Terahertz time-domain spectroscopy-based wheat stripe rust uredospore optical parameter extraction method

Abstract

The invention discloses a terahertz time-domain spectroscopy-based wheat stripe rust uredospore optical parameter extraction method. The method comprises the following steps: mixing wheat stripe rust uredospore powder and polyethylene powder, and tabletting; a transmission module of the terahertz time-domain spectrometer is used for collecting time-domain spectral data of the mixed tablets in a nitrogen environment; carrying out filtering processing on the collected time-domain spectral data to eliminate high-frequency noise; main waves of the time domain signals are reserved, the echo position is estimated, sampling points are randomly generated through normal distribution to dynamically correct the echo signals, and time domain interference is reduced; and performing fast Fourier transform on the corrected time domain signal, and calculating the refractive index, the absorption coefficient and the dielectric constant of the wheat stripe rust uredospores through frequency domain amplitude and phase information on the basis of an improved Dorne-Duvillaret model. Through sample preparation optimization, echo correction and model improvement, the measurement precision can be remarkably improved.

Description

Technical Field

[0001] The invention belongs to the technical field of biomaterial optical parameter measurement, and in particular relates to a method for extracting optical parameters of wheat stripe rust uredospores based on terahertz time-domain spectroscopy. Background Art

[0002] Traditional optical parameter measurement methods such as ellipsometry and Kramers-Kronig transform have significant limitations in biological sample detection. Ellipsometry requires a highly flat sample surface, while biological samples such as wheat stripe rust uredia have irregular shapes and are fragile. The surface roughness causes optical path distortion and large measurement errors. In addition, the theories used in the above measurement methods all assume that the material is a homogeneous medium, but biological samples are porous, layered, or contain complex components such as proteins and water, resulting in significant deviations in optical parameter extraction and the inability to monitor the dynamic changes of biological samples in real time. Although the Kramers-Kronig transform can invert optical parameters through phase information, it has strict requirements on data integrity and requires full-band phase information. The signal-to-noise ratio of biological samples in the terahertz high-frequency band is low, and data extrapolation is prone to introduce systematic errors. This is especially true for active uredia with a high water content, which strongly absorbs terahertz waves, making it difficult to accurately obtain phase information and invalidating parameter calculations. While existing terahertz time-domain spectroscopy offers the advantage of being non-destructive, it still faces numerous bottlenecks in its application to biological samples. For example, directly using time-domain signals to calculate optical parameters fails to eliminate echo interference caused by multiple reflections within the sample; using pure biological powder to compress the sample into pellets results in sample fragility and severe scattering, leading to poor measurement repeatability; and the effective frequency band is not optimized for biological samples, so full-band calculations introduce high-frequency noise, resulting in significant errors in optical parameters. These limitations limit the practical application of terahertz technology in biomaterial testing, necessitating a calculation method that balances sample preparation stability with signal processing accuracy. Summary of the Invention

[0003] In order to overcome the shortcomings of the above-mentioned prior art, the purpose of the present invention is to provide a method for extracting optical parameters of wheat stripe rust uredospores based on terahertz time-domain spectroscopy, which significantly improves the measurement accuracy through sample preparation optimization, echo correction and model improvement.

[0004] In order to achieve the above object, the technical solution adopted by the present invention is:

[0005] A method for extracting optical parameters of wheat stripe rust uredospores based on terahertz time-domain spectroscopy comprises the following steps:

[0006] Step 1, mixing wheat stripe rust uredia powder and polyethylene powder and pressing the mixture into tablets;

[0007] Step 2, using the transmission module of the terahertz time-domain spectrometer to collect time-domain spectrum data of the mixed pressed tablet in a nitrogen environment;

[0008] Step 3: Filter the collected time-domain spectral data to eliminate high-frequency noise;

[0009] Step 4: Use the time domain window method to retain the main wave in the time domain signal, use the normal distribution to randomly generate sampling points, estimate the echo position and replace the original echo in the time domain signal to achieve echo signal correction and reduce time domain interference;

[0010] Step 5: Perform fast Fourier transform on the corrected time domain signal, and calculate the refractive index, absorption coefficient and dielectric constant of wheat stripe rust uredospores through frequency domain amplitude and phase information based on the improved Dorney-Duvillaret model.

