CN114386456A - Terahertz time-domain signal noise reduction method, image reconstruction method and system - Google Patents
Terahertz time-domain signal noise reduction method, image reconstruction method and system Download PDFInfo
- Publication number
- CN114386456A CN114386456A CN202111538693.6A CN202111538693A CN114386456A CN 114386456 A CN114386456 A CN 114386456A CN 202111538693 A CN202111538693 A CN 202111538693A CN 114386456 A CN114386456 A CN 114386456A
- Authority
- CN
- China
- Prior art keywords
- noise
- terahertz
- signal
- terahertz time
- imaging
- 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
- 238000000034 method Methods 0.000 title claims abstract description 54
- 230000009467 reduction Effects 0.000 title claims abstract description 49
- 238000003384 imaging method Methods 0.000 claims abstract description 55
- 238000000354 decomposition reaction Methods 0.000 claims abstract description 49
- 238000001914 filtration Methods 0.000 claims abstract description 46
- 238000012880 independent component analysis Methods 0.000 claims abstract description 33
- 238000012545 processing Methods 0.000 claims abstract description 20
- 238000001228 spectrum Methods 0.000 claims description 13
- 230000003287 optical effect Effects 0.000 claims description 8
- 238000001328 terahertz time-domain spectroscopy Methods 0.000 claims description 6
- 230000005540 biological transmission Effects 0.000 claims description 4
- 238000005259 measurement Methods 0.000 claims description 4
- 230000003044 adaptive effect Effects 0.000 claims description 3
- 230000000295 complement effect Effects 0.000 claims description 3
- 238000012360 testing method Methods 0.000 claims description 2
- 230000007547 defect Effects 0.000 description 20
- 229920002430 Fibre-reinforced plastic Polymers 0.000 description 8
- 239000002131 composite material Substances 0.000 description 8
- 239000011151 fibre-reinforced plastic Substances 0.000 description 8
- 238000001514 detection method Methods 0.000 description 6
- 230000000694 effects Effects 0.000 description 6
- 238000005516 engineering process Methods 0.000 description 5
- 238000004364 calculation method Methods 0.000 description 4
- 238000004458 analytical method Methods 0.000 description 2
- 230000007613 environmental effect Effects 0.000 description 2
- 230000006872 improvement Effects 0.000 description 2
- 238000007689 inspection Methods 0.000 description 2
- 239000000463 material Substances 0.000 description 2
- 230000005457 Black-body radiation Effects 0.000 description 1
- 238000010521 absorption reaction Methods 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 238000003745 diagnosis Methods 0.000 description 1
- 238000010586 diagram Methods 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 238000012544 monitoring process Methods 0.000 description 1
- 238000005457 optimization Methods 0.000 description 1
- 230000008569 process Effects 0.000 description 1
- 238000003672 processing method Methods 0.000 description 1
- 238000011946 reduction process Methods 0.000 description 1
- 239000000126 substance Substances 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
- G06F2218/02—Preprocessing
- G06F2218/04—Denoising
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/21—Design or setup of recognition systems or techniques; Extraction of features in feature space; Blind source separation
- G06F18/213—Feature extraction, e.g. by transforming the feature space; Summarisation; Mappings, e.g. subspace methods
- G06F18/2135—Feature extraction, e.g. by transforming the feature space; Summarisation; Mappings, e.g. subspace methods based on approximation criteria, e.g. principal component analysis
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Data Mining & Analysis (AREA)
- Artificial Intelligence (AREA)
- Physics & Mathematics (AREA)
- General Engineering & Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Signal Processing (AREA)
- Life Sciences & Earth Sciences (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Bioinformatics & Computational Biology (AREA)
- Evolutionary Biology (AREA)
- Evolutionary Computation (AREA)
- Investigating Or Analysing Materials By Optical Means (AREA)
Abstract
The invention discloses a terahertz time-domain signal noise reduction method, an image reconstruction method and an image reconstruction system, and belongs to the field of terahertz signal processing and image reconstruction. The noise-containing mode obtained by empirical mode decomposition is screened by a threshold filtering method to obtain a virtual noise channel, and then the noise-containing signal is separated into noise and a noise-reduced signal by independent component analysis, so that the problem of mode aliasing in the empirical mode decomposition is well solved, and compared with the existing method, the signal-to-noise ratio of the noise-reduced signal is greatly improved. When the method is further used for imaging, compared with the existing imaging method, the signal-to-noise ratio and the contrast ratio of the imaging are obviously improved.
Description
Technical Field
The invention belongs to the field of terahertz signal processing and image reconstruction, and particularly relates to a terahertz time-domain signal noise reduction method, an image reconstruction method and an image reconstruction system.
