CN113297879A - Acquisition method of measurement model group, measurement method and related equipment - Google Patents
Acquisition method of measurement model group, measurement method and related equipment Download PDFInfo
- Publication number
- CN113297879A CN113297879A CN202010111118.7A CN202010111118A CN113297879A CN 113297879 A CN113297879 A CN 113297879A CN 202010111118 A CN202010111118 A CN 202010111118A CN 113297879 A CN113297879 A CN 113297879A
- Authority
- CN
- China
- Prior art keywords
- sample
- detected
- vector
- vectors
- structural information
- 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.)
- Pending
Links
- 238000005259 measurement Methods 0.000 title claims abstract description 120
- 238000000034 method Methods 0.000 title claims abstract description 74
- 238000000691 measurement method Methods 0.000 title abstract description 18
- 239000013598 vector Substances 0.000 claims abstract description 507
- 238000003062 neural network model Methods 0.000 claims description 25
- 238000012360 testing method Methods 0.000 claims description 25
- 238000001228 spectrum Methods 0.000 claims description 20
- 238000012545 processing Methods 0.000 claims description 19
- 238000012549 training Methods 0.000 claims description 11
- 230000003595 spectral effect Effects 0.000 claims description 9
- 230000006835 compression Effects 0.000 claims description 8
- 238000007906 compression Methods 0.000 claims description 8
- 238000004590 computer program Methods 0.000 claims description 5
- 238000001514 detection method Methods 0.000 claims description 5
- 239000010410 layer Substances 0.000 description 45
- 238000010586 diagram Methods 0.000 description 11
- 230000006870 function Effects 0.000 description 10
- 230000003287 optical effect Effects 0.000 description 8
- 239000012780 transparent material Substances 0.000 description 7
- 238000004364 calculation method Methods 0.000 description 5
- 238000004891 communication Methods 0.000 description 5
- 238000012876 topography Methods 0.000 description 5
- 239000000463 material Substances 0.000 description 4
- 239000011159 matrix material Substances 0.000 description 4
- 238000004422 calculation algorithm Methods 0.000 description 3
- 238000000354 decomposition reaction Methods 0.000 description 3
- 210000002569 neuron Anatomy 0.000 description 3
- 230000009286 beneficial effect Effects 0.000 description 2
- 230000005540 biological transmission Effects 0.000 description 2
- 238000006243 chemical reaction Methods 0.000 description 2
- 230000000694 effects Effects 0.000 description 2
- 238000005516 engineering process Methods 0.000 description 2
- 238000010606 normalization Methods 0.000 description 2
- 238000005457 optimization Methods 0.000 description 2
- 239000002356 single layer Substances 0.000 description 2
- 230000004913 activation Effects 0.000 description 1
- 230000002457 bidirectional effect Effects 0.000 description 1
- 230000005684 electric field Effects 0.000 description 1
- 230000008030 elimination Effects 0.000 description 1
- 238000003379 elimination reaction Methods 0.000 description 1
- 238000005305 interferometry Methods 0.000 description 1
- 230000001131 transforming effect Effects 0.000 description 1
- 238000013519 translation Methods 0.000 description 1
- 238000002834 transmittance Methods 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F2218/00—Aspects of pattern recognition specially adapted for signal processing
- G06F2218/12—Classification; Matching
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/04—Architecture, e.g. interconnection topology
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/08—Learning methods
- G06N3/084—Backpropagation, e.g. using gradient descent
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T17/00—Three dimensional [3D] modelling, e.g. data description of 3D objects
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F2218/00—Aspects of pattern recognition specially adapted for signal processing
- G06F2218/08—Feature extraction
Abstract
The embodiment of the invention discloses an acquisition method of a measurement model group, a measurement method and related equipment. In addition, after the classification of the signal vector to be measured is determined, the measurement model corresponding to the classification is selected to obtain the structural information of the signal vector to be measured, so that the accuracy and the acquisition speed of the structural information can be improved.
Description
Technical Field
The present invention relates to the field of measurement technologies, and in particular, to a method and an apparatus for acquiring a measurement model set, a measurement device, a terminal device, and a computer storage medium.
Background
The white light interference measurement method is a high-precision nanoscale surface profile measurement method, adopts a low-coherence light interference test technology, uses a scanning module of a high-precision method to scan, image and acquire interference signals on a surface to be measured, realizes non-contact three-dimensional reconstruction, and has wide application in integrated circuit detection, micro-opto-electro-mechanical systems, micro-nano optical systems and other surface topography measurements.
In the prior art, the three-dimensional reconstruction process of the surface topography measurement method needs huge calculation amount, so that the corresponding surface profile information cannot be rapidly acquired, and needs to be improved.
Disclosure of Invention
The embodiment of the invention provides an acquisition method of a measurement model group, a measurement method and related equipment, which can improve the measurement speed of structural information.
In a first aspect, an embodiment of the present invention provides a method for obtaining a measurement model group, including:
acquiring a plurality of groups of sample signal vector groups, wherein each group of sample signal vector group comprises a plurality of sample signal vectors, the sample signal vectors are used for representing the structural information of the sample, and the structural information similarity of the sample signal vectors in each group meets a preset condition;
and modeling the structural information of the sample according to the plurality of sample signal vectors in each group of sample signal vector groups respectively to obtain a measurement model of the structural information.
Optionally, the step of obtaining a plurality of sets of sample signal vectors includes:
obtaining a plurality of sample signal vectors;
obtaining a sample sparse dictionary, wherein the sample signal vector can be linearly represented by atoms in the sample sparse dictionary;
carrying out sparse representation on the sample signal vector according to the sample sparse dictionary to obtain a sample sparse coefficient vector of the sample signal vector;
and classifying the sample signal vectors according to the sample sparse coefficient vectors, and dividing the sample signal vectors of which the sample sparse coefficient vectors meet the same condition into a group to form a plurality of groups of sample signal vector groups.
Optionally, classifying the sample signal vector according to the sparse coefficient vector comprises:
obtaining a similar representation of the sample sparse coefficient vector;
the similar representative same sample signal vectors are classified into one class.
Optionally, the step of obtaining a plurality of sample signal vectors comprises:
providing a plurality of sample light intensity signal vectors;
respectively carrying out Fourier transform on the sample light intensity signal vectors to obtain sample spectrum signal vectors;
and respectively carrying out real number and dimensionality compression on the sample spectrum signal vector to obtain the sample signal vector.
Optionally, the step of modeling processing comprises:
acquiring structure information of each sample signal vector group;
establishing a neural network model;
and training the neural network model through the sample signal vectors and the corresponding structural information to obtain the neural network model parameters of each sample signal vector, wherein the neural network model and the neural network model parameters of each sample signal vector form a measurement model of each structural information.
Optionally, the steps of obtaining a plurality of sets of sample signal vectors and obtaining structural information of each set of sample signal vectors include:
providing a plurality of structural information of a sample;
establishing a relation model between the structural information of the sample and the signal vector of the sample;
obtaining a sample signal vector corresponding to each piece of structural information according to the plurality of pieces of structural information and the relationship model;
classifying the sample signal vectors to obtain a plurality of groups of sample signal vector groups;
alternatively, the first and second electrodes may be,
providing a plurality of sets of sample signal vectors, the sets of sample signal vectors comprising one or more sample signal vectors;
establishing a relation model between the structural information of the sample and the signal vector of the sample;
and acquiring the structural information corresponding to each sample signal vector according to the plurality of sample signal vectors and the relation model.
Optionally, the sample signal vector comprises a reflected light signal vector, and the relationship model comprises a correspondence between a signal formed by light reflected by the sample and structural information of the sample.