[0011] The present invention can significantly improve measurement accuracy through the sample preparation optimization in step 1, the echo correction in step 4, and the model improvement in step 5, providing a theoretical basis for obtaining optical parameters of wheat stripe rust urediospores and biological samples.

[0012] The wheat stripe rust spores used in the present invention are all cultured and collected in a low-temperature greenhouse. The collected samples are stored in a drying dish containing silica gel and preserved under appropriate conditions.

[0013] In one embodiment, step 1 is to mix wheat stripe rust uredia powder with polyethylene powder in a mass ratio of 1:9, and use a tablet press to press at 2 tons of pressure for 5 minutes to form a stable mixed tablet with a diameter of 12 mm and a thickness of 0.5 to 2 mm as a biological sample. This process can significantly reduce the scattering loss of the biological sample, while enhancing the mechanical strength of the tablet, avoiding the fragility problem of pure biological sample tablets. Further preferably, the tablet of the present invention adopts a staged pressure gradient, i.e., 0.5 tons pre-pressing for 1 minute → 2 tons main pressing for 3 minutes → 1 ton final pressing for 1 minute, gradually removing the air between the powders, reducing the internal porosity, and facilitating subsequent accurate calculations.

[0014] In one embodiment, the operating frequency range of the terahertz time-domain spectrometer is 0.05 to 5 THz, the dynamic range is ≥90 dB, and the spectral resolution is ≤0.85 GHz. Before collecting time-domain spectral data, nitrogen is introduced until the water vapor absorption peak disappears, thereby eliminating the strong absorption interference of moisture on terahertz waves. When collecting time-domain spectral data, the pressed sheet is placed in the transmission light path, and the sample is rotated in four different directions and the measurement is repeated to reduce the local scattering error caused by uneven mixing.

[0015] In one embodiment, in step 3, a Gaussian window function is applied to the collected time domain spectrum, and the collected time domain spectrum is filtered using the Gaussian window to filter out high-frequency noise and improve the data signal-to-noise ratio.

[0016] In one embodiment, the correction echo signal is implemented by:

[0017] First, the main wave signal in the time domain is retained and the echo area is truncated; then, using statistical signals, a normally distributed random signal is generated in the echo area to correct the original echo signal, thereby achieving the purpose of dynamic correction.

[0018] By the formula Δτ 理论 =2nd / cCalculate the theoretical time difference Δτ between the main wave and the echo 理论 , where n is the refractive index of the mixed sheet, d is the thickness of the sheet, and c is the speed of light. The echo position is adjusted according to the actual measurement value. This is mainly done by dynamically calibrating the deviation between the theoretical model and the actual signal to accurately locate and correct the echo interference.

[0019] Specifically, first detect the main wave peak position t of the time domain signal 主波 , calculate the theoretical time difference Δτ between the main wave and the echo based on the formula 理论 , and then directly measure the actual time difference Δτ between the main wave peak and the echo peak from the time domain signal obtained by the spectrometer 测量 , by calculating Δτ 理论 and Δτ 测量 Deviation δτ=Δτ 测量 -Δτ 理论 , for subsequent echo positioning, adjust the echo position to the interval range: t 回波 ∈[t 主波 ±δτ]. Subsequently, the amplitude distribution statistics of the original signal within the echo interval are analyzed to dynamically adjust the normal distribution of the replacement signal. Specifically, the statistical signal is used to generate a normally distributed random signal within the adjusted echo position interval, replacing the original echo in the time domain signal. This dynamically corrects the original echo signal. Ultimately, the corrected echo signal better matches the actual noise characteristics, reducing model errors caused by manually preset parameters.

