A wavelet threshold denoising method and optical time domain reflectometer based on the method
By adopting the adaptive wavelet threshold denoising method in the optical time domain reflector, the problems of denoising discontinuity and low accuracy in the traditional method are solved, and higher signal-to-noise ratio and smaller mean square error are achieved, which improves the Denoising signal quality.
Patent Information
- Application Number
- CN202110627515.4
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- Filing Date
- 2021-06-04
- Publication Date
- 2024-11-08
- Estimated Expiration
- 2041-06-04
AI Technical Summary
The traditional wavelet threshold denoising method has problems such as discontinuity of hard threshold function in the denoising process, resulting in oscillation and pseudo-Gibbs effect, and the soft threshold function has constant deviation, resulting in reduced accuracy, and unsatisfactory signal-to-noise ratio and mean square error.
A wavelet threshold denoising method adapted to different scales is designed. Through wavelet decomposition, appropriate threshold selection and wavelet threshold processing, the denoising signal is reconstructed, combined with the wavelet decomposition module, threshold selection module, threshold processing module and signal reconstruction module. Used for denoising processing in optical time domain reflectors.
The accuracy after signal denoising is improved, the signal-to-noise ratio is improved, and the mean square error is reduced, and a higher quality denoising signal is obtained.
Smart Images

Figure CN113556169B_ABST
Abstract
Description
Technical Field
[0001] The invention relates to the technical field of optical fiber testing, and in particular to a wavelet threshold denoising method and an optical time domain reflectometer based on the method. Background Art
[0002] Optical Time Domain Reflectometry (OTDR) is a non-destructive test instrument designed according to the transmission characteristics of optical fiber. It is one of the main instruments in optical fiber testing. It can quantitatively test the length of optical fiber, transmission attenuation of optical fiber and joint attenuation. It has the advantages of distributed measurement, remote measurement, high sensitivity, strong anti-electromagnetic interference ability, good insulation, light weight and small size. It has been widely used in the construction, maintenance and operation of communication optical fiber and sensor optical fiber. However, the actual OTDR test curve contains a lot of noise, especially at the tail of the curve. Due to the large attenuation of optical power, the curve is seriously polluted by noise, and the events in the curve are difficult to identify. Noise is an important factor affecting the detection and recognition performance of target signals. Especially in the analysis of some high-precision data, even very weak noise will have a huge impact on the analysis results. Therefore, in the process of signal analysis, the first thing to do is to denoise the signal. Wavelet analysis is a new signal processing tool developed in recent years. This method is derived from Fourier analysis. Wavelet, that is, a wave in a small area, has non-zero values only in a very limited interval, rather than being endless like sine and cosine waves. Wavelets can be translated forward and backward along the time axis, and can also be stretched and compressed proportionally to obtain low-frequency and high-frequency wavelets. The constructed wavelet function can be used to filter or compress signals, so that useful signals can be extracted from noise signals. Therefore, wavelet theory has developed very rapidly. Commonly used wavelet denoising methods include: wavelet threshold denoising, wavelet correlation denoising and wavelet modulus maximum denoising. Wavelet threshold denoising was proposed by DL Donoho and IM Johnstone in 1992. The basic idea of this method is that when a noisy signal is decomposed by wavelet, the energy of the signal itself exists in some wavelet decomposition coefficients, while the energy of the noise is distributed in all wavelet decomposition coefficients. After multi-layer wavelet decomposition of the noisy signal, the modulus value of the wavelet coefficient of the signal itself is greater than the modulus value of the coefficient of the wavelet transform of the noise signal. After threshold processing and wavelet reconstruction, the denoised signal is obtained. Among all wavelet denoising methods, wavelet threshold denoising method is the most widely used and has become one of the research hotspots in the field of signal denoising in recent years.
