Spectral confocal multi-peak extraction method, module, computer device and storage medium

Abstract

Embodiments of the present application disclose a spectrum confocal multi-peak extraction method, a module, a computer device and a storage medium. The method comprises: performing frequency domain filtering and space domain filtering on collected spectrum signals respectively, and obtaining frequency domain filtered signals and space domain filtered signals correspondingly; processing the frequency domain filtered signals by using a symmetric zero area method to obtain symmetric zero area differential signals; screening target regions in the space domain filtered signals according to the symmetric zero area differential signals and the space domain filtered signals; performing peak extraction on each target region by using a weighted centroid method to obtain a peak set composed of peaks corresponding to each target region; and screening a plurality of required peaks from the peak set. The spectrum confocal multi-peak extraction method can accurately realize spectrum confocal multi-peak extraction, and has good adaptability to signals with low smoothness.

Description

TECHNICAL FIELD

[0001] The present application relates to the field of spectral confocal measurement technology, and particularly relates to a spectral confocal multi-peak extraction method, a module, computer equipment and a storage medium. BACKGROUND

[0002] The spectral confocal measurement technology is a new type of non-contact optical measurement technology using optical dispersion technology, which has the advantages of high measurement accuracy, wide material adaptability, multi-layer measurement for transparent and semi-transparent materials, etc. At present, some spectral confocal sensors have been applied in many industrial fields such as wafer surface measurement, circuit board surface measurement, etc.

[0003] The basic principle of the spectral confocal measurement technology is: using white light (composite light) to disperse along the axial direction near the surface of the measured object (usually using special optical components to enlarge the dispersion effect of light of different wave bands), and using a spectrometer system to obtain the reflection spectrum of different spectral light on the surface of the measured object, analyzing the relative intensity of the reflection spectrum (echo), extracting the peak position of the spectral signal, and converting the peak value into a corresponding measurement distance value.

[0004] Among the currently disclosed spectral confocal peak extraction methods, the commonly used methods are the centroid method, the weighted centroid method, and the symmetric zero area method. In addition, there is an automatic threshold segmentation algorithm, but it is rarely used. The existing spectral confocal peak extraction methods all have limitations in the scope of application, and each has its own advantages and disadvantages.

[0005] Among them, the centroid method and the weighted centroid method have high accuracy in the scene where the relative intensity distribution of the spectrum is relatively symmetrical, but the centroid method and the weighted centroid method cannot be directly applied to the extraction of multiple peaks.

[0006] The symmetric zero area method is suitable for finding small peaks next to large peaks and can be applied to the extraction of multiple peaks, but the symmetric zero area method requires a high degree of signal smoothness, and the direct calculation result is still a one-dimensional signal, which still needs to be analyzed to obtain the single or multiple peak positions corresponding to single or multi-layer measurement. Therefore, this method is prone to local small signal optimal solution, resulting in incorrect peak position determination.

[0007] The automatic threshold segmentation method can be applied to the determination of multiple peak positions, but this method is limited to determining the distribution area of each peak, and must be supplemented by other peak extraction methods to extract the accurate peak position. SUMMARY

[0008] The present application provides a spectral confocal multi-peak extraction method, a module, computer equipment and a storage medium, which can accurately realize spectral confocal multi-peak extraction and has good adaptability to low smoothness signals.

[0009] In a first aspect, the present application provides a spectral confocal multi-peak extraction method, comprising: performing frequency domain filtering and spatial domain filtering on the collected spectral signal respectively, to obtain a frequency domain filtered signal and a spatial domain filtered signal; processing the frequency domain filtered signal using a symmetric zero area method to obtain a symmetric zero area differential signal; screening a target region in the spatial domain filtered signal according to the symmetric zero area differential signal and the spatial domain filtered signal; performing peak extraction on each target region using a weighted centroid method to obtain a peak set composed of peaks corresponding to each target region; and screening a plurality of peaks required from the peak set.

[0010] In some specific embodiments, the processing of the frequency domain filtered signal using the symmetric zero area method to obtain the symmetric zero area differential signal comprises: determining a window value; generating a corresponding one-dimensional convolution operator using a symmetric zero area transform function according to the window value; and performing convolution processing on the frequency domain filtered signal according to the one-dimensional convolution operator.

[0011] Further, the determination of the window value comprises: determining the window value according to a full width at half maximum of a peak of the frequency domain filtered signal.