Background
The terahertz has the characteristics of no ionization, no damage and the like, the vibration and rotation energy levels of various organic molecules fall in the terahertz waveband, the requirements of the terahertz spectrum detection and imaging technology on the working environment and the sample form are not high, the operation is simple, and the terahertz spectrum detection and imaging technology can be conveniently combined with other technologies for use. The terahertz spectrum detection and imaging technology has wide application prospects in the aspects of researching biomacromolecule characteristics, medical diagnosis, security inspection, material detection, environmental monitoring and the like. The terahertz time-domain signal of the detected sample acquired by the terahertz time-domain spectroscopy system can reflect the amplitude and phase information of the terahertz pulse, and the photoelectric property of the detected sample in the terahertz waveband can be extracted. When the measured sample is further scanned in two dimensions, terahertz imaging can be performed on the measured sample according to the maximum value and the minimum value of terahertz time-domain signals of each point of the sample or the amplitude value at a specific time point, or based on the amplitude value at a certain frequency in a frequency spectrum obtained after Fourier transform is performed on the time-domain signals, or based on the size of optical parameters of the sample extracted from the time-domain signals at a certain frequency.
However, in the terahertz time-domain spectroscopy system, various noises originating from instability of the laser and the photoelectric device, environmental interference, black body radiation, electronic noise in the photoelectric device, and the like may be affected, so that the signal-to-noise ratio of the terahertz time-domain signal is low in some cases. Especially, when the terahertz source intensity is low or the absorption of the detected sample in the terahertz waveband is strong, the influence of noise on the terahertz time-domain signal analysis and imaging is more serious. The terahertz time-domain signal is subjected to effective noise reduction treatment, and the method has very important significance for sample property analysis and imaging identification.
Researchers have proposed various methods for noise reduction of terahertz time-domain signals. For example, in order to filter high-frequency noise, the most common means is to filter the influence of noise of a specific frequency band on a terahertz signal through a low-pass filter, but this method can only remove high-frequency noise in a noisy signal, and cannot remove noise in a low-frequency region, and a noise-reduced frequency region needs to be set by itself and has no self-adaptability; researchers also use empirical mode decomposition to decompose the original terahertz signal into a plurality of eigenmode functions in a self-adaptive manner, and specific eigenmode functions are selected to be added and reconstructed, so that noise influence is reduced.
Zhou et al in 2019 further combined empirical mode decomposition with independent component analysis, applied to noise reduction of photoacoustic imaging signals, constructed a noise channel through a noise-dominated eigenmode function decomposed by the empirical mode, and separated noise and signals by independent component analysis, thereby achieving better noise reduction effect than that achieved by using empirical mode decomposition alone. The method combining the empirical mode decomposition and the independent component analysis also has the obvious defect that because of the inherent defects of the empirical mode decomposition, serious mode aliasing phenomenon exists in noise-containing modes, the mode aliasing phenomenon causes noise and signals to be mixed in the intrinsic mode functions leading to the noise, and the noise channel is constructed by utilizing the intrinsic mode functions, so that useful signal information is inevitably lost, and the noise reduction effect is greatly influenced. Particularly, when terahertz spectrum detection and imaging are applied, a plurality of useful weak multilayer reflection signal components exist in the acquired terahertz signals due to multiple reflections during testing of layered samples, and if noise reduction is performed based on the empirical mode decomposition and independent component analysis combined method, a plurality of useful multilayer reflection signal information is lost due to a mode aliasing phenomenon, signal noise reduction and image reconstruction effects are not ideal, and improvement and optimization are urgently needed.
Disclosure of Invention
Aiming at the defects of the related technology, the invention aims to provide a terahertz time-domain signal noise reduction method, an image reconstruction method and an image reconstruction system, and aims to solve the problem that the noise reduction effect is not ideal due to mode aliasing in terahertz spectrum detection and imaging application.
In order to achieve the above object, the present invention provides a terahertz time-domain signal noise reduction method, which is characterized by comprising the following steps:
s1, carrying out empirical mode decomposition on terahertz time-domain signals containing noise to obtain a plurality of eigenmode functions;
s2, selecting an eigenmode function with leading noise by using a correlation coefficient;
s3, filtering the eigen mode function of the noise leading by using a threshold value, and then filtering a signal part to obtain a modal component of the noise part;
s4, constructing a virtual noise channel by using the modal components of the noise part;
and S5, separating the original signal into noise and a noise-reduced signal through independent component analysis by using the virtual noise channel.