[0003] The functions used in traditional wavelet threshold denoising methods include hard threshold function and soft threshold function, but both methods have certain defects: the hard threshold function is discontinuous at the threshold, resulting in oscillation and pseudo-Gibbs effect in the reconstructed signal; although the soft threshold function is continuous at the threshold, there is a constant deviation between the processed wavelet coefficients and the true wavelet coefficients, resulting in reduced accuracy of wavelet coefficient reconstruction and poor denoising effect. In order to overcome the defects of soft and hard threshold functions, researchers at home and abroad have proposed a ratio smoothing threshold function. The ratio smoothing threshold function is continuous at the threshold, and can also solve the problem of poor denoising effect of the soft threshold function to a certain extent, but the signal-to-noise ratio and mean square error obtained by denoising with this function are not ideal compared with the hard threshold function.
[0004] The function forms and images of the above three methods are as follows.
[0005] (1) Hard threshold function:
[0006]
[0007] (2) Soft threshold function:
[0008]
[0009] (3) Ratio smoothing threshold function:
[0010]
[0011] Among them, w j,k is the kth wavelet coefficient at the jth scale of decomposition, w j,k The estimated wavelet coefficients, n is the adjustment factor, which is a positive integer. p=(e λ +e -λ ) / 2, λ is the threshold, N is the length of the signal, σ n is the standard deviation of the noise contained in the first layer, σ n =Median(|w 1,k |) / 0.6745.
[0012] The soft threshold function, hard threshold function and ratio smoothing threshold function are shown in the attached figure. Figure 1 shown. Summary of the invention
[0013] The purpose of the present invention is to design a wavelet threshold denoising method and an optical time domain reflectometer based on the method, so that the accuracy of the signal after denoising can be improved, the denoising effect of the signal can be effectively improved, a higher signal-to-noise ratio and a smaller mean square error can be obtained, and a high-quality denoised signal can be obtained.
[0014] To achieve the above object, the present invention claims a wavelet threshold denoising method, comprising the following steps:
[0015] Step 1: Perform wavelet decomposition on the noisy signal to obtain the wavelet decomposition coefficient w j,k ;
[0016] Step 2: Select an appropriate threshold λ j ; In the actual denoising process, after the noisy signal is decomposed by wavelet, as the scale increases, the amplitude of the wavelet coefficient of the noise becomes smaller and smaller, while the amplitude of the wavelet coefficient of the signal becomes larger and larger. Therefore, the selection of thresholds at different scales should also be different so that it can adapt to the noise distribution of each layer;
[0017] Step 3: Decompose the wavelet coefficients w j,k Perform wavelet threshold processing to obtain the estimated wavelet decomposition coefficients
[0018] Step 4: Estimate the wavelet decomposition coefficients Reconstruct and obtain the denoised signal.
[0019] Specifically, in step 1, w j,k is the kth wavelet coefficient at the jth scale of decomposition.
[0020] Specifically, in step 2, Among them, N j is the length of the j-th layer wavelet decomposition coefficients, σ j is the standard deviation of the noise contained in the jth layer, σ j =Median(|w j,k ) / 0.6745. It can be seen that as the decomposition scale j increases, the threshold decreases, which is consistent with the characteristic that the amplitude of the noise decreases when the decomposition scale increases.
[0021] Specifically, in step 3,
[0022]
[0023] Among them, n and m are adjustment factors, both of which are positive integers;
[0024] The present invention also claims protection for an optical time domain reflectometer based on the above-mentioned wavelet threshold denoising method, comprising a pulse generator, a laser driving module, a fiber coupler, a photoelectric conversion module, an A\D conversion module and a microcontroller module connected in sequence; the microcontroller module is connected to the threshold denoising module.
[0025] The pulse generator mainly includes an electronic circuit for generating electric pulses, and the electric pulses generated by the pulse generator correspond to the optical pulses required for OTDR measurement; the laser driving module is used to convert the electric pulses generated by the pulse generator into optical pulses, and send the optical pulses to the optical fiber coupler through the optical fiber; the optical fiber coupler sends the optical pulses to the optical fiber to be tested through the optical fiber so that the operator can provide detection feedback, and at the same time sends the optical pulses reflected in the optical fiber to be tested to the photoelectric conversion module for detection; the photoelectric conversion module receives the optical signal to be tested and converts it into an analog electrical signal; the A\D conversion module converts the analog electrical signal into a digital electrical signal and sends it to the microcontroller module for processing; the microcontroller module uses a storage device to save the noisy signal data sent from the A\D conversion module, and then transmits the data to the threshold denoising module through the bus, and controls the working process of the threshold denoising module.