[0012] In some specific embodiments, the screening of the target region in the spatial domain filtered signal according to the symmetric zero area differential signal and the spatial domain filtered signal comprises: dividing a zero value upper interval and a zero value lower interval of the symmetric zero area differential signal according to a zero value position line of the symmetric zero area differential signal; and determining a region in the spatial domain filtered signal corresponding to the zero value lower interval as the target region.

[0013] In some specific embodiments, the peak extraction on each target region using the weighted centroid method comprises:

[0014]

[0015] wherein x is a peak, n is a position of a signal, and I n is a signal size at the signal n.

[0016] In some specific embodiments, the peak extraction on each target region using the weighted centroid method comprises:

[0017]

[0018] wherein x is a peak, n is a position of a signal, and I n is a signal size at the signal n, and m>2.

[0019] In a second aspect, the present application provides a spectral confocal multi-peak extraction module, comprising: a frequency domain filtering unit configured to perform frequency domain filtering on a spectral signal to obtain a frequency domain filtered signal; a symmetric zero area processing unit configured to perform symmetric zero area processing on the frequency domain filtered signal to obtain a symmetric zero area differential signal; a spatial domain filtering unit configured to perform spatial domain filtering on the spectral signal to obtain a spatial domain filtered signal; a target region screening unit configured to screen a target region in the spatial domain filtered signal according to the symmetric zero area differential signal and the spatial domain filtered signal; a weighted centroid processing unit configured to perform peak extraction on each target region using a weighted centroid method to obtain a peak set composed of peaks corresponding to each target region; and a peak screening unit configured to screen a plurality of required peaks from the peak set.

[0020] In a third aspect, the present application provides a computer device, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor implements the steps of any of the above spectral confocal multi-peak extraction methods when executing the computer program.

[0021] In a fourth aspect, the present application provides a computer storage medium, comprising a computer program, wherein the computer program implements the steps of any of the above spectral confocal multi-peak extraction methods when executed by a processor.

[0022] The present application provides a spectral confocal multi-peak extraction method, module, computer device, and storage medium. Frequency domain filtering on the collected spectral signal can obtain a relatively smooth frequency domain filtered signal, and the smooth frequency domain filtered signal can be processed by the symmetric zero area method to obtain a symmetric zero area differential signal. On the other hand, spatial domain filtering on the collected spectral signal can obtain a relatively less smooth spatial domain filtered signal that can better preserve the original characteristics of the spectral signal. Based on the joint processing of the symmetric zero area differential signal and the spatial domain filtered signal, the target region can be screened from the spatial domain filtered signal, thereby avoiding the problem of local optimal solution and preventing peak position judgment errors. For the target region, the weighted centroid method can be used to perform peak extraction on each target region to obtain a peak set composed of peaks corresponding to each target region, and finally a plurality of required peaks can be screened from the peak set. The technical solution of the present application can accurately implement spectral confocal multi-peak extraction, and has good adaptability to low-smoothness signals compared with the prior art. BRIEF DESCRIPTION OF DRAWINGS

[0023] In order to more clearly illustrate the technical solutions of the embodiments of the present application, the drawings needed to be used in the embodiments description will be briefly introduced. Obviously, the drawings in the following description are some embodiments of the present application, and for those skilled in the art, other drawings can be obtained without creative labor.

[0024] Figure 1 A schematic diagram of a spectrum signal collected by a spectrum confocal multi-peak extraction method provided by the embodiments of the present application;

[0025] Figure 2 A schematic diagram of a frequency domain filtered signal of a spectrum confocal multi-peak extraction method provided by the embodiments of the present application;

[0026] Figure 3 A schematic diagram of a symmetric zero area differential signal of a spectrum confocal multi-peak extraction method provided by the embodiments of the present application;

[0027] Figure 4 A schematic diagram of a spatial domain filtered signal of a spectrum confocal multi-peak extraction method provided by the embodiments of the present application;

[0028] Figure 5 A schematic diagram of screening a target region in a spatial domain filtered signal of a spectrum confocal multi-peak extraction method provided by the embodiments of the present application;

[0029] Figure 6 A flowchart of a spectrum confocal multi-peak extraction method provided by the embodiments of the present application;

[0030] Figure 7 A schematic diagram of screening a plurality of peaks needed by a spectrum confocal multi-peak extraction method provided by the embodiments of the present application;

[0031] Figure 8 A schematic block diagram of a spectrum confocal multi-peak extraction module provided by the embodiments of the present application. DETAILED DESCRIPTION

[0032] The technical solutions in the embodiments of the present application will be described clearly and completely with reference to the drawings in the embodiments of the present application. Obviously, the described embodiments are some embodiments of the present application, but not all the embodiments. Based on the embodiments in the present application, all the other embodiments obtained by those skilled in the art without creative labor are within the protection scope of the present application.