[0020] In one embodiment, in step 4, the theoretical time difference between the main wave and the echo is first calculated using a formula, and then the echo position interval is estimated. 800 sampling points are randomly generated using a normal distribution, corresponding to a 5ps time window, and the signal in the original echo interval is replaced with a mean value of -0.012 and a standard deviation of 0.001. The echo signal is dynamically corrected and combined into a complete time domain waveform.

[0021] Through pre-processing and echo correction, the present invention can effectively eliminate time domain interference caused by factors such as multiple reflections.

[0022] In one embodiment, in step 5, after the fast Fourier transform, effective frequency band data is extracted for calculation.

[0023] The absorption rate of wheat stripe rust uredospores in the terahertz band is mainly affected by their molecular vibration, lattice vibration, and residual moisture. Its core response mechanisms include:

[0024] The low-frequency band (0.1-1 THz) corresponds to macromolecular skeleton vibrations (such as protein-polysaccharide complexes) and hydrogen bond network relaxation, and is a characteristic absorption region for biological samples. The mid-frequency band (1-2 THz) reflects weak intermolecular interactions (such as van der Waals forces and dipole coupling) and is closely related to the activity state of spores (such as activity / inactivation). The high-frequency band (>2.5 THz): is mainly dominated by residual water (OH stretching vibration harmonics) and instrument noise (signal-to-noise ratio <50:1), and data reliability is significantly reduced. Especially for compressed samples, the influence of instrument noise is very large in the high-frequency band. Therefore, the present invention selects data from the 0.1-2.5 THz effective frequency band for subsequent calculations to reduce calculation errors caused by machine noise.

[0025] The specific upper and lower limits are selected as follows: Since the thickness of the pressed sheet is d = 2 mm and the maximum refractive index of polyethylene is n = 1.56, the minimum frequency for separating the main wave and the echo of the time domain signal can be calculated by the following formula.

[0026]

[0027] When the frequency is lower than 0.1 THz, the main wave and the echo are easily mixed in the time domain and are difficult to separate effectively.

[0028] The upper limit is mainly determined by the instrument's pulse half-width and signal-to-noise ratio.

[0029] In one embodiment, in step 5, the refractive index n of the sample tab (ω), absorption coefficient α tab (ω) and dielectric constant ε tab Solved by the calculation formula of the improved Dorney-Duvillaret model:

[0030]

[0031] Where c is the speed of light, d is the thickness of the sheet, φ(ω) is the phase difference between the sample signal and the reference signal, ω is the angular frequency, ρ(ω) is the amplitude ratio of the sample signal to the reference signal, and i is the imaginary unit.

[0032] In one embodiment, the wheat stripe rust uredia are active or inactive stripe rust uredia, wherein the refractive index of the active uredia is in the range of 1.40 to 1.42, and the real part of the dielectric constant is 1.95; the refractive index of the inactive uredia is in the range of 1.44 to 1.45, and the real part of the dielectric constant is 2.1; the corrected refractive index error is less than 0.2%, and the method and steps are also applicable to other biological spores.

[0033] Compared to existing technologies, this invention utilizes a terahertz time-domain spectrometer to filter and correct the collected time-domain spectrum of a biological sample (wheat stripe rust spores) mixed with polyethylene powder using a Gaussian window and echo correction. Based on the improved Dorney-Duvillaret model, optical parameters such as the refractive index, absorption coefficient, and dielectric constant are calculated from the spectral amplitude and phase information, enabling accurate extraction of the optical parameters of the pathogen spores. The proposed extraction method can also be applied to other spores and biological samples, providing a highly accurate method for determining optical parameters and laying a theoretical foundation for the subsequent development of precise detection devices for biological samples. BRIEF DESCRIPTION OF THE DRAWINGS

[0034] Figure 1 Schematic diagram of spectral parameter extraction of the present invention.

[0035] Figure 2 These are the THz absorption spectrum curves of uredia samples at different thicknesses.

[0036] Figure 3 is the time domain spectrum of the corrected polyethylene signal.

