Optical fiber distributed seismic wave data processing method based on CCRN

Abstract

The invention provides a CCRN-based optical fiber distributed seismic wave data processing method. The CCRN-based optical fiber distributed seismic wave data processing method comprises the following steps: step 1, constructing a seismic wave signal data set; step 2, constructing a CCRN network; 3, inputting the data into the network for training; and step 4, mapping the complex frequency spectrum containing the noise signal to the complex frequency spectrum of the reconstructed signal so as to restore a relatively clean signal. The optical fiber distributed seismic wave data processing method based on the CCRN can be used for enhancing signal noise reduction of a distributed acoustic wave sensing system, seismic wave signals collected by the distributed acoustic wave sensing system can be effectively processed, and the signal-to-noise ratio of the seismic wave signals is increased; and the method can also be effectively applied to the distributed sound wave sensing system for noise signal elimination, and the measurement capability of the distributed sound wave sensing system is improved.

Description

Technical Field

[0001] The present invention relates to the technical field of optical fiber distributed seismic wave signal noise reduction, and in particular to an optical fiber distributed seismic wave data processing method based on CCRN. Background Art

[0002] When seismic waves propagate inside the earth, they are affected by various complex effects of underground media, such as reflection, refraction, scattering, diffraction, etc. Therefore, seismic waves carry a lot of information during propagation. For geophysicists, seismic wave data is an important source of information for studying the internal structure of the earth and seismic activities. However, seismic signals often contain a lot of noise, which has a great impact on the quality of seismic data and hinders the analysis and interpretation of seismic wave data.

[0003] Seismic wave noise reduction is to remove the noise signal in the seismic signal through a series of digital signal processing methods to obtain clearer and more accurate seismic data. There are two main sources of noise: one is the noise from human activities on the surface, such as traffic noise, machine noise, construction, etc., and the other is from natural factors, such as wind, rain, waves, etc. In addition, nonlinear effects in underground media can also produce noise signals.

[0004] By reducing the noise of seismic waves, more accurate seismic data can be obtained, the accuracy of seismic imaging and positioning can be improved, and more accurate underground structures and earthquake mechanisms can be inferred. This has important application value in seismic exploration, seismic monitoring, earthquake prediction, and earthquake disaster prevention and control.

[0005] In the existing technology, the main methods for reducing noise of seismic waves are: (1) KL transform: This method extracts the coherent information between the waveform and energy of the effective components of adjacent channels and eliminates random noise. (2) Short-time Fourier transform: By obtaining the spectrum information in the time domain of the signal, the information is filtered. (3) Singular value decomposition (SVD): The noisy signal vector space is decomposed into two subspaces dominated by the pure signal and the noise signal respectively, and then the pure signal is estimated by removing the noisy signal vector component falling in the "noise space".

[0006] Most of the current seismic wave denoising methods target a single type of noise, have poor generalization performance, and have limited effects on seismic wave data with complex noise types. In addition, there are many problems such as the difficulty in effectively suppressing non-stationary noise in DAS systems.

[0007] Distributed fiber acoustic sensing (DAS) technology can realize long-distance, distributed monitoring of dynamic strain along optical fibers and is currently widely used in many fields such as vertical seismic profile (VSP) imaging.

[0008] When using DAS combined with VSP technology to collect seismic wave data, the irregular logging route makes it difficult for the optical cable to couple with the well wall, and the noise of spring waves downhole and other noises make the data blurred. Therefore, an effective noise reduction algorithm is urgently needed to remove various noises from the seismic wave data collected by DAS, so as to improve the performance of DAS and the signal-to-noise ratio of DAS data, which is beneficial to the subsequent earthquake reproduction work.

[0009] In the Chinese patent application with application number: CN202010080534.5, a method and device for removing wellbore wave interference in fiber optic acoustic wave sensing seismic data is involved. The method includes: obtaining seismic wave field data containing wellbore wave interference collected by the fiber optic acoustic wave sensor; intercepting the waveform data of a preset time window length downward on each seismic wave along the wellbore wave, calculating the root mean square amplitude of the waveform data intercepted on each seismic wave as the amplitude normalization factor of each seismic wave, so as to normalize the amplitude value of each seismic wave at each time sampling point; performing tail-removing mean filtering on the normalized amplitude value of each seismic wave at each time sampling point to obtain the predicted amplitude value of the wellbore wave at each time sampling point, and based on this, performing wellbore wave interference removal processing on each seismic wave at each time sampling point, and finally performing inverse normalization processing to obtain seismic wave field data with wellbore wave interference removed. The invention effectively suppresses wellbore wave interference in fiber optic acoustic wave sensing seismic data.