[0033] It should be understood that the terms "comprise" and "comprising" when used in this specification and accompanying claims, signify the presence of the stated features, integers, steps, operations, elements, and / or components but do not preclude the presence or addition of one or more other features, integers, steps, operations, elements, components, and / or groups thereof.

[0034] It should also be understood that the terms used in the specification and the appended claims are intended to be interpreted broadly and in a manner consistent with the principles of the application, and not necessarily in a narrow or legal sense. As used in this specification and the appended claims, the singular forms "a," "an," and "the" include plural referents unless the content clearly dictates otherwise.

[0035] It should also be further understood that the term "and / or" as used herein refers to and encompasses any and all possible combinations of one or more of the associated listed items, and that the term is intended to be interpreted in its broadest and intended sense.

[0036] Reference is made to Figure 6 , Figure 6 is a flowchart of a spectral confocal multi-peak extraction method, the method steps of the spectral confocal multi-peak extraction method comprising the following five steps:

[0037] Step 1: The collected spectral signal is subjected to frequency domain filtering and spatial domain filtering, and the frequency domain filtered signal and the spatial domain filtered signal are obtained correspondingly.

[0038] Reference is made to Figure 1 , Figure 1 is a schematic diagram of a collected spectral signal of a spectral confocal multi-peak extraction method provided by the embodiment of the present application. The spectral signal is usually collected by a spectral confocal sensor. The spectral signal is easily affected by factors such as stray light of the spectral confocal sensor, environmental noise, and image sensor resolution, resulting in unavoidable Gaussian white noise in the spectral signal. If the spectral signal is directly subjected to "second-order difference operator convolution", the result will have a large deviation.

[0039] The spectral signal is subjected to frequency domain filtering processing, and a frequency domain filtered signal with relatively high smoothness can be obtained, as shown in the schematic diagram of the frequency domain filtered signal in Figure 2 . The available frequency domain filter includes an ideal low-pass filter, a Butterworth low-pass filter, a Gaussian low-pass filter, etc., and those skilled in the art can apply as needed.

[0040] The spectral signal is subjected to frequency domain filtering processing, and a spatial domain filtered signal with relatively low smoothness can be obtained, as shown in the schematic diagram of the spatial domain filtered signal in Figure 4 .

[0041] The spatial domain filtering can eliminate the Gaussian white noise in the spectral signal to some extent, and can better preserve the original characteristics of the spectral signal, and can also ensure the accuracy of the weighted centroid method processing in the later stage. Among them, the spatial domain filtering algorithm can adopt median filtering, mean filtering and other common filtering algorithms.

[0042] Step 2: The symmetric zero area method is used to process the frequency domain filtered signal to obtain a symmetric zero area differential signal.

[0043] Since the smoothness of the frequency domain filtered signal is relatively high, the symmetric zero area method can directly process the frequency domain filtered signal. A specific scheme for processing the frequency domain filtered signal by the symmetric zero area method is: first, determine a fixed size window, that is, a window value, then generate a one-dimensional convolution operator corresponding to the symmetric zero area transformation function, and then perform convolution processing on the frequency domain filtered signal according to the one-dimensional convolution operator to obtain a symmetric zero area differential signal. The schematic diagram of the symmetric zero area differential signal can be seen from Figure 3 .

[0044] Among them, in this embodiment, specifically, the window value is determined according to the half-height width of the peak value of the frequency domain filtered signal. The larger the half-height width of the frequency domain filtered signal, the larger the window value. In specific operation, the variation range of the window value can be initially set according to the empirical value, and then the window value is gradually changed in the range to select the window with the best filtering effect as the fixed window size. It should be noted that the best filtering effect in this embodiment should be the effect of eliminating high-frequency signals and completely preserving low-frequency signals, which can reflect the trend of signal data change.

[0045] Step 3: According to the symmetric zero area differential signal and the spatial domain filtered signal, the target region in the spatial domain filtered signal is selected.

[0046] Please refer to Figure 5 The symmetric zero area differential signal and the spatial domain filtered signal are jointly processed, wherein first, the zero value upper interval and the zero value lower interval of the symmetric zero area differential signal are divided according to the zero value position line of the symmetric zero area differential signal.