Further, the empirical mode decomposition may be a collective empirical mode decomposition, a complementary collective empirical mode decomposition, or a variational modal decomposition.
Further, in step S1, the empirical mode decomposition result of the terahertz time-domain signal x (t) is represented as
Further, the step S2 specifically includes the following steps:
selecting noise dominant eigenmode function by using correlation coefficient, and selecting high-frequency eigenmode function c with correlation coefficient less than 0.2i(t) and the next eigenmode function ci+1(t) eigenmode functions considered to be noise dominated.
Further, the threshold filtering of step S3 is soft threshold filtering, hard threshold filtering or adaptive threshold filtering.
Further, the independent component analysis of the step S5 is a fast independent component analysis or a fourier independent component analysis.
Another aspect of the present invention provides a terahertz image reconstruction method, including the following steps:
performing two-dimensional imaging on a tested sample to obtain terahertz time-domain signals with noise at each point on the tested sample;
obtaining noise-reduced signals of each point on the measured sample by adopting the terahertz time-domain signal noise reduction method;
according to different application requirements, selecting the maximum value, the minimum value or the intensity at a specific time point of the noise-reduced signal, or selecting an amplitude value at a certain frequency in a frequency spectrum obtained after Fourier transform is performed on the noise-reduced signal, or selecting the size of an optical parameter extracted based on the noise-reduced signal at a certain frequency, and reconstructing the terahertz image of the detected sample.
Further, the two-dimensional imaging of the measured sample comprises: terahertz point-by-point scanning imaging or terahertz focal plane imaging is adopted, and transmission imaging or reflection imaging is combined.
Still another aspect of the present invention provides a terahertz image reconstruction system, including:
the imaging module is used for carrying out two-dimensional imaging on a tested sample to obtain terahertz time-domain signals with noise at each point on the tested sample;
the noise reduction processing module is used for carrying out noise reduction processing on the terahertz signals containing the noise of each point to obtain the signals of each point on the detected sample after noise reduction;
the image reconstruction module is used for selecting the maximum value and the minimum value of the noise-reduced signal or the intensity at a specific time point according to different application requirements, or selecting an amplitude value at a certain frequency in a frequency spectrum after Fourier transform is carried out on the noise-reduced signal, or selecting the size of an optical parameter at a certain frequency extracted based on the noise-reduced signal, and reconstructing the terahertz image of the detected sample;
the result comparison module is used for displaying and comparing the reconstructed terahertz images of different tested samples;
wherein the noise reduction processing module further comprises:
the empirical mode decomposition module is used for performing empirical mode decomposition on the terahertz time-domain signal containing the noise to obtain a plurality of eigenmode functions;
the threshold filtering module is used for receiving the processing result of the empirical mode decomposition module, selecting an eigenmode function with leading noise by using a correlation coefficient, filtering the selected eigenmode function with leading noise by using a threshold, filtering a signal part to obtain a modal component of the noise part, and constructing a virtual noise channel by using the part of the modal component of the noise obtained by filtering;
and the independent component analysis module receives the processing result of the threshold filtering module and separates the original signal into noise and a noise-reduced signal through independent component analysis.
The invention provides a method and a system for terahertz time-domain signal noise reduction and image reconstruction based on empirical mode decomposition, threshold filtering and independent component analysis. The mode containing the noise component obtained by empirical mode decomposition is screened in a threshold filtering mode, useful signal components are removed, a purer virtual noise channel is obtained, and then noise and effective signals in the noise-containing signals are decomposed through independent component analysis, so that the problem of mode aliasing existing in the empirical mode decomposition is well solved, and the signal-to-noise ratio of the processed signals is greatly improved. The signal noise reduction method can be further utilized to carry out two-dimensional imaging on the measured sample, and a terahertz image of the measured sample can be reconstructed according to the maximum value, the minimum value or the amplitude value at a specific time point of the terahertz signal after noise reduction at each point of the sample, or based on the amplitude value at a certain frequency in a frequency spectrum after Fourier transform of the noise reduction signal, or based on the size of the optical parameter of the sample at a certain frequency extracted by the noise reduction signal, so that the signal-to-noise ratio of the terahertz imaging is remarkably improved, and the contrast of different samples is enhanced.