In a second aspect, an embodiment of the present invention provides a measurement method, including:
providing a target to be detected, wherein the target to be detected comprises a plurality of regions to be detected;
respectively obtaining a signal vector to be detected of each region to be detected, wherein the signal vector to be detected is used for representing the structural information of the region to be detected;
classifying the signal vectors to be detected, wherein the signal vectors to be detected with structural information similarity meeting a preset condition are classified into the same vector set to be detected;
obtaining a measurement model group, wherein the measurement model group comprises measurement models which are in one-to-one correspondence with the vector sets to be measured;
and acquiring structural information corresponding to the signal vector to be measured in the vector set to be measured through the measurement model corresponding to the vector set to be measured.
Optionally, the structure information includes: height information of the surface of the target to be measured, film thickness information, refractive index or dielectric constant.
Optionally, the structure information includes: height information or film thickness information of the target surface to be measured, the method further comprising:
and performing three-dimensional reconstruction on the structural information to acquire the three-dimensional morphology information of the target to be detected.
Optionally, the classifying the signal vectors to be measured, wherein the classifying the signal vectors to be measured whose structural information similarity satisfies a preset condition into the same set of vectors to be measured includes:
acquiring a test sparse dictionary, wherein the signal vector to be tested can be linearly represented by atoms in the test sparse dictionary;
performing sparse representation on the signal vector to be tested according to the test sparse dictionary to obtain a sparse coefficient vector to be tested of the signal vector to be tested;
and classifying the signal vectors to be detected according to the sparse coefficient vectors to be detected, and dividing the signal vectors to be detected, of which the sparse coefficient vectors to be detected meet the same conditions, into a set to form a plurality of sets of the vectors to be detected.
Optionally, obtaining the measurement model set by the method of the first aspect;
the measurement models which are in one-to-one correspondence with the vector sets to be measured are that the similarity of the structural information of the vectors in the sample signal vector group of the measurement model and the structural information of the signal vectors to be measured in the corresponding vector sets to be measured meets the preset condition.
Optionally, the sample sparse dictionary is the same as the test sparse dictionary.
Optionally, before performing classification processing on the signal vector to be detected, the method further includes:
judging whether the number of the signal vectors to be detected is greater than or equal to a preset number threshold value or not;
when the number of the signal vectors to be detected is greater than or equal to the number threshold, executing the classification processing until structural information corresponding to the signal vectors to be detected in the vector set to be detected is obtained through a measurement model corresponding to the vector set to be detected;
and when the number of the signal vectors to be detected is smaller than a preset number threshold, performing nonlinear fitting on the signal vectors to be detected to obtain structural information corresponding to the signal vectors to be detected.
Optionally, the signal vector to be measured includes an interference light signal vector, a reflected light signal vector, or a scattered light signal vector.
Optionally, the step of respectively obtaining a signal vector to be detected of each region to be detected, where the signal vector to be detected is used to represent structural information of the region to be detected, includes:
respectively obtaining a light intensity signal vector to be detected of each region to be detected;
respectively carrying out Fourier transform on the light intensity signal vectors to be detected to obtain spectral signal vectors to be detected;
and respectively carrying out real number and dimensionality compression on the spectral signal vector to be detected to obtain the signal vector to be detected.
In a third aspect, an embodiment of the present invention provides an apparatus for obtaining a measurement model group, including:
the device comprises a first acquisition module, a second acquisition module and a third acquisition module, wherein the first acquisition module is used for acquiring a plurality of groups of sample signal vector groups, each group of sample signal vector groups comprises a plurality of sample signal vectors, the sample signal vectors are used for representing the structural information of the sample, and the structural information similarity of the sample signal vectors in each group meets a preset condition;
and the processing module is used for performing modeling processing on the structural information of the sample according to the plurality of sample signal vectors in each group of sample signal vector groups respectively to obtain a measurement model of the structural information.
In a fourth aspect, an embodiment of the present invention provides a measurement apparatus, including:
the system comprises an input module, a detection module and a control module, wherein the input module is used for providing a target to be detected, and the target to be detected comprises a plurality of regions to be detected;
the second acquisition module is used for respectively acquiring a signal vector to be detected of each region to be detected, and the signal vector to be detected is used for representing the structural information of the region to be detected;
the classification module is used for classifying the signal vectors to be detected, wherein the signal vectors to be detected with the structural information similarity meeting the preset conditions are classified into the same vector set to be detected;
a third obtaining module, configured to obtain a measurement model group, where the measurement model group includes measurement models that correspond to the set of vectors to be measured one by one;
and the fourth obtaining module is used for obtaining the structural information corresponding to the signal vector to be measured in the vector set to be measured through the measurement model corresponding to the vector set to be measured.
In a fifth aspect, an embodiment of the present invention provides a terminal device, including: a processor and a memory;
the processor is connected to the memory, wherein the memory is configured to store program code, and the processor is configured to call the program code to perform the method according to the first aspect or the second aspect.
In a sixth aspect, embodiments of the present invention provide a computer storage medium storing a computer program comprising program instructions that, when executed by a processor, perform a method according to the first or second aspect.
According to the embodiment of the invention, a plurality of groups of sample signal vector groups are obtained, each group of sample signal vector group comprises a plurality of sample signal vectors, the sample signal vectors are used for representing the structure information of a sample, and the similarity of the structure information of the sample signal vectors in each group meets a preset condition; and modeling the structural information of the sample according to the plurality of sample signal vectors in each group of sample signal vectors to obtain a measurement model of the structural information. Therefore, after the sample signal vectors are classified, different measurement models of the structural information are generated by using different sample signal vector groups, and the calculation accuracy and speed of the measurement models on the structural information can be improved.
Furthermore, the structure information of the object to be detected is obtained through the neural network model, and the detection speed can be improved.
In addition, the embodiment of the invention also provides a target to be detected, wherein the target to be detected comprises a plurality of areas to be detected; respectively obtaining a signal vector to be detected of each region to be detected, wherein the signal vector to be detected is used for representing the structural information of the region to be detected; classifying the signal vectors to be detected, wherein the signal vectors to be detected with structural information similarity meeting a preset condition are classified into the same vector set to be detected; obtaining a measurement model group, wherein the measurement model group comprises measurement models which are in one-to-one correspondence with the set of vectors to be measured; acquiring structural information corresponding to a signal vector to be measured in a vector set to be measured through a measurement model corresponding to the vector set to be measured; therefore, after the classification of the signal vector to be measured is determined, the measurement model corresponding to the classification is selected to obtain the structural information of the signal vector to be measured, and the accuracy and the obtaining speed of the structural information can be improved.
Drawings
In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings needed to be used in the description of the embodiments are briefly introduced below, and it is obvious that the drawings in the following description are some embodiments of the present invention, and it is obvious for those skilled in the art to obtain other drawings based on these drawings without creative efforts.
Fig. 1 is a schematic flowchart of a method for acquiring a measurement model set according to an embodiment of the present invention;
fig. 2 is a schematic flowchart of a method for acquiring a measurement model set according to an embodiment of the present invention;
fig. 3 is a schematic flowchart of a method for acquiring a measurement model set according to an embodiment of the present invention;
fig. 4 is a schematic diagram of a film structure of an acquisition method of a measurement model set according to an embodiment of the present invention;
FIG. 5 is a schematic diagram of a neural network model of a method for obtaining a measurement model set according to an embodiment of the present invention;
fig. 6 is a schematic flow chart of a measurement method according to an embodiment of the present invention;
fig. 7 is a schematic structural diagram of an apparatus for obtaining a measurement model set according to an embodiment of the present invention;
FIG. 8 is a schematic structural diagram of a measuring apparatus according to an embodiment of the present invention;
fig. 9 is a schematic structural diagram of a terminal device according to an embodiment of the present invention.