[0037] Figure 4 Comparison of the exact values of optical parameters before and after correction. (a) is the refractive index, (b) is the absorption coefficient, (c) is the real part of the dielectric constant, and (d) is the imaginary part of the dielectric constant. DETAILED DESCRIPTION

[0038] The embodiments of the present invention are described in detail below with reference to the accompanying drawings and examples.

[0039] The present invention is a method for extracting optical parameters of wheat stripe rust uredospores based on terahertz time-domain spectroscopy technology. The terahertz time-domain spectrometer is used to obtain biological sample data and then solve its optical parameters. The overall extraction method is as follows: Figure 1 By determining the biological sample, building the experimental collection platform, acquiring data, processing and calculating, and finally determining that the error meets the requirements, the optical parameters of stripe rust uredia can be accurately extracted.

[0040] The present invention specifically comprises the following steps:

[0041] Step 1: Collect wheat stripe rust uredospore samples, wherein the inactivated stripe rust spores are obtained by inactivating active stripe rust spores at high temperature. The samples are placed in a refrigerator for refrigeration and placed at room temperature of 20°C for 2 hours two hours before use to balance the temperature and humidity and eliminate the effects of temperature and humidity on the room temperature. After the spores are completely in a room temperature state, they can be used for measuring optical parameters.

[0042] First, the collected wheat stripe rust summer spore powder is prepared into discs using a grinder, a balance and a tablet press for later use. Specifically, the present invention takes wheat stripe rust summer spore powder (active or inactivated) and high-density polyethylene powder and accurately weighs and mixes them in a mass ratio of 1:9, pours the mixed powder into a tableting mold with a diameter of 12 mm, uses a tableting machine to apply a pressure of 2t and maintains it for 5 minutes, and cleans the mold surface with alcohol after each tableting to eliminate the interference of residual powder on the spectrum. During tableting, it was found that the detection accuracy would be affected by the thickness. If the tablet is too thick, it may cause spectral attenuation. If the tablet is too thin, it is easy to generate echoes and the tablet is fragile. Therefore, in order to explore the optimal thickness of the tablet in this experiment, sample tablets of different thicknesses are set to obtain their THz absorption coefficient spectra. The thickness range of the mixed tablet can be selected from 0.5 to 2 mm, such as Figure 2 As shown in the figure, the optimal tablet thickness is 0.7 mm. A staged pressure gradient can be used in the tableting process, such as 0.5 ton pre-pressing for 1 minute → 2 ton main pressing for 3 minutes → 1 ton final pressing for 1 minute, to gradually remove air between the powders and reduce internal porosity, facilitating subsequent accurate calculations.

[0043] Step 2: Use a terahertz time-domain spectrometer to collect the transmission spectrum data of the sample.

[0044] The present invention uses a CCT-1800 terahertz time-domain spectrometer transmission module with an operating frequency range of 0.05 to 5 THz, a dynamic range of 90 dB or higher, and a spectral resolution of 0.85 GHz or lower. The present invention sets the following operating parameters: a scanning range of 1200 ps and a spectral resolution of 0.85 GHz. High-purity nitrogen is introduced into the sample chamber before acquisition, and the reference spectrum is monitored in real time until the water vapor absorption peak disappears. The compressed sample is placed vertically in the terahertz optical path. During acquisition, the sample is rotated in four directions: 0°, 90°, 180°, and 270°. The time-domain signals are collected and recorded separately to reduce local scattering errors caused by uneven mixing.

[0045] Step 3: Preprocess and correct the obtained transmission spectrum data.

[0046] Specifically, a Gaussian window function operation is performed on the original time domain signal to eliminate high-frequency noise through filtering, improve the signal-to-noise ratio, and make subsequent calculations more accurate.