Drawings
FIG. 1 is a flow chart of a terahertz time-domain signal noise reduction process for a sample containing noise according to the invention;
FIG. 2 is a flow chart of terahertz imaging of a sample according to the present invention;
FIG. 3 is a graph comparing a noisy signal at a noise level of 8dB with a noise-reduced signal processed by a different method;
FIG. 4 is a comparison graph of the SNR and cross-correlation results for noise-reduced signals processed by different methods at different noise levels;
FIG. 5 is a schematic diagram of terahertz reflection imaging results of a layered fiber reinforced polymer composite material (with defects inside) provided by an embodiment of the invention; fig. 5(a) is a single-point original signal of a sample, fig. 5(b) is a corresponding signal obtained after noise reduction by using the method of the present invention, fig. 5(c) is a result obtained by performing terahertz imaging on a defect layer of a layered fiber-reinforced polymer composite material (containing a defect inside) based on an amplitude value at a specific time point of a reflected terahertz time-domain signal, and fig. 5(d) is a result obtained by performing terahertz imaging on the defect layer of the layered fiber-reinforced polymer composite material (containing a defect inside) based on the amplitude value at the specific time point of the signal obtained after noise reduction by using the method of the present invention.
Detailed Description
In order to make the objects, technical solutions and advantages of the present invention more apparent, the present invention is described in further detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative of the invention and are not intended to limit the invention. In addition, the technical features involved in the embodiments of the present invention described below may be combined with each other as long as they do not conflict with each other.
The invention provides a method for denoising terahertz time-domain signals based on empirical mode decomposition, threshold filtering and independent component analysis, which comprises the following steps:
1) acquiring a terahertz time-domain signal containing noise of a detected sample;
2) carrying out empirical mode decomposition on the terahertz time-domain signal containing noise to obtain a plurality of eigenmode functions;
3) selecting a noise-dominated eigenmode function by using the correlation coefficient;
4) filtering the selected noise-dominant eigenmode function by using a threshold value, and filtering a signal part to obtain a modal component of the noise part;
5) constructing a virtual noise channel by using the modal components of the noise part obtained by filtering;
6) separating the original signal into noise and a noise-reduced signal by independent component analysis by using a virtual noise channel;
the terahertz time-domain signal in the step 1) can be a terahertz time-domain signal of a detected sample obtained based on measurement of a transmission-type terahertz time-domain spectroscopy system or based on measurement of a reflection-type terahertz time-domain spectroscopy system.
The empirical mode decomposition in the step 2) may be collective empirical mode decomposition, complementary collective empirical mode decomposition, or variational modal decomposition.
The threshold filtering in step 4) may be various threshold filtering methods such as soft threshold filtering, hard threshold filtering, or adaptive threshold filtering.
The independent component analysis in step 6) may be fast independent component analysis or fourier independent component analysis.
In addition, the invention also provides a method for reconstructing the terahertz image based on empirical mode decomposition, threshold filtering and independent component analysis, which comprises the following steps:
1) performing two-dimensional imaging on a tested sample to obtain terahertz time-domain signals with noise at each point on the tested sample;
2) obtaining noise-reduced signals of each point on the tested sample based on the signal noise reduction method;
3) according to different application requirements, selecting the maximum value, the minimum value or the intensity at a specific time point of the noise-reduced signal, or selecting an amplitude value at a certain frequency in a frequency spectrum obtained after Fourier transform of the noise-reduced signal, or selecting the size of an optical parameter extracted based on the noise-reduced signal at a certain frequency, and reconstructing the terahertz image of the detected sample; specifically selecting an index which can be set according to the comparison requirement of a user; the specifically selected time point and frequency point can also be adjusted according to the comparison requirement of the user, and the method is not limited to this;
and 2, performing two-dimensional imaging on the detected sample in the step 1, performing terahertz point-by-point scanning imaging or terahertz focal plane imaging, and combining transmission imaging or reflection imaging.
Wherein, the sample in the step 1 is in various sample forms capable of carrying out terahertz imaging.
In addition, the invention also provides a system for reconstructing the terahertz image based on empirical mode decomposition, threshold filtering and independent component analysis, which comprises
1) The imaging module is used for carrying out two-dimensional imaging on a tested sample to obtain terahertz time-domain signals with noise at each point on the tested sample;
2) the noise reduction processing module is used for carrying out noise reduction processing on the terahertz signals containing the noise of each point to obtain the signals of each point on the detected sample after noise reduction;
3) the image reconstruction module is used for selecting the maximum value and the minimum value of the noise-reduced signal or the intensity at a specific time point according to different application requirements, or selecting an amplitude value at a certain frequency in a frequency spectrum after Fourier transform is carried out on the noise-reduced signal, or selecting the size of an optical parameter at a certain frequency extracted based on the noise-reduced signal, and reconstructing the terahertz image of the detected sample;
4) and the result comparison module is used for displaying and comparing the reconstructed terahertz images of different tested samples.