Detailed Description
The technical solution in the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings in the embodiments of the present invention.
It should be understood that the terms "comprises" and "comprising," and any variations thereof, in the description and claims of this application and the drawings, are intended to cover non-exclusive inclusions. For example, a process, method, system, article, or apparatus that comprises a list of steps or elements is not limited to only those steps or elements listed, but may alternatively include other steps or elements not listed, or inherent to such process, method, article, or apparatus.
Reference in the specification to "an embodiment" means that a particular feature, structure, or characteristic described in connection with the embodiment can be included in at least one embodiment of the invention. The appearances of the phrase in various places in the specification are not necessarily all referring to the same embodiment, nor are separate or alternative embodiments mutually exclusive of other embodiments. It is explicitly and implicitly understood by the person skilled in the art that the described embodiments of the invention can be combined with other embodiments.
In the prior art, in the article three-dimensional profile measurement method based on the white light interferometry, the profile measurement speed is low due to the fact that the data volume required to be processed in the three-dimensional reconstruction process of the profile is too large, and the profile measurement of an article cannot be completed quickly. Therefore, the present application provides a method for obtaining a measurement model set and a measurement method, which can quickly obtain structural information of an object to be measured.
Referring to fig. 1, fig. 1 is a schematic flow chart illustrating a method for obtaining a measurement model set according to an embodiment of the present invention; the method for acquiring the measurement model group comprises the following steps:
101. acquiring a plurality of groups of sample signal vector groups, wherein each group of sample signal vector group comprises a plurality of sample signal vectors, the sample signal vectors are used for representing the structure information of a sample, and the similarity of the structure information of the sample signal vectors in each group meets a preset condition;
specifically, the sample signal vectors include interference light signal vectors, reflection light signal vectors, or scattering light signal vectors, the preset conditions may be set as needed, and the preset conditions are used to classify the sample signal vectors to obtain a plurality of sets of sample signal vectors. The classification can be performed according to the size of the similarity of the structural information between the sample signal vectors, and the sample signal vectors with the same or similar structural information are grouped into one group.
102. And modeling the structural information of the sample according to the plurality of sample signal vectors in each group of sample signal vector groups respectively to obtain a measurement model of the structural information.
Specifically, a measurement model of the structural information can be obtained by modeling the structural information of the sample with a plurality of sample signal vectors in a set of sample signal vectors. Based on the sets of sample signal vectors, a plurality of measurement models of structural information can be obtained. Wherein the structural information comprises height information of the sample surface, film thickness information, refractive index or dielectric constant.
In this embodiment, after the sample signal vectors are classified, different measurement models of the structural information are generated by using different sample signal vector groups, so that the calculation accuracy and speed of the measurement models on the structural information can be improved.
Further, in a possible embodiment, referring to fig. 2, fig. 2 is a schematic flowchart of a method for obtaining a measurement model set according to an embodiment of the present invention, where step 101 includes:
201. obtaining a plurality of sample signal vectors;
specifically, a plurality of sample signal vectors of the sample are acquired, and the manner of acquiring the sample signal vectors is not limited.
202. Acquiring a sample sparse dictionary, wherein a sample signal vector can be linearly represented by atoms in the sample sparse dictionary;
specifically, sample sparse dictionary training is performed according to the plurality of sample signal vectors of step 201. The sparse dictionary D to be learned comprises K atoms, and the dimension of each atom is N which is the same as the sample signal vectorc. Learning from sparse dictionaries is finding optimal wordsTypically D, making the sparse coefficient X as sparse as possible, the sparsity being T0I.e. by
Wherein Y is a sample signal vector, and the solution of the sparse dictionary can be obtained by a K-SVD method. The concrete implementation steps are as follows:
s1, initializing sparse dictionary D according to sample signal vector(0)And its second order norm is normalized. The number of iterations J equals 1. S2, sparse representation of the training set is carried out by using an arbitrary tracking algorithm based on the sparse dictionary D, a sparse coefficient X is obtained, and DX is approximate representation of the training set.
S3, for the iteration sparse dictionary D(J-1)Ith atom d iniAnd sparse coefficient ith row xiThe iteration is carried out with the error E defined by the second order norm distance, i.e.
From EiBy a matrix of logical values omegaiDefine to obtain non-zero terms
To pair
Performing singular value decomposition to obtain
Using the first column of U as the ith atom d of the sparse dictionaryi,Using the product of the first column of V and Δ (1,1) on the ith row x of the sparse coefficientR iAnd (6) updating.
And S4, changing J to J +1, and repeating the iteration process until the iteration termination condition is met, namely the training of the sample sparse dictionary is completed.
The iteration termination condition includes: the iteration times reach the maximum iteration times, or the error E between the reconstructed signal and the sample signal vector is smaller than a threshold value.
203. Carrying out sparse representation on the sample signal vector according to the sample sparse dictionary to obtain a sample sparse coefficient vector of the sample signal vector;
specifically, sparse representation is performed on the sample signal vectors according to the sample sparse dictionary obtained in step 202, so as to obtain a sample sparse coefficient vector of each sample signal vector.
204. And classifying the sample signal vectors according to the sample sparse coefficient vectors, and dividing the sample signal vectors of which the sample sparse coefficient vectors meet the same condition into a group to form a plurality of groups of sample signal vector groups.
Specifically, the same condition means that similar representations of sample sparse coefficient vectors are the same, wherein the similar representations of the sample sparse coefficient vectors are obtained first, and the sample signal vectors with the same similar representations are classified into one class. In other embodiments, the similarity of the sample sparse coefficient vectors may be such that the error of the similar representation of the sample signal vectors is less than a set threshold.
Further, in a possible embodiment, referring to fig. 3, fig. 3 is a schematic flowchart of a method for obtaining a measurement model set according to an embodiment of the present invention, where step 102 includes:
301. acquiring structure information of each sample signal vector group;
in this embodiment, the steps of obtaining a plurality of sets of sample signal vectors and obtaining structural information of each sample signal vector set include:
providing a plurality of structural information of a sample;
establishing a relation model between the structural information of the sample and the signal vector of the sample;
obtaining a sample signal vector corresponding to each piece of structural information according to the plurality of pieces of structural information and the relationship model;
classifying the sample signal vectors to obtain a plurality of groups of sample signal vector groups;
alternatively, the first and second electrodes may be,
providing a plurality of sets of sample signal vectors, the sets of sample signal vectors comprising one or more sample signal vectors;
establishing a relation model between the structural information of the sample and the signal vector of the sample;
and acquiring the structural information corresponding to each sample signal vector according to the plurality of sample signal vectors and the relation model.
In other embodiments, a sample signal vector is obtained by detecting samples of known structural information. Obtaining a plurality of sample signal vectors comprises: providing a plurality of samples of which structural information is known; carrying out optical detection on the samples to obtain sample light intensity signal vectors of all the samples; and obtaining the sample signal vector according to the sample light intensity signal vector. Specifically, in this embodiment, a pixel coordinate (x, y) on the area array detector is set to scan the sample along the vertical direction z to obtain a plurality of discrete sample light intensity signal vectors I. Respectively carrying out discrete Fourier transform on the sample light intensity signal vector I to obtain a sample spectrum signal vector S; by transforming the matrix TclusterAnd respectively carrying out real number and dimensionality compression on the sample spectrum signal vector S to obtain a sample signal vector.