[0010] In the Chinese patent application with application number: CN202010180125.2, a well-ground seismic data joint acquisition system based on distributed fiber optic acoustic sensing and a well-drive data processing method are involved, in order to solve the problems existing in the prior art of inconsistent energy, inconsistent spectrum, and inconsistent coupling between the seismic source and the ground caused by repeated excitation of each seismic source point when the well and ground seismic data are collected separately; the invention is based on a well-ground seismic data joint stereo acquisition system composed of a distributed fiber optic acoustic sensing seismic data acquisition unit in the well and a ground seismic data acquisition unit, to conduct well-ground joint stereo exploration and synchronous acquisition of ground and well seismic data, to realize high-density, high-efficiency, high-resolution, and low-cost well-ground joint stereo seismic exploration technology, and to conduct oil and gas resource exploration and comprehensive evaluation.

[0011] In the Chinese patent application with application number: CN202110562514.6, a fiber-distributed seismic wave signal denoising method based on a generative adversarial network is involved. The random noise and cable wave noise of the fiber-distributed seismic wave signal are effectively removed by using a generative adversarial network. The network includes an attention generation network and an attention discrimination network, wherein the attention generation network is composed of an attention recurrent neural network and an autoencoder. The attention recurrent neural network is constructed using the attention mechanism to generate a spatial attention weight matrix for the random noise and cable wave noise in the seismic wave image signal; then the context convolutional autoencoder is used to generate the denoised DAS seismic wave signal; when the attention discrimination network cannot determine whether the input denoised DAS seismic wave signal is real data from the database or data generated by the generation network, the fiber-distributed seismic wave signal denoising algorithm based on the generative adversarial network is trained.

[0012] In the Chinese patent application with application number: CN202111068897.8, a distributed fiber optic seismic data acquisition system and its data collection and conversion method are involved, including performing time differential processing on the fiber optic particle displacement data collected by downhole or ground optical fiber, converting it into DAS seismic data with the same dimension as the seismic data recorded by the velocity detector, and then performing discrete Fourier transform (DFT) on the seismic data recorded by the detector and the DAS seismic data after the differential processing, comparing the amplitude spectra of the two data after the discrete Fourier transform, and obtaining the ratio ΔAG / F(ω) between the amplitude spectrum of the seismic data recorded by the detector corresponding to each discrete angular frequency ω and the amplitude spectrum of the converted DAS seismic data on the amplitude spectrum of the full frequency band, multiplying the coefficient corresponding to each frequency point of the DAS seismic data by ΔAG / F(ω), and finally performing inverse discrete Fourier transform on the corrected amplitude spectrum of the DAS seismic data that has been subjected to amplitude spectrum compensation processing, so as to obtain DAS seismic data that is exactly the same as the seismic data recorded by the detector at the same receiving position.

[0013] The above existing technologies are greatly different from the present invention and fail to solve the technical problem we want to solve. For this reason, we have invented a new fiber-optic distributed seismic wave data processing method based on CCRN. Summary of the invention

[0014] The purpose of the present invention is to provide a new algorithm for effectively reducing the noise of seismic wave signals collected by DAS and improving the data signal-to-noise ratio.

[0015] The object of the present invention can be achieved by the following technical measures: a fiber-optic distributed seismic wave data processing method based on CCRN, the fiber-optic distributed seismic wave data processing method based on CCRN comprising:

[0016] Step 1, constructing a seismic wave signal data set;

[0017] Step 2: Build CCRN network;

[0018] Step 3, input data into the network for training;

[0019] Step 4: Map the complex spectrum of the noisy signal to the complex spectrum of the reconstructed signal, thereby restoring a cleaner signal.

[0020] The purpose of the present invention can also be achieved by the following technical measures:

[0021] In step 1, vertical seismic profiling technology is used in combination with the DAS system to collect seismic wave signals.

[0022] In step 2, a CCRN network is constructed based on a convolutional recurrent network to reconstruct the signal through complex domain mapping.