[0047] Regarding the division of the zero value position line, the specific implementation method is: when a local signal composed of a specific position in the signal and its adjacent values behaves as a "convex" function, the final calculation result is positive, and the greater the curvature, the greater the calculated value. Similarly, the part of the "concave" function is calculated as negative.

[0048] According to the "0" position, the symmetric zero area differential signal can be divided into several sub-regions. According to the region in the foregoing several sub-regions whose value is greater than "0", the target region corresponding to the position in the spatial domain filtered signal is selected, and the target region is shown as the gray shadow covered region in 5.

[0049] Step 4: Use the weighted centroid method to extract peak values ​​for each target region, and obtain a peak value set consisting of the peak values ​​corresponding to each target region.

[0050] After obtaining each target area (e.g.) Figure 6 After identifying regions 1, 2, 3, ..., n as shown, it is necessary to extract the peak values ​​for each target region separately. The peak values ​​corresponding to each target region constitute a peak set (e.g., ...). Figure 6 The peaks shown are 1, 2, 3, ..., n.

[0051] In this embodiment, the weighted centroid method is used to obtain the peak value of the target region. Since each target region is unique and definite, the weighted centroid method can be used directly to obtain the peak value of each target region separately, effectively overcoming the limitation that the weighted centroid method cannot be directly applied to the extraction of multiple peak values.

[0052] Specifically, the method for extracting peak values ​​for each target region using the weighted centroid method is as follows:

[0053]

[0054] Where x is the peak value, n is the position of the signal, and I n Let n be the signal magnitude at point n.

[0055] In actual implementation, in order to highlight the semaphore I n The impact on the final peak value can be expressed in the above formula. Promoted as Right now:

[0056]

[0057] Where m>2.

[0058] The value of m is appropriately increased based on 2. When the signal value is relatively small, it is beneficial to amplify the difference in peak value, which makes it easier to select the required peak value later. At the same time, it can also be used to meet different accuracy requirements.

[0059] Step 5: Select the desired peaks from the peak set.

[0060] Select the desired peaks from the obtained peak set. The specific selection strategy can be based on a reference value such as the semaphore size P at the peak, the full width at half maximum (FWHM), or the relative distance.

[0061] Please see Figure 7 ,by Figure 7 To illustrate the peak selection process, an example is provided to help those skilled in the art understand it.Figure 7 In the signal in FIG. 7, there are four peaks, and it is assumed that two peaks need to be screened out as the final multi-layer measurement peaks. If the peaks are screened out by the signal amount P at the peak, and the value of the screened-out peak is required to be the largest, obviously P3>P1>P2>P4, then the two screened-out peaks should be peak 3 and peak 1. If the peaks are screened out by the half-height width FWHM, and the half-height width at the peak is required to be as small as possible, obviously FWHM3<FWHM4<FWHM2<FWHM1, and thus the two screened-out peaks should be peak 3 and peak 4. It can be seen that if the focus and the screening strategy used in screening are different, the final screening result is also different. In actual peak screening, a person skilled in the art can select a screening strategy according to specific needs.

[0062] Referring to FIG. 8, Figure 8 FIG. 8 is a schematic block diagram of a spectrum confocal multi-peak extraction module according to the embodiment. The spectrum confocal multi-peak extraction module includes: a frequency domain filtering unit, configured to perform frequency domain filtering on the spectrum signal to obtain a frequency domain filtered signal; a symmetric zero area processing unit, configured to process the frequency domain filtered signal by using a symmetric zero area method to obtain a symmetric zero area differential signal; a spatial domain filtering unit, configured to perform spatial domain filtering on the spectrum signal to obtain a spatial domain filtered signal; a target region screening unit, configured to screen out a target region in the spatial domain filtered signal according to the symmetric zero area differential signal and the spatial domain filtered signal; a weighted centroid processing unit, configured to extract a peak value of each target region by using a weighted centroid method to obtain a peak value set composed of peak values corresponding to each target region; and a peak screening unit, configured to screen out a plurality of required peak values from the peak value set.

[0063] The module can be a separate module and be sold directly in the market as a spare part.

[0064] In addition, the spectrum confocal multi-peak extraction method mentioned in the embodiment can be integrated in a computer device, and the computer device includes a memory, a processor, and a computer program stored in the memory and executable on the processor. The computer program is executed by the processor to implement the aforementioned spectrum confocal multi-peak extraction method.