Specifically, the sample signal vector includes a reflected light signal vector, and the reflected light signal vector is subjected to a non-linear fitting process to obtain structural information corresponding to the reflected light signal vector. Wherein, the reflected light signal vector comprises a reflected light intensity signal vector or a reflected spectrum signal vector. The relational model includes a correspondence between a signal formed by light reflected by the sample and structural information of the sample.
In this embodiment, taking the example that the sample includes a multilayer film structure and the reflected light signal vector is a reflected spectrum signal vector, establishing the relationship model between the structure information of the sample and the sample signal vector includes:
solving the reflection spectrum signal vector can regard the sample with the L-layer film layer structure as a multi-stage transmission module by using a transmission matrix method, and the structural information is the film layer structure parameter at the moment. Referring to fig. 4, fig. 4 is a diagram illustrating a method for obtaining a measurement model set according to an embodiment of the present inventionThe structural diagram of the film layer of (1); wherein, the characteristic matrix M of the l-th layerlCan be expressed as a number of times,
in the formula tl,rlRespectively the first layer medium transmittance and reflectance,
introducing a phase difference into the film layer, and then the relation model is as follows:
E0 +,E0 -,EL+1 +,EL+1 -is a bidirectional electric field vector on two sides of the film layer; reflection coefficient r ═ s21/s11The reflection coefficient of light at each frequency is r (k).
Obtaining structural information corresponding to each sample signal vector according to the plurality of sample signal vectors and the relation model, wherein the structural information comprises:
carrying out nonlinear fitting according to the relation model to obtain film layer structure parameter calculation;
searching the peak of the light intensity signal vector I of the sample, searching and recording the phase of the surface signal
Translating and standardizing the sample spectrum signal vector S to obtain a normalized reflection coefficient vector rtrain。
Solving non-linear optimization problems, i.e.
And obtaining a film layer structure parameter vector x. Where r (x) is the reflection spectrum vector under the film structure parameter vector x.
302. Establishing a neural network model;
in this embodiment, the test model group is obtained by establishing a neural network model; in other embodiments, different test models may be established for different sets of sample signal vectors by linear or nonlinear fitting.
Specifically, fig. 5 is a schematic diagram of a neural network model of an obtaining method of a measurement model set according to an embodiment of the present invention, and taking a film structure as an example, in fig. 5, an input layer is a reflection spectrum signal vector, and an output layer is a film structure parameter vector.
The input to the pth neuron at layer o is:
the output of the pth neuron at layer o is:
wherein f is a neuron activation function, W is a weight coefficient, and b is a bias term.
303. Training the neural network model through the sample signal vectors and the corresponding structural information to obtain the neural network model parameters of each sample signal vector, and forming the measurement model of each structural information by the neural network model and the neural network model parameters of each sample signal vector.
Taking a set of sample signal vectors as an example, and taking the film layer structure parameters as an example, a set of film layer structure parameters corresponding to the sample signal vector set can be obtained by using step 301, that is, one sample signal vector (in this case, a reflection spectrum signal vector) corresponds to one film layer structure parameter. Taking a film structure parameter vector x as a center, randomly or regularly selecting a group of discrete vectors (namely sample film structure parameter vectors) in the maximum variation range of film structure parameters to be divided into a training set and a testing set according to proportion, and calculating each weight coefficient W on the training set by applying a back propagation algorithmpq (o)And offset bp (o)And then, verifying on the test set, and storing the final result parameters as the neural network model parameters. A set of neural network model parameters may form a measurement model of the structural information.
Referring to fig. 6, fig. 6 is a schematic flow chart of a measurement method according to an embodiment of the present invention; the measuring method comprises the following steps:
601. providing a target to be detected, wherein the target to be detected comprises a plurality of regions to be detected;
specifically, the object to be measured may be a transparent (such as a lens) or non-transparent (such as a circuit board) article, and the surface profile of the article may be composed of any multiple layers of transparent or non-transparent films (more than one layer), and the thickness of the surface profile is not limited.
602. Respectively obtaining a signal vector to be detected of each region to be detected, wherein the signal vector to be detected is used for representing the structural information of the region to be detected;
specifically, the signal vector to be measured includes one or more combinations of interference light signal vectors, reflected light signal vectors, or scattered light signal vectors. The type of the signal vector to be measured is the same as that of the sample signal vector. And when the sample signal vector is a reflection signal vector, the signal vector to be detected is a reflection light signal vector.
In this embodiment, the step of respectively obtaining a signal vector to be detected of each region to be detected, where the signal vector to be detected is used to represent structural information of the region to be detected, includes:
respectively obtaining light intensity signal vectors to be detected of each region to be detected;
respectively carrying out Fourier transform on the light intensity signal vectors to be detected to obtain spectral signal vectors to be detected;
and respectively carrying out real number and dimensionality compression on the spectral signal vector to be detected to obtain the signal vector to be detected.
603. Classifying the signal vectors to be detected, wherein the signal vectors to be detected with structural information similarity meeting a preset condition are classified into the same vector set to be detected;
specifically, the preset condition may be set as required, that is, the signal vectors to be measured are classified according to the preset condition to obtain a plurality of sets of vectors to be measured. The signal vectors to be tested corresponding to the regions to be tested with the same or similar structures can be classified into a group according to the similarity of the structural information among the signal vectors to be tested.
604. Obtaining a measurement model group, wherein the measurement model group comprises measurement models which are in one-to-one correspondence with the set of vectors to be measured;
specifically, different sets of vectors to be measured correspond to different measurement models, wherein the similarity between the training data of the measurement model corresponding to the set of vectors to be measured and the structural information of the signal vectors to be measured of the set of vectors to be measured meets a preset condition.
605. And acquiring structural information corresponding to the signal vector to be measured in the vector set to be measured through the measurement model corresponding to the vector set to be measured.
Specifically, the structural information includes height information of the surface of the target to be measured, film thickness information, refractive index, or dielectric constant. After the classification of the signal vector to be detected is determined, the measurement model corresponding to the classification is selected to obtain the structural information of the signal vector to be detected, the calculated amount is reduced, and the accuracy and the acquisition speed of the structural information can be effectively improved.
Further, in one possible embodiment, step 603 includes:
6031. acquiring a test sparse dictionary, wherein the signal vector to be tested can be linearly represented by atoms in the test sparse dictionary;
in this embodiment, the sample sparse dictionary is the same as the test sparse dictionary. The test sparse dictionary is obtained by using the method for obtaining the measurement model group, which is not described herein in detail.
6032. And performing sparse representation on the signal vector to be tested according to the test sparse dictionary to obtain a sparse coefficient vector to be tested of the signal vector to be tested.
Before the classification processing, the measurement method further includes: and performing standard normalization processing on all the signal vectors to be measured.
Performing sparse representation on the signal vector to be tested according to the test sparse dictionary comprises:
then carrying out sparse decomposition on the normalized signal vector Y to be detected in a sparse dictionary D, and solving the sparse decomposition
6033. And classifying the signal vectors to be detected according to the sparse coefficient vectors to be detected, and dividing the signal vectors to be detected, of which the sparse coefficient vectors to be detected meet the same conditions, into a set to form a plurality of sets of the vectors to be detected.