Wherein, the noise reduction processing module further comprises:
an empirical mode decomposition module: the terahertz time-domain signal processing method comprises the following steps of performing empirical mode decomposition on a terahertz time-domain signal containing noise to obtain a plurality of eigenmode functions;
a threshold filtering module: receiving a processing result of the empirical mode decomposition module, selecting an eigenmode function with leading noise by using a correlation coefficient, filtering the selected eigenmode function with leading noise by using a threshold value, filtering a signal part to obtain a modal component of the noise part, and constructing a virtual noise channel by using the modal component of the noise part obtained by filtering;
an independent component analysis module: and receiving a processing result of the threshold filtering module, and separating the original signal into noise and a noise-reduced signal through independent component analysis.
The contents of the above embodiments will be described with reference to a preferred embodiment.
The embodiment of the invention provides a method for denoising terahertz time-domain signals based on empirical mode decomposition, soft threshold filtering and rapid independent component analysis, which comprises the following steps:
step 1: acquiring a terahertz time-domain signal of a sample containing noise;
step 2: and carrying out decomposition calculation on the obtained noise-containing signals by adopting empirical mode decomposition to obtain a plurality of eigenmode functions. The calculation method is as follows:
(1) finding all local maximum and minimum points in the noisy signal x (t);
(2) respectively obtaining the upper envelope line x in (1)max(t) and the lower envelope xmin(t)；
(3) Mean of upper and lower envelope:
m1(t)＝(xmax(t)+xmin(t))/2；
(4) obtaining a new data sequence h with x (t) and the mean value removed1(t)：
h1(t)＝x(t)-m1(t)；
(5) Inspection h1(t) whether or not the condition of the eigenmode function is satisfied, in general, h is required to be satisfied1(t) repeating the above process until the definition requirement of the eigenmode function is satisfied (1. within the whole time range, the number of local extremum points and zero crossing points must be equal or at most one difference; 2. at any time, the average value of the upper envelope line and the lower envelope line must be zero), finally obtaining the first eigenmode function c1(t)；
(6) Let r be1(t)＝x(t)-c1(t) with r1(t) is a time domain signal, and other eigenmode functions can be obtained by repeating the processing steps: c. C2(t)，c3(t)，...，cn(t) and remainder rn(t) of (d). To this end, the result of empirical mode decomposition of noisy signals can be expressed as:
and step 3: selecting noise dominant eigenmode function by using correlation coefficient, and selecting high-frequency eigenmode function c with correlation coefficient less than 0.2i(t) and the next eigenmode function ci+1(t) the eigenmode function which is regarded as the leading noise, and the calculation formula of the correlation coefficient of each eigenmode function and the original signal is as follows:
and 4, step 4: filtering the selected noise-dominant eigenmode function by adopting a soft threshold value, filtering a signal part, and obtaining a modal component y of the noise parti(t)，
The calculation method specifically comprises the following steps:
yi(t)＝ci(t)-Zi(t)；
wherein the content of the first and second substances,
σ＝median(|ci(t)-median(ci(t)) |)/0.6745, with N being the signal length.
And 5: constructing a virtual noise channel by using the modal component of the noise part obtained by the soft threshold;
step 6: and (3) distinguishing noise and useful signals by adopting rapid independent component analysis to obtain a noise reduction result.
In order to verify the noise reduction performance of the terahertz time-domain signal, the terahertz time-domain signal with different noise levels is formed by adding different white Gaussian noises to the pure terahertz time-domain signal. The existing noise reduction method (low-pass filter, empirical mode decomposition and independent component analysis combination) and the signal noise reduction method provided by the invention are respectively utilized for processing, when the noise level is 8dB, a pure signal, a noise-containing signal and a noise-reduced signal processed by different methods are shown in figure 3, and under different noise levels, the signal-to-noise ratio of the noise-reduced signal processed by different methods and the cross-correlation coefficient comparison result with the pure signal are shown in figure 4.
In order to verify the effect of the terahertz image reconstruction method provided by the invention, in a specific embodiment, a mode of combining two-dimensional point-by-point scanning with reflection imaging is adopted to image a laminated fiber reinforced polymer composite material (the interior of which contains defects), and the steps are as follows: (1) carrying out two-dimensional imaging on the layered fiber reinforced polymer composite material (with defects inside) to obtain noise-containing terahertz time-domain signals of each point on a detected sample; (2) obtaining noise-reduced signals of each point of the detected laminated fiber reinforced polymer composite material (the interior of which contains defects) by using the signal noise reduction method; (3) and respectively selecting the amplitude values of the original signal and the denoised signal under 22.03ps, and reconstructing a terahertz image of the detected layered fiber reinforced polymer composite material (with the defect inside) defect layer. In this imaging application, the time point 22.03ps was chosen for the following reasons: according to the layered distribution of the layered fiber reinforced polymer composite material (containing defects inside), the refractive index of the material and the depth position of the defect layer in the sample, the flight time required by the reflection of the terahertz pulse at the interface of the defect layer is calculated, and the corresponding time-domain signal time point is 22.03 ps.