[0023] In step 2, in the proposed network, the encoder includes five convolutional layers, while the decoder has five deconvolutional layers. The exponential linear unit (ELU) is applied to all convolutional and deconvolutional layers except the output layer, where softplus activation is used.

[0024] In step 2, batch normalization is applied after each convolution or deconvolution and before activation; and the number of kernels is kept symmetric: that is, the number of kernels gradually increases in the encoder and gradually decreases in the decoder; in order to utilize a larger context along the frequency direction, a stride of 2 is applied to all convolutional or deconvolution layers along the frequency dimension; that is, the frequency dimension size of the noisy signal is halved layer by layer in the encoder and doubled layer by layer in the decoder without changing the time dimension size of the noisy signal; in order to improve the information and gradient flow of the entire network, skip connections are used to connect the output of each encoder layer with the input of each decoder layer.

[0025] In step 2, in order to obtain a causal system for real-time speech enhancement, causal convolution is applied on the encoder-decoder structure; the input can be regarded as a sequence of feature vectors; in causal convolution, the output does not depend on future inputs; using causal convolution instead of non-causal convolution, the encoder-decoder architecture constitutes a recurrent structure.

[0026] In step 3, the phase information obtained by demodulating the collected data is extracted according to the positioning position, and it is input into the network in sequence form together with the original seismic wave signal for training.

[0027] In step 4, a coding and decoding method is used to iteratively estimate the nonlinear mapping relationship from the noisy signal to the clean sound signal, and the complex spectrum of the noisy signal is mapped to the complex spectrum of the reconstructed signal, thereby restoring a relatively clean signal.

[0028] In step 4, the purpose of complex spectrum mapping is to extract the real spectrum and imaginary spectrum of the clean acoustic signal from the noisy signal, while enhancing the amplitude response and phase response of the seismic wave signal.

[0029] In step 4, for a noisy signal:

[0030] y=x+n

[0031] Among them, y is the noisy signal; x is the clean sound signal; n is the noise signal;

[0032] The mapping of the amplitude spectrum is to perform a short-time Fourier transform on both sides to obtain:

[0033] Y=X+N

[0034] Where Y, X, and N represent the short-time Fourier transform of y, x, and n, respectively;

[0035] The proposed complex spectrum mapping is to apply the representation of short-time Fourier transform to the rectangular coordinate system, that is, to modify the above formula to:

[0036] (Y r +Y i )=(X r +N r )+i(X i +N i )

[0037] Among them, r represents the real part and i represents the imaginary part.

[0038] In step 4, the complex ideal ratio mask is estimated by the model, and the result is:

[0039]

[0040] Thus, the enhanced spectrum can be estimated by masking the noisy spectrum with complex ideal ratios: To obtain, and used to reconstruct the seismic wave signal:

[0041]

[0042] In step 4, after the enhanced real spectrum and the imaginary spectrum are combined and subjected to short-time inverse Fourier transform, the noise-reduced seismic wave signal collected by the DAS system can be obtained.

[0043] The fiber-optic distributed seismic wave data processing method based on CCRN (convolutional recurrent neural network) in the present invention can be used to enhance the signal noise reduction of the distributed acoustic wave sensing system, and can effectively process the seismic wave signals collected by the distributed acoustic sensing system, improve the signal-to-noise ratio of the seismic wave signals, and improve the quality of the collected signals. The CCRN technology proposed in the present invention can also be effectively applied to the distributed acoustic wave sensing system to eliminate noise signals and improve the measurement capability of the distributed acoustic wave sensing system.

[0044] The core of the present invention is a seismic wave processing algorithm based on the DAS system, which is original. The CCRN proposed in the present invention is improved on the basis of CRN. The commonly used CRN generally uses the amplitude spectrum to map or mask the signal. The recent DCCRN and U-NET networks have used complex spectra to improve the performance of the network. The complex spectrum can not only improve the quality of the network-enhanced voice signal, but also reduce the error of the phase of the signal after reconstruction, which is very important for a device such as the DAS system that is positioned by a phase signal. Therefore, the present invention also uses the complex spectrum to improve the mapping operation of CRN. In addition, in order to continue to improve the performance of the network and reduce the interference of the network to the noise part, an attention mechanism and a smooth transition hole convolution module are added to CCRN. Finally, this algorithm is used in DAS seismic wave data processing. BRIEF DESCRIPTION OF THE DRAWINGS