Specifically, the same condition means that the similar representations of the sparse coefficient vectors to be measured are the same, wherein the similar representations of the sparse coefficient vectors to be measured are obtained first, and the signal vectors to be measured with the same similar representations are classified into one class. Step 603 is to divide the signal vectors to be measured corresponding to the regions to be measured with the same or similar structures into a set. Specifically, the signal vector to be measured is classified into Class (Y) according to the minimum reconstruction errorM)：
Taking the structure of the region to be measured as the film layer structure as an example, the film layer structure refers to film layer optical parameters such as the number of film layers, the material of the film layers, and the parameters of a single-layer structure (such as thickness and material duty ratio), the same film layer structure refers to the number of film layers, the material of the film layers, and the parameters of the single-layer structure are completely the same, and the close film layer structure refers to that although some film layer optical parameters have slight differences (the specific values of the differences can be set according to actual requirements), some film layer optical parameters are close to the film layer optical parameters of the category to which the film layer structure belongs. Taking the circuit board as an example, in the surface profile of the circuit board, there may be a portion of the area covered with a layer of transparent material, a portion of the area covered with two layers of transparent material, or a portion of the area covered with a non-transparent material, or a portion of the area covered with two layers of transparent material. The signal vectors to be measured can be classified into four types, one type is a set of signal vectors to be measured in which the film layer is a layer of transparent material, one type is a set of signal vectors to be measured in which the film layer is two layers of transparent material, one type is a set of signal vectors to be measured in which the film layer is a non-transparent material, and the other type is a set of signal vectors to be measured in which the film layer is two layers of material which are densely arranged but can be used as an equivalent film layer.
Further, in a possible embodiment, in step 604, the measurement model set is obtained by the above-mentioned method for obtaining the measurement model set; the measurement models which are in one-to-one correspondence with the vector sets to be measured are that the similarity of the structural information of the vectors in the sample signal vector group of the measurement model and the corresponding signal vectors to be measured in the vector sets to be measured meets the preset condition. When the structural information is obtained by using the measurement model, the sample signal vector at the moment is a reflected light signal vector, and the reflected light signal vector comprises a reflected spectrum signal vector or a reflected light intensity signal vector. And normalizing all the spectrum signal vectors to obtain the reflection spectrum signal vector.
Further, in a possible embodiment, before performing the classification processing on the signal vector to be detected, the method further includes:
judging whether the number of the signal vectors to be detected is greater than or equal to a preset number threshold value or not; the preset number threshold value can be set according to needs without limitation;
when the number of the signal vectors to be measured is greater than or equal to the number threshold, executing classification processing until structural information corresponding to the signal vectors to be measured in the vector set to be measured is obtained through a measurement model corresponding to the vector set to be measured; when the number of the signal vectors to be detected is greater than or equal to the number threshold, executing the steps 603-605 to obtain structural information;
and when the number of the signal vectors to be detected is smaller than a preset number threshold, performing nonlinear fitting on the signal vectors to be detected to obtain structural information corresponding to the signal vectors to be detected.
Specifically, for a small amount of data such as single-point data or multi-point data, namely, when the number of the signal vectors to be detected is smaller than a preset number threshold, the nonlinear fitting algorithm is directly applied to the calculation of the signal to be detected. Taking the signal vector to be measured as an interference signal vector as an example, taking the acquisition of film layer structure parameters as an example, firstly, the interference signal is subjected to peak searching, and the estimated phase phi of the surface signal is searched and recordedijThe signal phase is set to zero. Then Fourier transform is carried out on the interference signal to obtain a frequency spectrum, and normalization is carried out to obtain an n-dimensional reflection spectrum vector rm∈Cn(ii) a And finally solving the nonlinear optimization problem to obtain a film structure parameter vector x.
Where r (x) is the reflection spectrum vector under the film structure parameter vector x.
Further, in a possible embodiment, when the structural information includes height information of the target surface to be measured or film thickness information, the measuring method further includes:
and performing three-dimensional reconstruction on the structural information to obtain the three-dimensional shape information of the target to be measured.
Specifically, taking the film structure as an example, the initial three-dimensional morphology information of the target to be detected is obtained according to the film structure parameters of the target to be detected and the pixel coordinates of the target to be detected. The thickness (also called ordinate) of the film layer can be obtained according to the film layer structure parameters, and then the initial three-dimensional shape information of the target to be detected can be obtained by integrating all pixel point data (pixel coordinates (such as abscissa and ordinate)). And then, carrying out space coordinate position conversion processing on the initial three-dimensional shape information to obtain the three-dimensional shape information of the target to be detected. Based on the parameters of the optical system, the initial three-dimensional topography information is converted into the final three-dimensional topography information of the target to be measured, such as a three-dimensional topography map, through spatial coordinate position conversion. Furthermore, the obtained three-dimensional shape information can be subjected to translation elimination, inclination and high-order deformation processing, so that a user can conveniently check the three-dimensional shape information.
Based on the description of the above method for acquiring a measurement model group, an apparatus for acquiring a measurement model group is further disclosed in the embodiments of the present invention, and referring to fig. 7, fig. 7 is a schematic structural diagram of an apparatus for acquiring a measurement model group provided in the embodiments of the present invention, and the apparatus for acquiring a measurement model group includes:
a first obtaining module 701, configured to obtain multiple sets of sample signal vector sets, where each set of sample signal vector set includes multiple sample signal vectors, the sample signal vectors are used to represent structure information of a sample, and a similarity of the structure information of the sample signal vectors in each set meets a preset condition;
the processing module 702 is configured to perform modeling processing on the structural information of the sample according to the plurality of sample signal vectors in each group of sample signal vectors, respectively, to obtain a measurement model of the structural information.
Further, in one possible embodiment, the first obtaining module includes:
a first sub-module for obtaining a plurality of sample signal vectors;
the second submodule is used for acquiring a sample sparse dictionary, and a sample signal vector can be represented by atom linearity in the sample sparse dictionary;
the third sub-module is used for carrying out sparse representation on the sample signal vector according to the sample sparse dictionary to obtain a sample sparse coefficient vector of the sample signal vector;
and the fourth sub-module is used for classifying the sample signal vectors according to the sample sparse coefficient vectors, and dividing the sample signal vectors of which the sample sparse coefficient vectors meet the same condition into a group to form a plurality of groups of sample signal vector groups.
Further, in a possible embodiment, the first sub-module is specifically configured to:
providing a plurality of sample light intensity signal vectors;
respectively carrying out Fourier transform on the light intensity signal vectors of the samples to obtain spectral signal vectors of the samples;
and respectively carrying out real number and dimensionality compression on the sample spectrum signal vector to obtain a sample signal vector.
Further, in a possible embodiment, the fourth submodule is specifically configured to:
obtaining a similar representation of a sample sparse coefficient vector;
similar sample signal vectors representing the same are classified into one class.
Further, in one possible embodiment, the processing module includes:
the fifth submodule is used for acquiring the structural information of each sample signal vector group;
the sixth submodule is used for establishing a neural network model;
and the seventh sub-module is used for training the neural network model through the sample signal vectors and the corresponding structural information to obtain the neural network model parameters of each sample signal vector, and the neural network model parameters of each sample signal vector form the measurement model of each structural information.
In particular, the sample signal vector comprises a reflected light signal vector and the relational model comprises a correspondence between a signal formed by light reflected by the sample and structural information of the sample.
Further, in one possible embodiment, the method of obtaining a plurality of sets of sample signal vectors and obtaining structural information of each set of sample signal vectors includes:
providing a plurality of structural information of a sample;
establishing a relation model between the structural information of the sample and the signal vector of the sample;
obtaining a sample signal vector corresponding to each piece of structural information according to the plurality of pieces of structural information and the relation model;
classifying the sample signal vectors to obtain a plurality of groups of sample signal vector groups;
alternatively, the first and second electrodes may be,
providing a plurality of sets of sample signal vectors, a set of sample signal vectors comprising one or more sample signal vectors;
establishing a relation model between the structural information of the sample and the signal vector of the sample;
and acquiring the structural information corresponding to each sample signal vector according to the plurality of sample signal vectors and the relation model.