The results before and after the noise reduction processing of the sample single-point terahertz signal and the corresponding terahertz imaging result are shown in fig. 5. By adopting the original terahertz time-domain signal, the defect position is invisible. By adopting the novel imaging method provided by the invention, the defect position is clearly visible (shown by a black dotted line frame), and the contrast between different samples is obviously enhanced. The signal-to-noise ratio of the terahertz image of the imaging method provided by the invention in the target defect area is 14.15dB, and the signal-to-noise ratio of the terahertz image of the traditional imaging method in the target defect area is 9.25dB, so that the signal-to-noise ratio of the terahertz image is obviously improved, the contrast among different samples is obviously enhanced, and beneficial technical effects are achieved. The invention provides a new imaging method for terahertz imaging and promotes the development of the field.
In summary, the invention provides a method and a system for terahertz signal noise reduction and image reconstruction based on empirical mode decomposition, threshold filtering and independent component analysis. The noise-containing mode obtained by empirical mode decomposition is screened by a threshold filtering method to obtain a virtual noise channel, and then noise and effective signals in the noise-containing signals are distinguished by independent component analysis, so that the problem of mode aliasing in the empirical mode decomposition is well solved, and compared with the existing method, the signal-to-noise ratio of the noise-reduced signals is greatly improved. When the method is further used for imaging, compared with the existing imaging method, the imaging signal-to-noise ratio and the contrast are obviously improved.
It will be understood by those skilled in the art that the foregoing is only a preferred embodiment of the present invention, and is not intended to limit the invention, and that any modification, equivalent replacement, or improvement made within the spirit and principle of the present invention should be included in the scope of the present invention.
Claims (10)
1. A terahertz time-domain signal noise reduction method is characterized by comprising the following steps:
s1, carrying out empirical mode decomposition on terahertz time-domain signals containing noise to obtain a plurality of eigenmode functions;
s2, selecting an eigenmode function with leading noise by using a correlation coefficient;
s3, filtering the eigen mode function of the noise leading by using a threshold value, and then filtering a signal part to obtain a modal component of the noise part;
s4, constructing a virtual noise channel by using the modal components of the noise part;
and S5, separating the original signal into noise and a noise-reduced signal through independent component analysis by using the virtual noise channel.
2. The terahertz time-domain signal noise reduction method of claim 1, wherein the empirical mode decomposition is a classical empirical mode decomposition, an ensemble empirical mode decomposition, a complementary ensemble empirical mode decomposition, or a variational mode decomposition.
3. The terahertz time-domain signal noise reduction method of claim 1, wherein in step S1, the empirical mode decomposition result of the terahertz time-domain signal x (t) is represented as
4. The terahertz time-domain signal noise reduction method according to claim 3, wherein the step S2 specifically includes the steps of:
selecting noise dominant eigenmode function by using correlation coefficient, and selecting high-frequency eigenmode function c with correlation coefficient less than 0.2i(t) and the next eigenmode function ci+1(t) eigenmode functions considered to be noise dominated.
5. The terahertz time-domain signal noise reduction method of claim 1, wherein the threshold filtering of step S3 is soft threshold filtering, hard threshold filtering, or adaptive threshold filtering.
6. The method for reducing the noise of the terahertz time-domain signal according to claim 1, wherein the terahertz time-domain signal is the terahertz time-domain signal of the sample to be measured, which is obtained based on a transmission type terahertz time-domain spectroscopy system measurement or based on a reflection type terahertz time-domain spectroscopy system measurement.
7. The terahertz time-domain signal noise reduction method of claim 1, wherein the independent component analysis of step S5 is a fast independent component analysis or a fourier independent component analysis.
8. A terahertz image reconstruction method is characterized by comprising the following steps:
performing two-dimensional imaging on a tested sample to obtain terahertz time-domain signals with noise at each point on the tested sample;
obtaining noise-reduced signals of each point on the sample to be detected by adopting the terahertz time-domain signal noise reduction method of any one of claims 1 to 7;
according to different application requirements, selecting the maximum value, the minimum value or the intensity at a specific time point of the noise-reduced signal, or selecting an amplitude value at a certain frequency in a frequency spectrum obtained after Fourier transform is performed on the noise-reduced signal, or selecting the size of an optical parameter extracted based on the noise-reduced signal at a certain frequency, and reconstructing the terahertz image of the detected sample.