[0047] By the formula Δτ 理论 =2nd / cCalculate the theoretical time difference Δτ between the main wave and the echo 理论 ,like Figure 3 As shown. Where n is the refractive index of the mixed tablet, d is the thickness of the tablet, and c is the speed of light. 理论 =18.7ps, inactivated urediospore Δτ 理论=19.3ps. The primary echo of the terahertz signal passing through the polyethylene sample has a small amplitude and low energy. To reduce the error caused by the echo in optical parameter extraction, the main wave of the time domain signal is retained and the part that generates the echo is reduced. First, the echo position is estimated. Second, 800 sampling points (about 5ps) are randomly generated using a normal distribution. The echo signal is corrected with an average value of -0.012 and a standard deviation of 0.001. Finally, the effective time domain window (main wave center ±10ps) is intercepted to eliminate the influence of edge signal jitter on the frequency domain transformation. The range of the echo position adjustment is: t 回波 ∈[t 主波 ±δτ].

[0048] In step 4, the time-domain spectral data after the corrected echo is subjected to fast Fourier transform, and the optical parameters of the sample, including refractive index, absorption coefficient, and dielectric constant, are solved using the improved Dorney-Duvillaret model.

[0049] Specifically, a fast Fourier transform (FFT) is first performed on the time-domain waveform of the terahertz pulse passing through free space or a reference sample, along with the time-domain waveform collected in this experiment, to determine the sample's absorbance. To minimize the impact of echoes on the results, the main wave of the time-domain signal is retained, while the portion that generates echoes is reduced, and the echo location is estimated. Finally, the amplitude and phase information in the frequency domain is converted into optical parameters such as the sample's refractive index, absorption coefficient, and dielectric constant using a modified Dorney-Duvillaret model.

[0050] The present invention generates frequency domain amplitude spectrum ρ(ω) and phase spectrum After screening the effective frequency band, the 0.1-2.5 THz band was selected based on the improved Dorney-Duvillaret model (such as Equations 1 to 3) to complete the extraction of optical parameters.

[0051]

[0052] According to formula (1) combined with the terahertz spectrum value of stripe rust uredia, the refractive index of the sample can be calculated. The refractive index of active uredia is 1.40, and that of inactive uredia is about 1.44. Similarly, its absorption coefficient and dielectric constant can be calculated, and the absorption coefficient has a strong frequency dependence. Finally, the method is repeated to calculate the optical parameters of polyethylene powder, and the calculated values are compared with the actual values. The results show that the error after correction is only 0.2%, as shown in Figure 2. Figure 4 As shown, the feasibility of the model is verified.

[0053] Based on the above analysis, the optical parameter extraction method of wheat stripe rust uredia based on terahertz time-domain spectroscopy technology proposed in the present invention can effectively eliminate the influence of echoes. The spectrum calculation results after removing echoes are closer to the precise value. The calculation of the optical parameters of wheat stripe rust biological samples is more accurate and effective. Terahertz time-domain spectroscopy data can be used to quickly and accurately solve the optical parameters of pathogen spores, providing a new idea for the determination of optical parameters of biological samples.

Claims

1. A method for extracting optical parameters of wheat stripe rust uredospores based on terahertz time-domain spectroscopy, characterized in that: The steps include: Step 1, mixing wheat stripe rust uredia powder and polyethylene powder and pressing the mixture into tablets; Step 2, using the transmission module of the terahertz time-domain spectrometer to collect time-domain spectrum data of the mixed pressed tablet in a nitrogen environment; Step 3: Filter the collected time-domain spectral data to eliminate high-frequency noise; Step 4: Use the time domain window method to retain the main wave in the time domain signal, use the normal distribution to randomly generate sampling points, estimate the echo position and replace the original echo in the time domain signal to achieve echo signal correction and reduce time domain interference; Step 5: Perform fast Fourier transform on the corrected time domain signal, and calculate the refractive index, absorption coefficient and dielectric constant of wheat stripe rust uredospores through frequency domain amplitude and phase information based on the improved Dorney-Duvillaret model.

2. The method for extracting optical parameters of wheat stripe rust uredospores based on terahertz time-domain spectroscopy according to claim 1, wherein: In the step 1, wheat stripe rust uredia powder and polyethylene powder are mixed in a mass ratio of 1:9, and pressed using a tablet press at a pressure of 2 tons for 5 minutes to form a mixed tablet with a diameter of 12 mm and a thickness of 0.5 to 2 mm; and the tableting adopts a staged pressure gradient, i.e., 0.5 tons pre-pressing for 1 minute → 2 tons main pressing for 3 minutes → 1 ton final pressing for 1 minute, gradually removing air between the powders and reducing the internal porosity.