It should be noted that, for a specific implementation manner of the function of the obtaining apparatus of the measurement model group, reference may be made to the description of the obtaining method of the measurement model group, and details are not described here. The units or modules in the acquiring apparatus of the measurement model group may be respectively or completely combined into one or several other units or modules to form, or some unit(s) or module(s) may be further split into multiple units or modules with smaller functions to form, which may implement the same operation without affecting implementation of technical effects of embodiments of the present invention. The above units or modules are divided based on logic functions, and in practical applications, the functions of one unit (or module) may also be implemented by a plurality of units (or modules), or the functions of a plurality of units (or modules) may be implemented by one unit (or module).
Based on the description of the above embodiment of the measurement method, the embodiment of the present invention further discloses a measurement apparatus, and referring to fig. 8, fig. 8 is a schematic structural diagram of the measurement apparatus provided in the embodiment of the present invention, where the measurement apparatus includes:
an input module 801, configured to provide a target to be detected, where the target to be detected includes multiple regions to be detected;
a second obtaining module 802, configured to obtain to-be-detected signal vectors of each to-be-detected region, where the to-be-detected signal vectors are used to represent structural information of the to-be-detected region;
the classification module 803 is configured to classify the signal vectors to be detected, where the signal vectors to be detected whose structural information similarity satisfies a preset condition are classified into the same set of vectors to be detected;
a third obtaining module 804, configured to obtain a measurement model group, where the measurement model group includes measurement models that correspond to the sets of vectors to be measured one by one;
a fourth obtaining module 805, configured to obtain, through the measurement model corresponding to the vector set to be measured, structural information corresponding to a signal vector to be measured in the vector set to be measured.
Further, in one possible embodiment, the signal vector to be measured includes an interference light signal vector, a reflected light signal vector, or a scattered light signal vector. The structure information includes: height information of the surface of the target to be measured, film thickness information, refractive index or dielectric constant.
Further, in one possible embodiment, the structure information includes: when the height information or the film thickness information of the target surface to be measured, the measuring device further comprises:
and the fifth acquisition module is used for carrying out three-dimensional reconstruction on the structural information and acquiring the three-dimensional appearance information of the target to be detected.
Further, in one possible embodiment, the classification module includes:
the eighth submodule is used for acquiring a test sparse dictionary, and the signal vector to be tested can be linearly represented by atoms in the test sparse dictionary;
the ninth submodule is used for carrying out sparse representation on the signal vector to be tested according to the test sparse dictionary to obtain a sparse coefficient vector to be tested of the signal vector to be tested;
and the tenth submodule is used for classifying the signal vectors to be measured according to the sparse coefficient vectors to be measured, dividing the signal vectors to be measured, of which the sparse coefficient vectors to be measured meet the same conditions, into a set and forming a plurality of sets of the vectors to be measured.
Specifically, the third obtaining module obtains the measurement model group through the obtaining device of the measurement model group; the measurement models which are in one-to-one correspondence with the vector sets to be measured are that the similarity of the structural information of the vectors in the sample signal vector group of the measurement model and the corresponding signal vectors to be measured in the vector sets to be measured meets the preset condition.
Further, in one possible embodiment, the sample sparse dictionary is the same as the test sparse dictionary.
Further, in one possible embodiment, the second obtaining module includes:
the eleventh submodule is used for respectively acquiring the light intensity signal vector to be detected of each region to be detected;
the twelfth submodule is used for respectively carrying out Fourier transform on the light intensity signal vectors to be detected to obtain spectral signal vectors to be detected;
and the thirteenth submodule is used for respectively carrying out real number and dimension compression on the spectral signal vector to be detected to obtain the signal vector to be detected.
Further, in a possible embodiment, the measuring device further comprises:
the judging module is used for judging whether the number of the signal vectors to be detected is greater than or equal to a preset number threshold value or not;
the first execution module is used for executing classification processing until the structural information corresponding to the signal vector to be measured in the vector set to be measured is obtained through the measurement model corresponding to the vector set to be measured when the number of the signal vectors to be measured is greater than or equal to the number threshold;
and the second execution module is used for performing nonlinear fitting on the signal vectors to be detected when the number of the signal vectors to be detected is smaller than a preset number threshold value, and acquiring the structural information corresponding to the signal vectors to be detected.
It is to be noted that, for a specific functional implementation of the measurement apparatus, reference may be made to the description of the measurement method, and details are not described here. The units or modules in the measuring apparatus may be respectively or completely combined into one or several other units or modules to form the measuring apparatus, or some unit(s) or module(s) thereof may be further split into multiple functionally smaller units or modules to form the measuring apparatus, which may implement the same operation without affecting the implementation of the technical effects of the embodiments of the present invention. The above units or modules are divided based on logic functions, and in practical applications, the functions of one unit (or module) may also be implemented by a plurality of units (or modules), or the functions of a plurality of units (or modules) may be implemented by one unit (or module).
Based on the description of the method embodiment and the device embodiment, the embodiment of the invention also provides the terminal equipment. Fig. 9 is a schematic structural diagram of a terminal device according to an embodiment of the present invention. As shown in fig. 9, the above-mentioned obtaining apparatus or measuring apparatus of the measurement model group may be applied to the terminal device 900, and the terminal device 900 may include: the processor 901, the network interface 904 and the memory 905, the terminal device 900 may further include: a user interface 903, and at least one communication bus 902. Wherein a communication bus 902 is used to enable connective communication between these components. The user interface 903 may include a Display (Display) and a Keyboard (Keyboard), and the optional user interface 903 may also include a standard wired interface and a standard wireless interface. The network interface 904 may optionally include a standard wired interface, a wireless interface (e.g., WI-FI interface). The memory 905 may be a high-speed RAM memory or a non-volatile memory (e.g., at least one disk memory). The memory 905 may optionally be at least one memory device located remotely from the processor 901. As shown in fig. 9, the memory 905, which is a kind of computer storage medium, may include therein an operating system, a network communication module, a user interface module, and a device control application program.
In the terminal apparatus 900 shown in fig. 9, the network interface 904 may provide a network communication function; and the user interface 903 is primarily an interface for providing input to a user; and the processor 901 may be configured to invoke a device control application stored in the memory 905 to implement the above-mentioned acquisition method of the measurement model set or the steps in the measurement method.
It should be understood that the terminal device 900 described in the embodiment of the present invention may perform the aforementioned method for acquiring a measurement model set or the measurement method, and may also perform the aforementioned apparatus for acquiring a measurement model set or the description of the measurement apparatus, which is not described herein again. In addition, the beneficial effects of the same method are not described in detail.
Further, here, it is to be noted that: an embodiment of the present invention further provides a computer storage medium, where the aforementioned apparatus for acquiring a measurement model group or a computer program executed by a measurement apparatus is stored in the computer storage medium, and the computer program includes program instructions, and when the processor executes the program instructions, the processor can execute the aforementioned description of the method for acquiring a measurement model group or the measurement method, and therefore, details will not be repeated here. In addition, the beneficial effects of the same method are not described in detail. For technical details not disclosed in the embodiments of the computer storage medium to which the present invention relates, reference is made to the description of the method embodiments of the present invention.
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 a computer program, which can be stored in a computer-readable storage medium, and when executed, can include the processes of the embodiments of the methods described above. The storage medium may be a magnetic disk, an optical disk, a Read-Only Memory (ROM), a Random Access Memory (RAM), or the like.
The above disclosure is only for the purpose of illustrating the preferred embodiments of the present invention, and it is therefore to be understood that the invention is not limited by the scope of the appended claims.