[0045] Figure 1 It is a structural schematic diagram of the distributed acoustic wave sensing system of the present invention;

[0046] Figure 2 A schematic diagram of a CCRN network structure in a specific embodiment of the present invention;

[0047] Figure 3 It is a flow chart of a specific embodiment of the fiber optic distributed seismic wave data processing method based on CCRN of the present invention. DETAILED DESCRIPTION

[0048] It should be noted that the following detailed descriptions are exemplary and are intended to provide further explanation of the present invention. Unless otherwise specified, all technical and scientific terms used herein have the same meanings as those commonly understood by those skilled in the art to which the present invention belongs.

[0049] It should be noted that the terms used herein are only for describing specific embodiments and are not intended to limit exemplary embodiments according to the present invention. As used herein, unless the context clearly indicates otherwise, the singular form is also intended to include the plural form. In addition, it should be understood that when the terms "comprising" and / or "including" are used in this specification, it indicates the presence of features, steps, operations and / or combinations thereof.

[0050] like Figure 3 As shown, Figure 3 The figure is a flow chart of the fiber-optic distributed seismic wave data processing method based on CCRN of the present invention. The fiber-optic distributed seismic wave data processing method based on CCRN includes:

[0051] Step 101, constructing a seismic wave signal data set.

[0052] Step 102, constructing a CCRN network based on a convolutional recurrent network to reconstruct the signal through complex domain mapping.

[0053] Step 103, extract the phase information obtained by demodulating the collected data according to the positioning position, and input it and the original seismic wave signal into the network in a sequence form for training.

[0054] Step 104 , adopting a coding and decoding method, by iteratively estimating the nonlinear mapping relationship from the noisy signal to the clean sound signal, the complex spectrum of the noisy signal is mapped to the complex spectrum of the reconstructed signal, thereby restoring a relatively clean signal.

[0055] The following are several specific embodiments of the present invention.

[0056] Example 1

[0057] In a specific embodiment 1 of the present invention, the fiber-optic distributed seismic wave data processing method based on CCRN includes the following steps:

[0058] 1) Construct a seismic wave signal data set. Use vertical seismic profile technology and DAS system to collect seismic wave signals.

[0059] 2) A CCRN network was constructed based on a convolutional recurrent network to reconstruct signals through complex domain mapping.

[0060] like Figure 2 As shown in the CRNN network structure, in the proposed network, Figure 2The convolutional encoder in includes five convolutional layers, while the decoder has five deconvolutional layers. Exponential linear units (ELUs) are applied to all convolutional and deconvolutional layers except the output layer. ELUs have been shown to lead to faster convergence and better generalization than ReLUs. In the output layer, softplus activation is used, which is a smooth approximation of the ReLU function and can constrain the network output to always be positive. In addition, batch normalization is used after each convolution (or deconvolution) and before activation. The number of kernels is kept symmetric: the number of kernels gradually increases in the encoder and decreases in the decoder. In order to exploit a larger context along the frequency direction, a stride of 2 is applied to all convolutional (or deconvolution) layers along the frequency dimension. That is, the size of the frequency dimension of the noisy signal is halved layer by layer in the encoder and doubled layer by layer in the decoder without changing the size of the time dimension of the noisy signal. In order to improve the flow of information and gradients throughout the network, skip connections are used to connect the output of each encoder layer with the input of each decoder layer.

[0061] To obtain a causal system for real-time speech enhancement, causal convolutions are imposed on the encoder-decoder structure. The input can be viewed as a sequence of feature vectors. In causal convolutions, the output does not depend on future inputs. Using causal convolutions instead of non-causal convolutions, the encoder-decoder architecture will lead to a causal system. Causal deconvolution can be easily applied to the decoder since deconvolution is essentially a convolution operation.

[0062] 3) The phase information obtained by demodulating the collected data is extracted according to the positioning position, and it is input into the network in sequence form together with the original seismic wave signal for training.

[0063] 4) Using the encoding and decoding method, the nonlinear mapping relationship from the noisy signal to the clean sound signal is iteratively estimated, and the complex spectrum of the noisy signal is mapped to the complex spectrum of the reconstructed signal, thereby restoring a relatively clean signal.

