, among others. 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 a variety of forms such as Static RAM (SRAM), Dynamic RAM (DRAM), Synchronous DRAM (SDRAM), Double Data Rate SDRAM (DDRSDRAM), Enhanced SDRAM (ESDRAM), Synchronous Link DRAM (SLDRAM), Rambus Direct RAM (RDRAM), direct bus dynamic RAM (DRDRAM), and memory bus dynamic RAM (RDRAM).
The technical features of the above embodiments can be arbitrarily combined, and for the sake of brevity, all possible combinations of the technical features in the above embodiments are not described, but should be considered as the scope of the present specification as long as there is no contradiction between the combinations of the technical features.
The above-mentioned embodiments only express several embodiments of the present application, and the description thereof is more specific and detailed, but not construed as limiting the scope of the invention. It should be noted that, for a person skilled in the art, several variations and modifications can be made without departing from the concept of the present application, which falls within the scope of protection of the present application. Therefore, the protection scope of the present patent shall be subject to the appended claims.
Claims (9)
1. An electromagnetic image extraction method, characterized in that the method comprises:
acquiring electromagnetic field original image data obtained by scanning the surface of an object to be detected through a near-field surface, wherein the electromagnetic field original image data are electromagnetic signal emission values of the surface of the object to be detected under each frequency point through the near-field surface scanning;
respectively storing the electromagnetic signal emission value of each frequency point in the electromagnetic field original image data into an emission value matrix corresponding to each frequency point according to the frequency point;
respectively solving the standard deviation of the emission value matrix of each frequency point;
screening the emission value matrix of each frequency point according to a preset standard deviation threshold value, and screening out the emission value matrix with the standard deviation reaching the standard deviation threshold value as the emission value matrix of the object to be detected at the specific frequency point;
and generating an electromagnetic image of the object to be detected according to the emission value matrix of the object to be detected at the specific frequency point.
2. The electromagnetic image extraction method of claim 1, further comprising:
respectively carrying out dimension reduction processing on the emission value matrix of the object to be detected at each specific frequency point;
and clustering the emission value matrix of the object to be measured at each specific frequency point after dimension reduction to obtain a clustering result of the electromagnetic image of the object to be measured.
3. The electromagnetic image extraction method of claim 2, wherein the performing the dimension reduction processing on the emission value matrix of the object to be measured at each specific frequency point respectively comprises:
recombining the emission value matrix of the object to be detected at each specific frequency point into an input matrix;
performing morphological segmentation processing on the input matrix to obtain a binarization matrix;
and carrying out dimension reduction processing on the binarization matrix.
4. The electromagnetic image extraction method of claim 2, wherein the dimension reduction processing of the emission value matrix of the specific frequency point of the object to be measured comprises:
and performing dimensionality reduction on the emission value matrix of the object to be detected at each specific frequency point by adopting a principal component analysis method.
5. The method for extracting an electromagnetic image according to claim 2, wherein clustering the emission value matrix of the specific frequency points of the object to be measured after the dimension reduction comprises:
and clustering the emission value matrix of the object to be measured at each specific frequency point by adopting a K-Means clustering method.
6. The electromagnetic image extraction method of claim 5, further comprising:
and establishing an electromagnetic test model by using the clustering result of the electromagnetic image of the object to be tested.
7. An electromagnetic image extraction apparatus, characterized in that the apparatus comprises:
the device comprises an original image data acquisition module, a data acquisition module and a data acquisition module, wherein the original image data acquisition module is used for acquiring electromagnetic field original image data obtained by scanning the surface of an object to be measured through a near-field surface, and the electromagnetic field original image data are emission values of the surface of the object to be measured scanned through the near-field surface under each frequency point;
the data matrixing module is used for storing the emission value of each frequency point in the electromagnetic field original image data as an emission value matrix of a corresponding frequency point according to the frequency point;
the standard deviation processing module is used for respectively solving the standard deviation of the emission value matrix of each frequency point;
the screening module is used for screening the emission value matrix of each frequency point according to a preset standard deviation threshold value, screening the emission value matrix of which the standard deviation reaches the standard deviation threshold value as the emission value matrix of the object to be tested at a specific frequency point;
and the image generation module is used for generating an electromagnetic image of the object to be detected according to the emission value matrix of the object to be detected at the specific frequency point.
8. A computer device comprising a memory and a processor, the memory storing a computer program, wherein the processor implements the steps of the method of any one of claims 1 to 6 when executing the computer program.