9. The terahertz image reconstruction method of claim 8, wherein the two-dimensional imaging of the sample under test comprises: terahertz point-by-point scanning imaging or terahertz focal plane imaging is adopted, and transmission imaging or reflection imaging is combined.
10. A terahertz image reconstruction system, comprising:
the imaging module is used for carrying out two-dimensional imaging on a tested sample to obtain terahertz time-domain signals with noise at each point on the tested sample;
the noise reduction processing module is used for carrying out noise reduction processing on the terahertz signals containing the noise of each point to obtain the signals of each point on the detected sample after noise reduction;
the image reconstruction module is used for selecting the maximum value and the minimum value of the noise-reduced signal or the intensity at a specific time point according to different application requirements, or selecting an amplitude value at a certain frequency in a frequency spectrum after Fourier transform is carried out on the noise-reduced signal, or selecting the size of an optical parameter at a certain frequency extracted based on the noise-reduced signal, and reconstructing the terahertz image of the detected sample;
the result comparison module is used for displaying and comparing the reconstructed terahertz images of different tested samples;
wherein the noise reduction processing module further comprises:
the empirical mode decomposition module is used for performing empirical mode decomposition on the terahertz time-domain signal containing the noise to obtain a plurality of eigenmode functions;
the threshold filtering module is used for receiving the processing result of the empirical mode decomposition module, selecting an eigenmode function with leading noise by using a correlation coefficient, filtering the selected eigenmode function with leading noise by using a threshold, filtering a signal part to obtain a modal component of the noise part, and constructing a virtual noise channel by using the modal component of the noise part obtained by filtering;
and the independent component analysis module receives the processing result of the threshold filtering module and separates the original signal into noise and a noise-reduced signal through independent component analysis.
Priority Applications (2)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202111538693.6A CN114386456A (en) | 2021-12-15 | 2021-12-15 | Terahertz time-domain signal noise reduction method, image reconstruction method and system |
| PCT/CN2022/138212 WO2023109717A1 (en) | 2021-12-15 | 2022-12-09 | Terahertz time domain signal noise reduction method, and terahertz image reconstruction method and system |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202111538693.6A CN114386456A (en) | 2021-12-15 | 2021-12-15 | Terahertz time-domain signal noise reduction method, image reconstruction method and system |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN114386456A true CN114386456A (en) | 2022-04-22 |
Family
ID=81198521
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202111538693.6A Pending CN114386456A (en) | 2021-12-15 | 2021-12-15 | Terahertz time-domain signal noise reduction method, image reconstruction method and system |
Country Status (2)
| Country | Link |
|---|---|
| CN (1) | CN114386456A (en) |
| WO (1) | WO2023109717A1 (en) |
Cited By (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN114936375A (en) * | 2022-06-02 | 2022-08-23 | 北京理工大学 | Asymmetric integrated optical encryption system based on two-dimensional empirical mode decomposition |
| WO2023109717A1 (en) * | 2021-12-15 | 2023-06-22 | 深圳先进技术研究院 | Terahertz time domain signal noise reduction method, and terahertz image reconstruction method and system |
| CN114936375B (en) * | 2022-06-02 | 2024-06-04 | 北京理工大学 | Asymmetric integrated optical encryption system based on two-dimensional empirical mode decomposition |
Families Citing this family (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN116776168B (en) * | 2023-08-22 | 2023-11-21 | 惠州帝恩科技有限公司 | Intelligent analysis method and system for production data of reagent tubes |
| CN116843596B (en) * | 2023-08-28 | 2023-11-14 | 浙江大学杭州国际科创中心 | Method, system and device for adaptive fusion of multi-mode images of X-ray grating |
Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN108618773A (en) * | 2017-03-15 | 2018-10-09 | 深圳市理邦精密仪器股份有限公司 | A kind of Denoising of ECG Signal, device and a kind of ecg signal acquiring equipment |
| CN110553998A (en) * | 2019-07-31 | 2019-12-10 | 西安交通大学 | nondestructive testing method for blade test piece of aero-engine based on terahertz technology |
Family Cites Families (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN107464226B (en) * | 2017-07-31 | 2019-10-15 | 东南大学 | A kind of image de-noising method based on improvement two-dimensional empirical mode decomposition algorithm |