[0064] The purpose of complex spectrum mapping is to extract the real spectrum and imaginary spectrum of the clean acoustic signal from the noisy signal, while enhancing the amplitude response and phase response of the seismic wave signal. In other words, for the noisy signal:

[0065] y=x+n

[0066] Among them, y is the noisy signal; x is the clean sound signal; n is the noise signal.

[0067] The mapping of the amplitude spectrum is to perform a short-time Fourier transform on both sides to obtain:

[0068] Y=X+N

[0069] Where Y, X, and N represent the short-time Fourier transform of y, x, and n, respectively.

[0070] The proposed complex spectrum mapping is to apply the representation of short-time Fourier transform to the rectangular coordinate system, that is, to modify the above formula to:

[0071] (Y r +Y i )=(X r +N r )+i(X i +N i )

[0072] Among them, r represents the real part and i represents the imaginary part.

[0073] The complex ideal ratio mask is estimated by the model, and we get:

[0074]

[0075] Thus, the enhanced spectrum can be estimated by masking the noisy spectrum with complex ideal ratios: To obtain, and used to reconstruct the seismic wave signal:

[0076]

[0077] Finally, after combining the enhanced real spectrum with the imaginary spectrum and performing a short-time inverse Fourier transform, the de-noised seismic wave signal collected by the DAS system can be obtained.

[0078] Example 2

[0079] In the specific embodiment 2 of the present invention, in the network convolutional coding layer and the decoding layer, the number of convolutional layers is adjusted according to the actual application requirements. The increase of convolutional layers and the deepening of the network will lead to a deeper mining of signal features, but will increase the calculation efficiency. By adjusting the number of layers to achieve effective mining of signal features, the number of convolutional layers can be increased to 8 or 10 in actual applications.

[0080] Example 3

[0081] In the specific embodiment 3 of the present invention, two convolutions are performed in parallel. The number of channels can be increased and the size of the channel convolution kernel can be adjusted. The small convolution kernel is used to extract the low-frequency features of the signal, and the large convolution kernel is used to extract the high-frequency features of the signal.

[0082] The present invention is a fiber-optic distributed seismic wave data processing method based on CCRN. The DAS system is the hardware system that completes the entire process, and the CCRN algorithm is based on the DAS system. Figure 1As shown, the distributed acoustic wave sensing system device is composed of a laser, an optical isolator, an acousto-optic modulator, a function generator, an erbium-doped fiber amplifier, a filter, a circulator, a sensing fiber, a 50:50 coupler, a Faraday rotating mirror, a piezoelectric ceramic, a photodetector, and an acquisition card. The laser generates a narrow linewidth laser light source and enters the optical isolator. The function of the optical isolator is to prevent the end face reflected light from reflecting back to the laser. The laser light enters the acousto-optic modulator that can generate a modulated pulse signal through the optical isolator. The function generator plays the role of driving the acousto-optic modulator. Then, the pulse signal generated by the acousto-optic modulator enters the erbium-doped fiber amplifier for pulse amplification, and the spontaneous radiation noise is eliminated through the filter. Then, the pulse signal enters the circulator, and enters the sensing fiber through the circulator to monitor the voice signal. The backward Rayleigh scattered signal enters the circulator from the sensing fiber and then enters the erbium-doped fiber amplifier for amplification. Then, the spontaneous radiation noise is eliminated through the filter. Then, the two detectors are used to detect the signal through an unbalanced Mach-Zehnder interferometer. Finally, the Rayleigh scattered signal is collected by the data acquisition card. The Mach-Zehnder interferometer consists of a 50:50 coupler, a piezoelectric ceramic, and two Faraday rotator mirrors.

[0083] Finally, it should be noted that the above is only a preferred embodiment of the present invention and is not intended to limit the present invention. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art can still modify the technical solutions recorded in the aforementioned embodiments or replace some of the technical features therein with equivalents. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present invention shall be included in the protection scope of the present invention.

[0084] Except for the technical features described in the specification, all other technical features are known technologies to those skilled in the art.

Claims

1. A fiber-optic distributed seismic wave data processing method based on CCRN, characterized in that: The fiber-optic distributed seismic wave data processing method based on CCRN includes: Step 1, constructing a seismic wave signal data set; Step 2: Build CCRN network; Step 3, input data into the network for training; Step 4: Map the complex spectrum of the noisy signal to the complex spectrum of the reconstructed signal, thereby restoring a cleaner signal.