3. The method for extracting optical parameters of wheat stripe rust uredospores based on terahertz time-domain spectroscopy according to claim 1, wherein: The terahertz time-domain spectrometer has an operating frequency range of 0.05 to 5 THz, a dynamic range of ≥90 dB, and a spectral resolution of ≤0.85 GHz. Before collecting time-domain spectral data, nitrogen is introduced until the water vapor absorption peak disappears, eliminating the strong absorption interference of moisture on terahertz waves. When collecting time-domain spectral data, the pressed sheet is placed in the transmission light path, and the sample is rotated in four different directions and measured repeatedly to reduce local scattering errors caused by uneven mixing.

4. The method for extracting optical parameters of wheat stripe rust uredospores based on terahertz time-domain spectroscopy according to claim 1, wherein: In step 3, a Gaussian window function is applied to the collected time domain spectrum to filter out high-frequency noise.

5. The method for extracting optical parameters of wheat stripe rust uredospores based on terahertz time-domain spectroscopy according to claim 1, wherein: In step 4, first detect the peak position t of the main wave of the time domain signal. 主波 , and calculate the theoretical time difference Δτ between the main wave and the echo 理论 ; Then directly measure the actual time difference Δτ between the main wave peak and the echo peak from the time domain signal 测量 , and calculate Δτ 理论 and Δτ 测量 Deviation δτ=Δτ 测量 -Δτ 理论 ; Adjust the echo position to t 回波 ∈[t 主波 ±δτ], using statistical signals, a normally distributed random signal is generated in the adjusted echo position interval to replace the original echo in the time domain signal, thereby achieving the purpose of dynamically correcting the original echo signal.

6. The method for extracting optical parameters of wheat stripe rust uredospores based on terahertz time-domain spectroscopy according to claim 5, characterized in that: The theoretical time difference Δτ between the main wave and the echo 理论 =2nd / c, where n is the refractive index of the mixed pellet, d is the pellet thickness, and c is the speed of light.

7. The method for extracting optical parameters of wheat stripe rust uredospores based on terahertz time-domain spectroscopy according to claim 6, characterized in that: In step 4, 800 sampling points are randomly generated using normal distribution, corresponding to a 5ps time window, and the signal in the original echo interval is replaced with a mean value of -0.012 and a standard deviation of 0.001, and the echo signal is dynamically corrected to form a complete time domain waveform.

8. The method for extracting optical parameters of wheat stripe rust uredospores based on terahertz time-domain spectroscopy according to claim 1, characterized in that: In step 5, after fast Fourier transform, data of the effective frequency band of 0.1 to 2.5 THz is extracted for calculation.

9. The method for extracting optical parameters of wheat stripe rust uredospores based on terahertz time-domain spectroscopy according to claim 1, characterized in that: In step 5, the refractive index n of the sample tab (ω), absorption coefficient α tab (ω) and dielectric constant ε tab Solved by the calculation formula of the improved Dorney-Duvillaret model: Where c is the speed of light, d is the thickness of the sheet, φ(ω) is the phase difference between the sample signal and the reference signal, ω is the angular frequency, ρ(ω) is the amplitude ratio of the sample signal to the reference signal, and i is the imaginary unit.

10. The method for extracting optical parameters of wheat stripe rust uredospores based on terahertz time-domain spectroscopy according to claim 1, characterized in that: The wheat stripe rust uredia are active or inactive stripe rust uredia, wherein the refractive index of the active uredia is in the range of 1.40 to 1.42, the real part of the dielectric constant is 1.95, the refractive index of the inactive uredia is in the range of 1.44 to 1.45, the real part of the dielectric constant is 2.1, and the refractive index error after correction is less than 0.2%.

Images