| CN110940409A (en) * | 2019-12-02 | 2020-03-31 | 天津市计量监督检测科学研究院 | Ultrasonic signal measurement method based on ICEEMDAN and ICA combined denoising |
| CN114386456A (en) * | 2021-12-15 | 2022-04-22 | 深圳先进技术研究院 | Terahertz time-domain signal noise reduction method, image reconstruction method and system |
-
2021
- 2021-12-15 CN CN202111538693.6A patent/CN114386456A/en active Pending
-
2022
- 2022-12-09 WO PCT/CN2022/138212 patent/WO2023109717A1/en unknown
Patent Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN108618773A (en) * | 2017-03-15 | 2018-10-09 | 深圳市理邦精密仪器股份有限公司 | A kind of Denoising of ECG Signal, device and a kind of ecg signal acquiring equipment |
| CN110553998A (en) * | 2019-07-31 | 2019-12-10 | 西安交通大学 | nondestructive testing method for blade test piece of aero-engine based on terahertz technology |
Non-Patent Citations (3)
| Title |
|---|
| 张婷 等: ""改进的EEMD算法在时域航空电磁信号降噪中的研究"", 《信号处理》, vol. 32, no. 07, 25 July 2016 (2016-07-25), pages 771 - 778 * |
| 张婷;李双田;: "改进的EEMD算法在时域航空电磁信号降噪中的研究", 信号处理, no. 07, 25 July 2016 (2016-07-25), pages 771 - 778 * |
| 李杰: ""基于经验模态分解与独立成分分析的航空瞬变电磁去噪方法研究"", 中国优秀硕士学位论文全文数据库基础科学辑, 15 June 2021 (2021-06-15), pages 011 - 309 * |
Cited By (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2023109717A1 (en) * | 2021-12-15 | 2023-06-22 | 深圳先进技术研究院 | Terahertz time domain signal noise reduction method, and terahertz image reconstruction method and system |
| CN114936375A (en) * | 2022-06-02 | 2022-08-23 | 北京理工大学 | Asymmetric integrated optical encryption system based on two-dimensional empirical mode decomposition |
| CN114936375B (en) * | 2022-06-02 | 2024-06-04 | 北京理工大学 | Asymmetric integrated optical encryption system based on two-dimensional empirical mode decomposition |
Also Published As
| Publication number | Publication date |
|---|---|
| WO2023109717A1 (en) | 2023-06-22 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN114386456A (en) | Terahertz time-domain signal noise reduction method, image reconstruction method and system | |
| CA2762848C (en) | System and method for detecting poor quality in 3d reconstructions | |
| Yugander et al. | MR image enhancement using adaptive weighted mean filtering and homomorphic filtering | |
| CN109697476B (en) | X-ray photon counting detector consistency calibration method based on deep learning | |
| CN105628675B (en) | A kind of removing method of the Raman fluorescence interference of power sensitive substance | |
| CN117056642A (en) | Microwave data processing method and system | |
| Gemechu et al. | Random noise attenuation using an improved anisotropic total variation regularization | |
| WO2023109713A1 (en) | Terahertz imaging method, system and device based on empirical wavelet transform | |
| CN116399850B (en) | Spectrum detection and identification system for optical signal processing and detection method thereof | |
| Subbuthai et al. | Reduction of types of noises in dental images | |
| Li et al. | Research on image denoising based on median filter | |
| Zhai et al. | Terahertz Nondestructive Stratigraphic Analysis of Complex Layered Structures: Reconstruction Techniques | |
| Shen et al. | Deep learning-based wavelet threshold function optimization on noise reduction in ultrasound images | |
| CN115307715A (en) | Improved wavelet denoising method based on sagnac optical fiber acoustic sensing system | |
| CN114444548A (en) | Wavelet sound denoising method and system based on robustness principal component analysis | |
| CN114970639A (en) | Gas flicker noise elimination method for open space external environment | |
| Quan | Quality evaluation method of agricultural product packaging image based on structural similarity and MTF | |
| KR20100121273A (en) | Apparatus and method for processing x-ray image | |
| Chinnathambi et al. | Despeckling of ultrasound images using directionally decimated wavelet packets with adaptive clustering | |
| Lee et al. | A hybrid approach of wavelet-based total variation and Wiener filter to denoise adventitious lung sound signal for an accurate assessment | |
| Jiao | An image denosing algorithm for strong noise image based on structural similarity of contourlet domain via grey relational analysis | |
| Wang et al. | A novel method to process the THz time-domain spectrum and software to make data visualization | |
| Li et al. | A Novel Method for the Enhancement of Composite Materials’ Terahertz Image Using Unsharp Masking and Guided Filtering Technology | |
| Feng et al. | Binarization method for on-line ferrograph image based on uniform curvelet transformation | |
| Hu et al. | Ultrasound image preprocessing method for deep-learning-based fatty liver diagnosis |
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 |