Claims (20)
1. A method for obtaining a measurement model group is characterized by comprising the following steps:
acquiring a plurality of groups of sample signal vector groups, wherein each group of sample signal vector group comprises a plurality of sample signal vectors, the sample signal vectors are used for representing the structural information of the sample, and the structural information similarity of the sample signal vectors in each group meets a preset condition;
and modeling the structural information of the sample according to the plurality of sample signal vectors in each group of sample signal vector groups respectively to obtain a measurement model of the structural information.
2. The method of claim 1, wherein the step of obtaining a plurality of sets of sample signal vectors comprises:
obtaining a plurality of sample signal vectors;
obtaining a sample sparse dictionary, wherein the sample signal vector can be linearly represented by atoms in the sample sparse dictionary;
carrying out sparse representation on the sample signal vector according to the sample sparse dictionary to obtain a sample sparse coefficient vector of the sample signal vector;
and classifying the sample signal vectors according to the sample sparse coefficient vectors, and dividing the sample signal vectors of which the sample sparse coefficient vectors meet the same condition into a group to form a plurality of groups of sample signal vector groups.
3. The method of claim 2, wherein classifying the sample signal vector according to the sample sparse coefficient vector comprises:
obtaining a similar representation of the sample sparse coefficient vector;
the similar representative same sample signal vectors are classified into one class.
4. The method of claim 2, wherein the step of obtaining a plurality of sample signal vectors comprises:
providing a plurality of sample light intensity signal vectors;
respectively carrying out Fourier transform on the sample light intensity signal vectors to obtain sample spectrum signal vectors;
and respectively carrying out real number and dimensionality compression on the sample spectrum signal vector to obtain the sample signal vector.
5. The method according to any one of claims 1 to 4, wherein the step of modeling comprises:
acquiring structure information of each sample signal vector group;
establishing a neural network model;
and training the neural network model through the sample signal vectors and the corresponding structural information to obtain the neural network model parameters of each sample signal vector, wherein the neural network model and the neural network model parameters of each sample signal vector form a measurement model of each structural information.
6. The method of claim 5, wherein the steps of obtaining a plurality of sets of sample signal vectors and obtaining structural information of each set of sample signal vectors comprise:
providing a plurality of structural information of a sample;
establishing a relation model between the structural information of the sample and the signal vector of the sample;
obtaining a sample signal vector corresponding to each piece of structural information according to the plurality of pieces of structural information and the relationship model;
classifying the sample signal vectors to obtain a plurality of groups of sample signal vector groups;
alternatively, the first and second electrodes may be,
providing a plurality of sets of sample signal vectors, the sets of sample signal vectors comprising one or more sample signal vectors;
establishing a relation model between the structural information of the sample and the signal vector of the sample;
and acquiring the structural information corresponding to each sample signal vector according to the plurality of sample signal vectors and the relation model.
7. The method of claim 6, wherein the sample signal vector comprises a reflected light signal vector, and wherein the relationship model comprises a correspondence between a signal formed by light reflected by the sample and structural information of the sample.
8. A method of measurement, comprising:
providing a target to be detected, wherein the target to be detected comprises a plurality of regions to be detected;
respectively obtaining a signal vector to be detected of each region to be detected, wherein the signal vector to be detected is used for representing the structural information of the region to be detected;
classifying the signal vectors to be detected, wherein the signal vectors to be detected with structural information similarity meeting a preset condition are classified into the same vector set to be detected;
obtaining a measurement model group, wherein the measurement model group comprises measurement models which are in one-to-one correspondence with the vector sets to be measured;
and acquiring structural information corresponding to the signal vector to be measured in the vector set to be measured through the measurement model corresponding to the vector set to be measured.
9. The method of claim 8, wherein the configuration information comprises: height information of the surface of the target to be measured, film thickness information, refractive index or dielectric constant.
10. The method of claim 8, wherein the configuration information comprises: height information or film thickness information of the target surface to be measured, the method further comprising:
and performing three-dimensional reconstruction on the structural information to acquire the three-dimensional morphology information of the target to be detected.
11. The method according to claim 8, wherein the classifying the signal vectors to be measured whose structural information similarity satisfies a predetermined condition into the same set of vectors to be measured comprises:
acquiring a test sparse dictionary, wherein the signal vector to be tested can be linearly represented by atoms in the test sparse dictionary;
performing sparse representation on the signal vector to be tested according to the test sparse dictionary to obtain a sparse coefficient vector to be tested of the signal vector to be tested;
and classifying the signal vectors to be detected according to the sparse coefficient vectors to be detected, and dividing the signal vectors to be detected, of which the sparse coefficient vectors to be detected meet the same conditions, into a set to form a plurality of sets of the vectors to be detected.
12. The method according to claim 11, wherein the set of measurement models is obtained by the method of any one of claims 1 to 7;
the measurement models which are in one-to-one correspondence with the vector sets to be measured are that the similarity of the structural information of the vectors in the sample signal vector group of the measurement model and the structural information of the signal vectors to be measured in the corresponding vector sets to be measured meets the preset condition.
13. The method of claim 12, wherein the sample sparse dictionary is the same as the test sparse dictionary.
14. The method according to any one of claims 8 to 13, wherein before the classification processing of the signal vector under test, the method further comprises:
judging whether the number of the signal vectors to be detected is greater than or equal to a preset number threshold value or not;
when the number of the signal vectors to be detected is greater than or equal to the number threshold, executing the classification processing until structural information corresponding to the signal vectors to be detected in the vector set to be detected is obtained through a measurement model corresponding to the vector set to be detected;
and when the number of the signal vectors to be detected is smaller than a preset number threshold, performing nonlinear fitting on the signal vectors to be detected to obtain structural information corresponding to the signal vectors to be detected.
15. The method of any one of claims 8 to 13, wherein the signal vector under test comprises an interference light signal vector, a reflected light signal vector, or a scattered light signal vector.
16. The method according to any one of claims 8 to 13, wherein the step of obtaining a signal vector to be measured of each region to be measured, the signal vector to be measured being used for characterizing the structural information of the region to be measured, comprises:
respectively obtaining a light intensity signal vector to be detected of each region to be detected;
respectively carrying out Fourier transform on the light intensity signal vectors to be detected to obtain spectral signal vectors to be detected;
and respectively carrying out real number and dimensionality compression on the spectral signal vector to be detected to obtain the signal vector to be detected.
17. An apparatus for obtaining a measurement model group, comprising:
the device comprises a first acquisition module, a second acquisition module and a third acquisition module, wherein the first acquisition module is used for acquiring a plurality of groups of sample signal vector groups, each group of sample signal vector groups comprises a plurality of sample signal vectors, the sample signal vectors are used for representing the structural information of the sample, and the structural information similarity of the sample signal vectors in each group meets a preset condition;
and the processing module is used for performing modeling processing on the structural information of the sample according to the plurality of sample signal vectors in each group of sample signal vector groups respectively to obtain a measurement model of the structural information.
18. A measuring device, comprising:
the system comprises an input module, a detection module and a control module, wherein the input module is used for providing a target to be detected, and the target to be detected comprises a plurality of regions to be detected;
the second acquisition module is used for respectively acquiring a signal vector to be detected of each region to be detected, and the signal vector to be detected is used for representing the structural information of the region to be detected;
the classification module is used for classifying the signal vectors to be detected, wherein the signal vectors to be detected with the structural information similarity meeting the preset conditions are classified into the same vector set to be detected;
a third obtaining module, configured to obtain a measurement model group, where the measurement model group includes measurement models that correspond to the set of vectors to be measured one by one;
and the fourth obtaining module is used for obtaining the structural information corresponding to the signal vector to be measured in the vector set to be measured through the measurement model corresponding to the vector set to be measured.