[0026] Specifically, the threshold denoising module includes a wavelet decomposition module, a threshold selection module, a threshold processing module and a signal reconstruction module which are connected in sequence.
[0027] The wavelet decomposition module selects a suitable wavelet basis function and performs wavelet decomposition on the noisy signal. After decomposition, a series of low-frequency wavelet coefficients and high-frequency wavelet coefficients are obtained; the low-frequency wavelet coefficients remain unchanged, and the high-frequency wavelet coefficients enter the threshold selection module to select a suitable threshold; the threshold processing module uses the above wavelet threshold function for processing; the signal reconstruction module combines the processed high-frequency wavelet coefficients and low-frequency wavelet coefficients to reconstruct the OTDR signal.
[0028] Compared with the prior art, the present invention has the following beneficial effects:
[0029] Compared with the traditional hard threshold function, soft threshold function and ratio smoothing threshold function, the wavelet threshold function used in the wavelet threshold denoising method of the present invention can obtain a better signal-to-noise ratio and a smaller mean square error for denoising noisy signals, thereby effectively improving the quality of the denoised signal.
[0030] The optical time domain reflectometer based on the wavelet threshold denoising method of the present application can effectively detect optical fibers and improve the detection effect under the mutual cooperation of the pulse generator, laser driving module, fiber coupler, photoelectric conversion module, A\D conversion module, microcontroller module and threshold denoising module. BRIEF DESCRIPTION OF THE DRAWINGS
[0031] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.
[0032] Figure 1 It is a comparison chart of soft threshold function, hard threshold function and ratio smoothing threshold function;
[0033] Figure 2 A comparison diagram of the wavelet threshold function, the soft threshold function and the hard threshold function of the present invention;
[0034] Figure 3 It is a schematic structural diagram of the optical time domain reflectometer of the present invention;
[0035] Figure 4 It is a structural schematic diagram of the threshold denoising module of the present invention; DETAILED DESCRIPTION
[0036] The following will be combined with the accompanying drawings in the present invention to clearly and completely describe the technical solution in the present invention. Obviously, what is described is only a part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments in the present invention, all other embodiments obtained by ordinary technicians in this field without making creative work are within the scope of protection of the present invention.
[0037] A wavelet threshold denoising method comprises the following steps:
[0038] Step 1: Perform wavelet decomposition on the noisy signal to obtain the wavelet decomposition coefficient w j,k ;
[0039] Step 2: Select an appropriate threshold λ j ; In the actual denoising process, after the noisy signal is decomposed by wavelet, as the scale increases, the amplitude of the wavelet coefficient of the noise becomes smaller and smaller, while the amplitude of the wavelet coefficient of the signal becomes larger and larger. Therefore, the selection of thresholds at different scales should also be different so that it can adapt to the noise distribution of each layer;
[0040] Step 3: Decompose the wavelet coefficients w j,k Perform wavelet threshold processing to obtain the estimated wavelet decomposition coefficients
[0041] Step 4: Estimate the wavelet decomposition coefficients Reconstruct and obtain the denoised signal.
[0042] Specifically, in step 1, w j,k is the kth wavelet coefficient at the jth scale of decomposition.
[0043] Specifically, in step 2, Among them, N j is the length of the j-th layer wavelet decomposition coefficients, σ j is the standard deviation of the noise contained in the jth layer, σ j =Median(|wj,k ) / 0.6745. It can be seen that as the decomposition scale j increases, the threshold decreases, which is consistent with the characteristic that the amplitude of the noise decreases when the decomposition scale increases.
[0044] Specifically, in step 3,
[0045]
[0046] Among them, n and m are adjustment factors, both of which are positive integers;
[0047] An optical time domain reflectometer based on the wavelet threshold denoising method comprises a pulse generator, a laser driving module, a fiber coupler, a photoelectric conversion module, an A\D conversion module and a microcontroller module connected in sequence; the microcontroller module is connected to the threshold denoising module.