[0065] Of course, the spectrum confocal multi-peak extraction method mentioned in the embodiment can also be integrated in a computer storage medium, and the computer storage medium includes a computer program. The computer program is executed by the processor to implement the aforementioned spectrum confocal multi-peak extraction method.

[0066] Those skilled in the art can realize that each algorithm step described in connection with the embodiments disclosed herein can be realized by electronic hardware, computer software, or a combination of both. In order to clearly illustrate the interchangeability of hardware and software, the general description of each example has been described in terms of functional generalities. Whether the functions are performed in hardware or software depends on the specific application and design constraints of the technical solution. A skilled person can use different methods to implement the described functions for each specific application, but such implementation should not be considered beyond the scope of the present application. The steps in the method of the embodiments of the present application can be adjusted, combined and deleted in sequence according to actual needs.

[0067] The above is only a specific implementation of the present application, but the protection scope of the present application is not limited to this. Any skilled person in the art can easily think of various equivalent modifications or replacements within the technical scope disclosed by the present application, and these modifications or replacements should be covered within the protection scope of the present application. Therefore, the protection scope of the present application should be subject to the protection scope of the claims.

Claims

1. A method for extracting multiple peak values ​​from a spectral confocal focal plane, characterized in that, include: The acquired spectral signals are subjected to frequency domain filtering and spatial domain filtering respectively, resulting in frequency domain filtered signals and spatial domain filtered signals. The frequency domain filtered signal is processed using the symmetrical zero-area method to obtain a symmetrical zero-area differential signal; Based on the symmetrical zero-area difference signal and the spatial filtering signal, the target region within the spatial filtering signal is selected. Peak values ​​are extracted from each target region using the weighted centroid method to obtain a peak set consisting of the peak values ​​corresponding to each target region. Select the desired peaks from the set of peaks.

2. The method for extracting multiple peak values ​​from a confocal spectrum according to claim 1, characterized in that, The process of applying the symmetric zero-area method to the frequency domain filtered signal to obtain a symmetric zero-area differential signal includes: Determine the window value; Based on the window value, a corresponding one-dimensional convolution operator is generated using a symmetric zero-area transformation function; The frequency domain filtered signal is convolved according to the one-dimensional convolution operator.

3. The method for extracting multiple peak values ​​from a confocal spectrum according to claim 2, characterized in that, The determined window value includes: The window value is determined based on the half-width at half-maximum (WHM) of the peak value of the frequency domain filtered signal.

4. The method for extracting multiple peak values ​​from a confocal spectrum according to claim 1, characterized in that, The step of filtering out the target region within the spatial filtering signal based on the symmetrical zero-area difference signal and the spatial filtering signal includes: Based on the zero-value position line of the symmetrical zero-area difference signal, the upper zero-value interval and the lower zero-value interval of the symmetrical zero-area difference signal are divided. The region within the spatial filtering signal that corresponds to the interval above zero is determined as the target region.

5. The method for extracting multiple peak values ​​from a confocal spectrum according to claim 1, characterized in that, The method for peak extraction of each target region using the weighted centroid method includes: Where x is the peak position, n is the signal position, and I n Let n be the signal magnitude at point n.

6. The method for extracting multiple peak values ​​from a confocal spectrum according to claim 1, characterized in that, The method for peak extraction of each target region using the weighted centroid method includes: Where x is the peak position, n is the position of the signal, In is the signal magnitude at position n, and m>2.

7. A spectral confocal multi-peak extraction module, characterized in that, include: The frequency domain filtering unit is used to perform frequency domain filtering on the spectral signal to obtain the frequency domain filtered signal. A symmetrical zero-area processing unit is used to process the frequency domain filtered signal using the symmetrical zero-area method to obtain a symmetrical zero-area differential signal. The spatial filtering unit is used to perform frequency domain filtering on the spectral signal to obtain the spatial filtered signal. The target region filtering unit is used to filter out the target region within the spatial filtering signal based on the symmetrical zero-area difference signal and the spatial filtering signal. A weighted centroid processing unit is used to extract peak values ​​from each target region using the weighted centroid method, thereby obtaining a set of peak values ​​consisting of the peak values ​​corresponding to each target region. A peak filtering unit is used to filter out multiple desired peaks from the set of peaks.

8. A computer device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that: When the processor executes the computer program, it implements the steps of the method as described in any one of claims 1-6.

9. A computer storage medium comprising a computer program, characterized in that: When the computer program is executed by a processor, it implements the steps of the method as described in any one of claims 1-6.

Images