19. A terminal device, comprising: a processor and a memory;
the processor is coupled to a memory, wherein the memory is configured to store program code and the processor is configured to invoke the program code to perform the method of any of claims 1-16.
20. A computer storage medium, characterized in that the computer storage medium stores a computer program comprising program instructions which, when executed by a processor, perform the method according to any one of claims 1-16.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202010111118.7A CN113297879A (en) | 2020-02-23 | 2020-02-23 | Acquisition method of measurement model group, measurement method and related equipment |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202010111118.7A CN113297879A (en) | 2020-02-23 | 2020-02-23 | Acquisition method of measurement model group, measurement method and related equipment |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN113297879A true CN113297879A (en) | 2021-08-24 |
Family
ID=77317815
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202010111118.7A Pending CN113297879A (en) | 2020-02-23 | 2020-02-23 | Acquisition method of measurement model group, measurement method and related equipment |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN113297879A (en) |
Citations (12)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103745239A (en) * | 2013-12-18 | 2014-04-23 | 广西生态工程职业技术学院 | Forest resource measuring method based on satellite remote sensing technology |
| CN104103060A (en) * | 2013-04-03 | 2014-10-15 | 华为技术有限公司 | Dictionary expression method and device in sparse model |
| CN106408656A (en) * | 2016-09-21 | 2017-02-15 | 江西天祥通用航空股份有限公司 | Building method and building system of 3D model |
| CN108090461A (en) * | 2017-12-29 | 2018-05-29 | 浙江大学宁波理工学院 | Three-dimensional face identification method based on sparse features |
| CN108763204A (en) * | 2018-05-21 | 2018-11-06 | 浙江大学 | A kind of multi-level text emotion feature extracting method and model |
| CN109543764A (en) * | 2018-11-28 | 2019-03-29 | 安徽省公共气象服务中心 | A kind of warning information legitimacy detection method and detection system based on intelligent semantic perception |
| CN109784351A (en) * | 2017-11-10 | 2019-05-21 | 财付通支付科技有限公司 | Data classification method, disaggregated model training method and device |
| CN109829505A (en) * | 2019-02-15 | 2019-05-31 | 北京工业大学 | A kind of threedimensional model classification method based on various features descriptor |
| CN109871896A (en) * | 2019-02-26 | 2019-06-11 | 北京达佳互联信息技术有限公司 | Data classification method, device, electronic equipment and storage medium |
| CN110516677A (en) * | 2019-08-23 | 2019-11-29 | 上海云绅智能科技有限公司 | A kind of neural network recognization model, target identification method and system |
| CN110782520A (en) * | 2018-07-30 | 2020-02-11 | 韦伯斯特生物官能(以色列)有限公司 | Left atrial shape reconstruction using neural networks based on sparse position measurements |
| CN210089643U (en) * | 2019-07-12 | 2020-02-18 | 长鑫存储技术有限公司 | Measuring device for measuring three-dimensional shape |
-
2020
- 2020-02-23 CN CN202010111118.7A patent/CN113297879A/en active Pending
Patent Citations (12)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104103060A (en) * | 2013-04-03 | 2014-10-15 | 华为技术有限公司 | Dictionary expression method and device in sparse model |
| CN103745239A (en) * | 2013-12-18 | 2014-04-23 | 广西生态工程职业技术学院 | Forest resource measuring method based on satellite remote sensing technology |
| CN106408656A (en) * | 2016-09-21 | 2017-02-15 | 江西天祥通用航空股份有限公司 | Building method and building system of 3D model |
| CN109784351A (en) * | 2017-11-10 | 2019-05-21 | 财付通支付科技有限公司 | Data classification method, disaggregated model training method and device |
| CN108090461A (en) * | 2017-12-29 | 2018-05-29 | 浙江大学宁波理工学院 | Three-dimensional face identification method based on sparse features |
| CN108763204A (en) * | 2018-05-21 | 2018-11-06 | 浙江大学 | A kind of multi-level text emotion feature extracting method and model |
| CN110782520A (en) * | 2018-07-30 | 2020-02-11 | 韦伯斯特生物官能(以色列)有限公司 | Left atrial shape reconstruction using neural networks based on sparse position measurements |
| CN109543764A (en) * | 2018-11-28 | 2019-03-29 | 安徽省公共气象服务中心 | A kind of warning information legitimacy detection method and detection system based on intelligent semantic perception |
| CN109829505A (en) * | 2019-02-15 | 2019-05-31 | 北京工业大学 | A kind of threedimensional model classification method based on various features descriptor |
| CN109871896A (en) * | 2019-02-26 | 2019-06-11 | 北京达佳互联信息技术有限公司 | Data classification method, device, electronic equipment and storage medium |
| CN210089643U (en) * | 2019-07-12 | 2020-02-18 | 长鑫存储技术有限公司 | Measuring device for measuring three-dimensional shape |
| CN110516677A (en) * | 2019-08-23 | 2019-11-29 | 上海云绅智能科技有限公司 | A kind of neural network recognization model, target identification method and system |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Gupta et al. | Non-Gaussian information from weak lensing data via deep learning | |
| CN105427300B (en) | A kind of hyperspectral image abnormal detection method based on low-rank representation and study dictionary | |
| KR100910003B1 (en) | Profile segmentation for integrated circuit instrumentation | |
| CN109685152A (en) | A kind of image object detection method based on DC-SPP-YOLO | |
| CN111126575B (en) | Gas sensor array mixed gas detection method and device based on machine learning | |
| AU2010291853B2 (en) | Estimating reflectance model parameters from an image | |
| CN104657717B (en) | A kind of pedestrian detection method based on layering nuclear sparse expression | |
| CN112287839A (en) | SSD infrared image pedestrian detection method based on transfer learning | |
| CN113283419B (en) | Convolutional neural network pointer instrument image reading identification method based on attention | |
| CN105103166A (en) | Systems and methods for texture assessment of a coating formulation | |
| CN110704652A (en) | Vehicle image fine-grained retrieval method and device based on multiple attention mechanism | |
| CN112950682A (en) | Cross-scale point cloud registration algorithm based on wavelet transformation and scale representation | |
| CN115019145A (en) | Method for monitoring settlement deformation of road surface of Qinghai-Tibet plateau highway | |
| CN107578448B (en) | CNN-based method for identifying number of spliced curved surfaces contained in calibration-free curved surface | |
| Hosny | Fast and low-complexity method for exact computation of 3D Legendre moments | |
| Yarovyi et al. | An intelligent system of neural networking recognition of multicolor spot images of laser beam profile | |
| Viel et al. | Hyperspectral image classification: an analysis employing CNN, LSTM, transformer, and attention mechanism | |
| CN114663755A (en) | Neural network training method for intelligent detection of semiconductor wafer roughness | |
| CN112288784B (en) | Descriptor neighborhood self-adaptive weak texture remote sensing image registration method | |
| Sun et al. | Pruning filters with L1-norm and standard deviation for CNN compression | |
| CN113297879A (en) | Acquisition method of measurement model group, measurement method and related equipment | |
| CN117011274A (en) | Automatic glass bottle detection system and method thereof | |
| CN115719363B (en) | Environment sensing method and system capable of performing two-dimensional dynamic detection and three-dimensional reconstruction | |
| CN116842827A (en) | Electromagnetic performance boundary model construction method for unmanned aerial vehicle flight control system | |
| Ji et al. | Faster region-based convolutional neural network method for estimating parameters from Newton’s rings |
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 |