[0048] The pulse generator mainly includes an electronic circuit for generating electric pulses, and the electric pulses generated by the pulse generator correspond to the optical pulses required for OTDR measurement; the laser driving module is used to convert the electric pulses generated by the pulse generator into optical pulses, and send the optical pulses to the optical fiber coupler through the optical fiber; the optical fiber coupler sends the optical pulses to the optical fiber to be tested through the optical fiber so that the operator can provide detection feedback, and at the same time sends the optical pulses reflected in the optical fiber to be tested to the photoelectric conversion module for detection; the photoelectric conversion module receives the optical signal to be tested and converts it into an analog electrical signal; the A\D conversion module converts the analog electrical signal into a digital electrical signal and sends it to the microcontroller module for processing; the microcontroller module uses a storage device to save the noisy signal data sent from the A\D conversion module, and then transmits the data to the threshold denoising module through the bus, and controls the working process of the threshold denoising module.
[0049] Specifically, the threshold denoising module includes a wavelet decomposition module, a threshold selection module, a threshold processing module and a signal reconstruction module which are connected in sequence.
[0050] The wavelet decomposition module selects a suitable wavelet basis function and performs wavelet decomposition on the noisy signal. After decomposition, a series of low-frequency wavelet coefficients and high-frequency wavelet coefficients are obtained; the low-frequency wavelet coefficients remain unchanged, and the high-frequency wavelet coefficients enter the threshold selection module to select a suitable threshold; the threshold processing module uses the above wavelet threshold function for processing; the signal reconstruction module combines the processed high-frequency wavelet coefficients and low-frequency wavelet coefficients to reconstruct the OTDR signal.
[0051] The comparison of soft threshold function, hard threshold function and wavelet threshold function of the present invention is shown in the attached figure. Figure 4 shown.
[0052] In order to verify the effectiveness of the denoising of this system, simulation experiments were carried out using the soft threshold function method, the hard threshold function method, the ratio smoothing threshold function method and the wavelet threshold function method of the present invention in the threshold denoising module of the system. The noisy signal was decomposed by wavelet with a scale of 6 using the Sym6 wavelet basis. The adjustment factors n and m in the improved threshold function proposed by the present invention were 3 and 1 respectively. The signal-to-noise ratio (SNR) and mean square error (RMSE) obtained by the four methods are shown in Table 1.
[0053] Table 1 Comparison of SNR and RMSE after denoising with different threshold functions
[0054]
[0055] As can be seen from Table 1, the system of the present invention can effectively improve the signal-to-noise ratio of the signal, and the improvement of the signal-to-noise ratio is greater than that of the hard threshold method, the soft threshold method and the ratio smoothing threshold method. In addition, the mean square error of the present invention is reduced compared with the hard threshold method, the soft threshold method and the ratio smoothing threshold method, and the result is better than other denoising systems.
Claims
1. A wavelet threshold denoising method, characterized in that: The steps include: Step 1: Perform wavelet decomposition on the noisy signal to obtain the wavelet decomposition coefficients is the kth wavelet coefficient at the jth scale of decomposition; Step 2: Choose an appropriate threshold ;in, is the length of the j-th layer wavelet decomposition coefficients, is the standard deviation of the noise contained in the jth layer, ; Step 3: Decompose the wavelet coefficients Perform wavelet threshold processing to obtain the estimated wavelet decomposition coefficients Among them, n and m are adjustment factors, both of which are positive integers; Step 4: Estimate the wavelet decomposition coefficients Reconstruct and obtain the denoised signal.
2. An optical time domain reflectometer based on the wavelet threshold denoising method according to claim 1, characterized in that: It comprises a pulse generator, a laser driving module, an optical fiber coupler, a photoelectric conversion module, an A\D conversion module and a microcontroller module which are connected in sequence; the microcontroller module is connected to a threshold denoising module.
3. The optical time domain reflectometer according to claim 2, characterized in that: The threshold denoising module comprises a wavelet decomposition module, a threshold selection module, a threshold processing module and a signal reconstruction module which are connected in sequence.