2. The fiber-optic distributed seismic wave data processing method based on CCRN according to claim 1, characterized in that: In step 1, vertical seismic profiling technology is used in combination with the DAS system to collect seismic wave signals.

3. The fiber-optic distributed seismic wave data processing method based on CCRN according to claim 1, characterized in that: In step 2, a CCRN network is constructed based on a convolutional recurrent network to reconstruct the signal through complex domain mapping.

4. The fiber-optic distributed seismic wave data processing method based on CCRN according to claim 3, characterized in that: In step 2, in the proposed network, the encoder includes five convolutional layers, while the decoder has five deconvolutional layers. The exponential linear unit (ELU) is applied to all convolutional and deconvolutional layers except the output layer, where softplus activation is used.

5. The fiber-optic distributed seismic wave data processing method based on CCRN according to claim 4, characterized in that: In step 2, batch normalization is applied after each convolution or deconvolution and before activation; and the number of kernels is kept symmetric: that is, the number of kernels gradually increases in the encoder and gradually decreases in the decoder; in order to utilize a larger context along the frequency direction, a stride of 2 is applied to all convolutional or deconvolution layers along the frequency dimension; that is, the frequency dimension size of the noisy signal is halved layer by layer in the encoder and doubled layer by layer in the decoder without changing the time dimension size of the noisy signal; in order to improve the information and gradient flow of the entire network, skip connections are used to connect the output of each encoder layer with the input of each decoder layer.

6. The fiber-optic distributed seismic wave data processing method based on CCRN according to claim 5, characterized in that: In step 2, in order to obtain a causal system for real-time speech enhancement, causal convolution is applied on the encoder-decoder structure; the input can be regarded as a sequence of feature vectors; in causal convolution, the output does not depend on future inputs; using causal convolution instead of non-causal convolution, the encoder-decoder architecture constitutes a recurrent structure.

7. The fiber-optic distributed seismic wave data processing method based on CCRN according to claim 1, characterized in that: In step 3, the phase information obtained by demodulating the collected data is extracted according to the positioning position, and it is input into the network in sequence form together with the original seismic wave signal for training.

8. The fiber-optic distributed seismic wave data processing method based on CCRN according to claim 1, characterized in that: In step 4, a coding and decoding method is used to iteratively estimate the nonlinear mapping relationship from the noisy signal to the clean sound signal, and the complex spectrum of the noisy signal is mapped to the complex spectrum of the reconstructed signal, thereby restoring a relatively clean signal.

9. The fiber-optic distributed seismic wave data processing method based on CCRN according to claim 8, characterized in that: In step 4, the purpose of complex spectrum mapping is to extract the real spectrum and imaginary spectrum of the clean acoustic signal from the noisy signal, while enhancing the amplitude response and phase response of the seismic wave signal.

10. The fiber-optic distributed seismic wave data processing method based on CCRN according to claim 9, characterized in that: In step 4, for a noisy signal: y=x+n Among them, y is the noisy signal; x is the clean sound signal; n is the noise signal; The mapping of the amplitude spectrum is to perform a short-time Fourier transform on both sides to obtain: Y=X+N Where Y, X, and N represent the short-time Fourier transform of y, x, and n, respectively; The proposed complex spectrum mapping is to apply the representation of short-time Fourier transform to the rectangular coordinate system, that is, to modify the above formula to: (Y r +Y i )=(X r +N r )+i(X i +N i ) Among them, r represents the real part and i represents the imaginary part.

11. The CCRN-based optical fiber distributed seismic wave data processing method according to claim 10, characterized in that: In step 4, the complex ideal ratio mask is estimated by the model, and the result is: Thus, the enhanced spectrum can be estimated by masking the noisy spectrum with complex ideal ratios: To obtain, and used to reconstruct the seismic wave signal:

12. The fiber-optic distributed seismic wave data processing method based on CCRN according to claim 11, characterized in that: In step 4, after the enhanced real spectrum and the imaginary spectrum are combined and subjected to short-time inverse Fourier transform, the noise-reduced seismic wave signal collected by the DAS system can be obtained.

Images