9. A computer-readable storage medium, on which a computer program is stored, which, when being executed by a processor, carries out the steps of the method of any one of claims 1 to 6.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201811542900.3A CN109727295B (en) | 2018-12-17 | 2018-12-17 | Electromagnetic image extraction method, electromagnetic image extraction device, computer equipment and storage medium |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201811542900.3A CN109727295B (en) | 2018-12-17 | 2018-12-17 | Electromagnetic image extraction method, electromagnetic image extraction device, computer equipment and storage medium |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN109727295A CN109727295A (en) | 2019-05-07 |
| CN109727295B true CN109727295B (en) | 2021-06-29 |
Family
ID=66297650
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201811542900.3A Active CN109727295B (en) | 2018-12-17 | 2018-12-17 | Electromagnetic image extraction method, electromagnetic image extraction device, computer equipment and storage medium |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN109727295B (en) |
Families Citing this family (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN112561821B (en) * | 2020-12-17 | 2024-05-17 | 中国电子产品可靠性与环境试验研究所（（工业和信息化部电子第五研究所）（中国赛宝实验室）） | Near field scanning-based chip surface electromagnetic data noise reduction method |
| CN112749668A (en) * | 2021-01-18 | 2021-05-04 | 上海明略人工智能(集团)有限公司 | Target image clustering method and device, electronic equipment and computer readable medium |
| CN113781332B (en) * | 2021-08-24 | 2023-09-29 | 中国电子产品可靠性与环境试验研究所（（工业和信息化部电子第五研究所）（中国赛宝实验室）） | Noise reduction method and device for electromagnetic data, computer equipment and storage medium |
Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104584326A (en) * | 2012-05-09 | 2015-04-29 | 杜克大学 | Metamaterial devices and methods of using the same |
| CN105026970A (en) * | 2013-01-08 | 2015-11-04 | 麻省理工学院 | Optical phased arrays |
Family Cites Families (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US7301199B2 (en) * | 2000-08-22 | 2007-11-27 | President And Fellows Of Harvard College | Nanoscale wires and related devices |
-
2018
- 2018-12-17 CN CN201811542900.3A patent/CN109727295B/en active Active
Patent Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104584326A (en) * | 2012-05-09 | 2015-04-29 | 杜克大学 | Metamaterial devices and methods of using the same |
| CN105026970A (en) * | 2013-01-08 | 2015-11-04 | 麻省理工学院 | Optical phased arrays |
Non-Patent Citations (2)
| Title |
|---|
| 一种新的自适应宽带扫频方法及其在分层微带结构全波分析中的应用;宋喆，吴仕飞，李卫东，张彦;《中国科学》;20170330;第47卷(第7期);全文 * |
| 基于近场声全息声像图纹理特征的机械故障诊断方法;鲁文波，蒋伟康，潘思伟，向上;《振动工程学报》;20130830;第26卷(第4期);全文 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN109727295A (en) | 2019-05-07 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN109727295B (en) | Electromagnetic image extraction method, electromagnetic image extraction device, computer equipment and storage medium | |
| Frery et al. | Analytic expressions for stochastic distances between relaxed complex Wishart distributions | |
| Gracia et al. | A methodology to compare dimensionality reduction algorithms in terms of loss of quality | |
| Mathiassen et al. | Texture similarity measure using Kullback-Leibler divergence between gamma distributions | |
| US10902088B2 (en) | Quantitative analysis of signal related measurements for trending and pattern recognition | |
| Rabinovich et al. | Model order selection and cue combination for image segmentation | |
| Zhang et al. | Low-rank-sparse subspace representation for robust regression | |
| CN110717806B (en) | Product information pushing method, device, equipment and storage medium | |
| Han et al. | l0-norm based structural sparse least square regression for feature selection | |
| CN112070008A (en) | Hyperspectral image feature identification method, device and equipment and storage medium | |
| Nowakowska et al. | Tractable measure of component overlap for gaussian mixture models | |
| Tirelli et al. | Unsupervised machine learning approaches to the q-state Potts model | |
| Cosmo et al. | The average mixing kernel signature | |
| Fursov et al. | Thematic classification with support subspaces in hyperspectral images | |
| CN111223570A (en) | Pathological data analysis method, device, equipment and storage medium | |
| Seth et al. | Model criticism in latent space | |
| Sáez et al. | Comparative study of probability distribution distances to define a metric for the stability of multi-source biomedical research data | |
| Srivatsa et al. | Application of least square denoising to improve admm based hyperspectral image classification | |
| CN114422450A (en) | Network flow analysis method and device based on multi-source network flow data | |
| Zhang et al. | A computational perspective on projection pursuit in high dimensions: feasible or infeasible feature extraction | |
| Vescan et al. | Cross-Project Defect Prediction using Supervised and Unsupervised Learning: a Replication Study | |
| Negri et al. | Stochastic approaches of minimum distance method for region based classification | |
| Wang et al. | Robust and Sparse Kernel PCA and Its Outlier Map | |
| Juneja et al. | Graph-based strategy for establishing morphology similarity | |
| Bressan et al. | Independent modes of variation in point distribution models |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| CB02 | Change of applicant information |
Address after: 511300 No.78, west of Zhucun Avenue, Zhucun street, Zengcheng District, Guangzhou City, Guangdong Province Applicant after: CHINA ELECTRONIC PRODUCT RELIABILITY AND ENVIRONMENTAL TESTING RESEARCH INSTITUTE ((THE FIFTH ELECTRONIC RESEARCH INSTITUTE OF MIIT)(CEPREI LABORATORY)) Address before: 510610 No. 110 Zhuang Road, Tianhe District, Guangdong, Guangzhou, Dongguan Applicant before: CHINA ELECTRONIC PRODUCT RELIABILITY AND ENVIRONMENTAL TESTING RESEARCH INSTITUTE ((THE FIFTH ELECTRONIC RESEARCH INSTITUTE OF MIIT)(CEPREI LABORATORY)) |
|
| CB02 | Change of applicant information | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |