Optical fiber sensing waterfall diagram data noise reduction method, device and processor

Abstract

The embodiment of the present application provides a method, device and processor for denoising data of a fiber optic sensor waterfall chart. The method includes: obtaining the original data of the fiber optic sensor waterfall chart, wherein the original data includes background noise data and intrusion data; inputting the original data into a denoising autoencoder to output predicted data of the original data through the denoising autoencoder; inputting the predicted data into an intrusion classifier to output first data and second data corresponding to the predicted data through the intrusion classifier; determining a first error based on the first data and the first preset data; determining a second error based on the second data and the second preset data; and determining that the original data denoising is completed when the first error is less than the first preset error and the second error is less than the second preset error. The above technical solution can perform denoising more effectively and improve the denoising effect on the sensor waterfall chart data.

Description

Technical Field

[0001] The present application relates to the field of data processing technology, and in particular to a method, device, storage medium and processor for reducing noise in optical fiber sensing waterfall diagram data. Background Art

[0002] In order to remove the noise data from the noisy data, the prior art usually adopts the method of subtracting the noise spectrum or reconstructing the noisy data by self-encoding. However, the prior art subtracts the noise signal spectrum from the original signal spectrum to obtain the denoised data, which will cause a certain amount of information loss. And it is impossible to obtain a clean and noise-free data sample. In addition, it is also impossible to guarantee that the data corresponding to all its signals are valid, so it is of low practicality and cannot achieve a good noise reduction effect. Summary of the invention

[0003] The purpose of the embodiments of the present application is to provide a method, device, storage medium and processor for reducing noise in fiber optic sensing waterfall diagram data.

[0004] In order to achieve the above-mentioned object, the present application provides a data noise reduction method in a first aspect, comprising:

[0005] Obtaining raw data of the optical fiber sensing waterfall diagram, wherein the raw data includes background noise data and intrusion data;

[0006] Inputting the original data into the denoising autoencoder to output predicted data of the original data through the denoising autoencoder;

[0007] Inputting the predicted data into the intrusion classifier so that the intrusion classifier outputs first data and second data corresponding to the predicted data;

[0008] determining a first error according to the first data and the first preset data;

[0009] determining a second error according to the second data and the second preset data;

[0010] When the first error is smaller than the first preset error and the second error is smaller than the second preset error, it is determined that the original data denoising is completed.

[0011] In an embodiment of the present application, the method also includes: when the first error is greater than or equal to the first preset error, and / or the second error is greater than or equal to the second preset error, again executing the step of inputting the original data into the denoising autoencoder to output predicted data of the original data through the denoising autoencoder, until it is determined that the denoising of the original data is completed.

[0012] In an embodiment of the present application, determining the first error according to the first data and the first preset data includes: inputting the first data and the first preset data into a binary cross entropy loss function to output the first error through the binary cross entropy loss function.

[0013] In an embodiment of the present application, determining the second error based on the second data and the second preset data includes: inputting the second data and the second preset data into a binary cross entropy loss function to output the second error through the binary cross entropy loss function.

[0014] In an embodiment of the present application, the method also includes: after inputting the original data into the denoising autoencoder to output the predicted data of the original data through the denoising autoencoder, determining the target intrusion data according to the intrusion data; determining the target noise data according to the background noise data; determining the first prediction error of the denoising autoencoder according to the predicted data and the target intrusion data; and determining the second prediction error of the denoising autoencoder according to the predicted data and the target noise data.

[0015] In an embodiment of the present application, determining that the noise reduction of the original data is completed includes: when the first error is less than the first preset error and the second error is less than the second preset error, judging whether the first prediction error is less than the third preset error and whether the second prediction error is less than the fourth preset error; when the first prediction error is less than the third preset error and the second prediction error is less than the fourth preset error, determining that the noise reduction of the original data is completed.

[0016] In an embodiment of the present application, determining target intrusion data according to the intrusion data includes: processing the intrusion data by spectral subtraction to obtain the target intrusion data.

[0017] A second aspect of the present application provides a processor configured to execute the above-mentioned fiber optic sensing waterfall diagram data denoising method.

[0018] A third aspect of the present application provides a fiber optic sensing waterfall diagram data noise reduction device, comprising the above-mentioned processor.

[0019] A fourth aspect of the present application provides a machine-readable storage medium having instructions stored thereon, which, when executed by a processor, configures the processor to execute the above-mentioned fiber optic sensing waterfall diagram data denoising method.

[0020] The above technical solution obtains the original data of the fiber optic sensor waterfall diagram, wherein the original data includes background noise data and intrusion data; inputs the original data into a denoising autoencoder to output predicted data of the original data through the denoising autoencoder; inputs the predicted data into an intrusion classifier to output first data and second data corresponding to the predicted data through the intrusion classifier; determines a first error according to the first data and the first preset data; determines a second error according to the second data and the second preset data; and determines that the original data denoising is completed when the first error is less than the first preset error and the second error is less than the second preset error. The above technical solution can perform denoising more effectively and improve the denoising effect on the sensor waterfall diagram data.

[0021] Other features and advantages of the embodiments of the present application will be described in detail in the subsequent specific implementation section. BRIEF DESCRIPTION OF THE DRAWINGS

[0022] The accompanying drawings are used to provide a further understanding of the embodiments of the present application and constitute a part of the specification. Together with the following specific implementations, they are used to explain the embodiments of the present application, but do not constitute a limitation on the embodiments of the present application. In the accompanying drawings:

[0023] Figure 1 The flowchart of the optical fiber sensing waterfall diagram data noise reduction method according to the embodiment of the present application is schematically shown;

[0024] Figure 2 A schematic diagram of a method for reducing noise in a fiber optic sensing waterfall diagram data according to an embodiment of the present application is schematically shown;

[0025] Figure 3 The internal structure diagram of a computer device according to an embodiment of the present application is schematically shown. DETAILED DESCRIPTION

[0026] In order to make the purpose, technical scheme and advantages of the embodiments of the present application clearer, the technical scheme in the embodiments of the present application will be clearly and completely described below in conjunction with the drawings in the embodiments of the present application. It should be understood that the specific implementation methods described herein are only used to illustrate and explain the embodiments of the present application, and are not used to limit the embodiments of the present application. Based on the embodiments in the present application, all other embodiments obtained by ordinary technicians in this field without making creative work are within the scope of protection of this application.

[0027] Figure 1 The flowchart of the optical fiber sensing waterfall diagram data noise reduction method according to the embodiment of the present application is schematically shown. Figure 1 As shown, in one embodiment of the present application, a method for reducing noise of optical fiber sensing waterfall diagram data is provided, comprising the following steps:

[0028] Step 101, obtaining raw data of the optical fiber sensing waterfall diagram, wherein the raw data includes background noise data and intrusion data.

[0029] Step 102: inputting the original data into a denoising autoencoder so as to output predicted data of the original data through the denoising autoencoder.

[0030] Step 103: input the predicted data into an intrusion classifier so that the intrusion classifier outputs first data and second data corresponding to the predicted data.

[0031] Step 104: determine a first error according to the first data and the first preset data.

[0032] Step 105: determine a second error according to the second data and the second preset data.

[0033] Step 106 , when the first error is less than the first preset error and the second error is less than the second preset error, it is determined that the original data denoising is completed.

[0034] The fiber optic sensor waterfall diagram is a three-dimensional spectrum diagram that includes information such as noise and intrusion vibration. Background noise data refers to all interference data in the occurrence, inspection, measurement or recording of optical fiber. Intrusion data refers to the vibration data generated when the optical fiber is damaged externally. The denoising autoencoder is able to process the input raw data containing noise to output raw data without noise. The intrusion classifier is a binary model composed of a deep neural network, which is a model that can classify input data.

[0035] The processor may obtain the raw data of the optical fiber sensing waterfall diagram, wherein the raw data may include background noise data and intrusion data. After obtaining the raw data of the optical fiber sensing waterfall diagram, the processor may input the raw data into the denoising autoencoder to output the predicted data of the raw data through the denoising autoencoder. Wherein, the predicted data may include intrusion data with high purity and / or normal data with approximately noise-free. After the denoising autoencoder outputs the predicted data, the processor may input the predicted data into the intrusion classifier to output the first data and the second data corresponding to the predicted data through the intrusion classifier. Wherein, the first data may be intrusion data with high purity included in the predicted data, and the second data may be normal data with approximately noise-free included in the predicted data. After the intrusion classifier outputs the first data and the second data, the processor may determine the first error according to the first data and the first preset data. Wherein, the first preset data may be completely pure intrusion data. That is, after processing, the first error may be determined according to the intrusion data with high purity and the completely pure intrusion data. The processor may also determine the second error according to the second data and the second preset data. Wherein, the second preset data may be normal data with no noise. That is, the processor can determine the second error based on the normal data that is approximately noise-free and the normal data that is noise-free. After determining the first error and the second error, the processor can determine whether the first error is less than the first preset error and whether the second error is less than the second preset error. When the first error is less than the first preset error and the second error is less than the second preset error, the processor can determine that the original data denoising is completed.

[0036] In one embodiment, the method further includes: when the first error is greater than or equal to the first preset error, and / or the second error is greater than or equal to the second preset error, again executing the step of inputting the original data into the denoising autoencoder to output predicted data of the original data through the denoising autoencoder, until it is determined that the denoising of the original data is completed.

[0037] The processor may obtain the raw data of the optical fiber sensing waterfall diagram, wherein the raw data may include background noise data and intrusion data. After obtaining the raw data of the optical fiber sensing waterfall diagram, the processor may input the raw data into the denoising autoencoder to output the predicted data of the raw data through the denoising autoencoder. Wherein, the predicted data may include intrusion data with high purity and / or normal data with approximately noise-free. After the denoising autoencoder outputs the predicted data, the processor may input the predicted data into the intrusion classifier to output the first data and the second data corresponding to the predicted data through the intrusion classifier. Wherein, the first data may be intrusion data with high purity included in the predicted data, and the second data may be normal data with approximately noise-free included in the predicted data. After the intrusion classifier outputs the first data and the second data, the processor may determine the first error according to the first data and the first preset data. Wherein, the first preset data may be completely pure intrusion data. That is, after processing, the first error may be determined according to the intrusion data with high purity and the completely pure intrusion data. The processor may also determine the second error according to the second data and the second preset data. Wherein, the second preset data may be normal data with no noise. That is, the processor can determine the second error based on the normal data that is approximately noise-free and the normal data that is noise-free. After determining the first error and the second error, the processor can determine whether the first error is less than the first preset error, and whether the second error is less than the second preset error. In the case where the first error is greater than or equal to the first preset error, and / or the second error is greater than or equal to the second preset error, the processor can again perform the step of inputting the original data into the denoising autoencoder to output the predicted data of the original data through the denoising autoencoder until it is determined that the denoising of the original data is completed.

[0038] In one embodiment, determining the first error according to the first data and the first preset data includes: inputting the first data and the first preset data into a binary cross entropy loss function to output the first error through the binary cross entropy loss function. The binary cross entropy loss function is a commonly used loss function in machine learning and can be used to measure the distance between two distributions.

[0039] The processor can use the predicted data output by the denoising autoencoder as the input of the intrusion classifier. After the predicted data is input into the intrusion classification classifier, the processor can control the intrusion classifier to output the first data and the second data corresponding to the predicted data. Among them, the first data can be the intrusion data with high purity included in the predicted data, and the second data can be the normal data with approximately noise-free included in the predicted data. After the intrusion classifier outputs the first data and the second data, the processor can determine the first error according to the first data and the first preset data. Among them, the first preset data can be completely pure intrusion data. Specifically, the processor can input the first data and the first preset data into a binary cross entropy loss function to output the first error through the binary cross entropy loss function. That is, the processor can input the intrusion data with high purity and the completely pure intrusion data into the binary cross entropy loss function to output the first error through the binary cross entropy loss function.

[0040] In one embodiment, determining the second error according to the second data and the second preset data includes: inputting the second data and the second preset data into a binary cross entropy loss function to output the second error through the binary cross entropy loss function.

[0041] The processor can use the predicted data output by the denoising autoencoder as the input of the intrusion classifier. After the predicted data is input into the intrusion classification classifier, the processor can control the intrusion classifier to output the first data and the second data corresponding to the predicted data. Among them, the first data can be the intrusion data with high purity included in the predicted data, and the second data can be the normal data with approximately noise-free included in the predicted data. After the intrusion classifier outputs the first data and the second data, the processor can determine the second error according to the second data and the second preset data. Among them, the second preset data can be the normal data with no noise. Specifically, the processor can input the second data and the second preset data into the binary cross entropy loss function to output the second error through the binary cross entropy loss function. That is, the processor can input the normal data with approximately noise-free and the normal data with no noise into the binary cross entropy loss function to output the second error through the binary cross entropy loss function.

[0042] In one embodiment, the method also includes: after inputting the original data into the denoising autoencoder to output the predicted data of the original data through the denoising autoencoder, determining the target intrusion data according to the intrusion data; determining the target noise data according to the background noise data; determining the first prediction error of the denoising autoencoder according to the predicted data and the target intrusion data; and determining the second prediction error of the denoising autoencoder according to the predicted data and the target noise data.

[0043] After the original data is input into the noise reduction autoencoder to output the predicted data of the original data through the noise reduction autoencoder, the processor can determine the target intrusion data according to the intrusion data. In one embodiment, the processor can process the intrusion data by spectral subtraction to obtain the target intrusion data. Among them, spectral subtraction is to subtract the spectrum of the noise from the spectrum of the noisy intrusion data to obtain the noise-free target intrusion data. The processor can also determine the target noise data according to the background noise data. After determining the target intrusion data, the processor can determine the first prediction error of the noise reduction autoencoder according to the predicted data and the target intrusion data. For example, the processor can use the mean square loss function to process the predicted data and the target intrusion data, and then determine the first prediction error of the noise reduction autoencoder. After determining the target noise data, the processor can determine the second prediction error of the noise reduction autoencoder according to the predicted data and the target noise data. For example, the processor can use the mean square loss function to process the predicted data and the target background noise data, and then determine the second prediction error of the noise reduction autoencoder.

[0044] In one embodiment, determining that the noise reduction of the original data is completed includes: when the first error is less than the first preset error and the second error is less than the second preset error, judging whether the first prediction error is less than the third preset error and whether the second prediction error is less than the fourth preset error; when the first prediction error is less than the third preset error and the second prediction error is less than the fourth preset error, determining that the noise reduction of the original data is completed.

[0045] When the first error is less than the first preset error and the second error is less than the second preset error, the processor can determine whether the first prediction error is less than the third preset error and whether the second prediction error is less than the fourth preset error. When the first prediction error is less than the third preset error and the second prediction error is less than the fourth preset error, the processor can determine that the original data denoising is completed.

[0046] In one embodiment, Figure 2 As shown, the processor can input the training data into the denoising autoencoder to output the denoised training data through the denoising autoencoder. After the denoising autoencoder outputs the denoised training data, the processor can input the denoised training data into the intrusion classifier G π , to pass the intrusion classifier G πThe output corresponds to the intrusion data and normal data of the denoised training data. After obtaining the intrusion data and the normal data, the processor can process the intrusion data and the normal data respectively through the Crossentropy function (binary cross entropy loss function) to obtain a first error corresponding to the intrusion data and a second error corresponding to the normal data. When the first error is less than the first preset error and the second error is less than the second preset error, the processor can determine that the denoising of the training data is completed. When the first error is greater than or equal to the first preset error, and / or the second error is greater than or equal to the second preset error, the processor can perform gradient back propagation to adjust the model parameters of the denoising autoencoder, the intrusion classifier G π The model parameters are then input into the denoising autoencoder until the denoising of the training data is completed.

[0047] The processor can also obtain the intrusion signal and background noise in the training data. The intrusion signal is processed by spectral subtraction to obtain an intrusion label corresponding to the intrusion signal. Among them, the intrusion label refers to a pure intrusion signal. The processor can perform background noise label all-zero processing on the background noise, that is, set the background noise to zero to obtain noise-free normal data. The processor can input the training data into the denoising autoencoder to output the denoised training data through the denoising autoencoder. On the one hand, the processor can process the denoised training data and the intrusion label, and the denoised training data and the noise-free normal data respectively through the Mse loss function (mean square loss function) to obtain the first prediction error and the second prediction error of the denoising autoencoder.

[0048] On the other hand, the processor can input the denoised training data into the intrusion classifier G π , to pass the intrusion classifier G π Output the intrusion data and normal data corresponding to the denoised training data. The processor can process the intrusion data and normal data respectively through the Crossentropy function to obtain the first error corresponding to the intrusion data and the second error corresponding to the normal data respectively. When the first prediction error is greater than or equal to the third preset error, and / or the second prediction error is greater than or equal to the fourth preset error, the processor can perform gradient back propagation to adjust the model parameters of the denoising autoencoder, and input the training data to the denoising autoencoder again until the first prediction error is less than the third preset error and the second prediction error is less than the fourth preset error. When the first error is the first preset error, the second error is less than the second preset error, the first prediction error is less than the third preset error, and the second prediction error is less than the fourth preset error, it is determined that the training data denoising is completed.

[0049] The above technical solution obtains the original data of the fiber optic sensor waterfall diagram, wherein the original data includes background noise data and intrusion data; inputs the original data into a denoising autoencoder to output predicted data of the original data through the denoising autoencoder; inputs the predicted data into an intrusion classifier to output first data and second data corresponding to the predicted data through the intrusion classifier; determines a first error according to the first data and the first preset data; determines a second error according to the second data and the second preset data; and determines that the original data denoising is completed when the first error is less than the first preset error and the second error is less than the second preset error. The above technical solution can perform denoising more effectively and improve the denoising effect on the sensor waterfall diagram data.

[0050] Figure 1 FIG. 1 is a flow chart of a method for reducing noise in a fiber optic sensor waterfall diagram in one embodiment. It should be understood that although Figure 1 The steps in the flowchart are shown in sequence as indicated by the arrows, but these steps are not necessarily executed in the order indicated by the arrows. Unless otherwise specified in this document, there is no strict order restriction for the execution of these steps, and these steps can be executed in other orders. Moreover, Figure 1 At least part of the steps may include multiple sub-steps or multiple stages. These sub-steps or stages are not necessarily executed at the same time, but can be executed at different times. The execution order of these sub-steps or stages is not necessarily sequential, but can be executed in turn or alternately with other steps or at least part of the sub-steps or stages of other steps.

[0051] An embodiment of the present application provides a processor, which is used to run a program, wherein the above-mentioned fiber optic sensor waterfall chart data denoising method is executed when the program is running.

[0052] An embodiment of the present application provides a fiber optic sensor waterfall diagram data noise reduction device, including the above-mentioned processor.

[0053] An embodiment of the present application provides a storage medium on which a program is stored. When the program is executed by a processor, the above-mentioned fiber optic sensor waterfall diagram data noise reduction method is implemented.

[0054] In one embodiment, a computer device is provided. The computer device may be a server, and its internal structure diagram may be as follows: Figure 3As shown. The computer device includes a processor A01, a network interface A02, a memory (not shown in the figure) and a database (not shown in the figure) connected through a system bus. Among them, the processor A01 of the computer device is used to provide computing and control capabilities. The memory of the computer device includes an internal memory A03 and a non-volatile storage medium A04. The non-volatile storage medium A04 stores an operating system B01, a computer program B02 and a database (not shown in the figure). The internal memory A03 provides an environment for the operation of the operating system B01 and the computer program B02 in the non-volatile storage medium A04. The database of the computer device is used to store the original data of the optical fiber sensing waterfall diagram. The network interface A02 of the computer device is used to communicate with an external terminal through a network connection. When the computer program B02 is executed by the processor A01, a method for reducing noise in optical fiber sensing waterfall diagram data is implemented.

[0055] Those skilled in the art will understand that Figure 3 The structure shown in the figure is only a block diagram of a part of the structure related to the solution of the present application, and does not constitute a limitation on the computer device to which the solution of the present application is applied. The specific computer device may include more or fewer components than those shown in the figure, or combine certain components, or have a different arrangement of components.

[0056] An embodiment of the present application provides a device, which includes a processor, a memory, and a program stored in the memory and executable on the processor. When the processor executes the program, the following steps are implemented: obtaining raw data of a fiber optic sensor waterfall diagram, wherein the raw data includes background noise data and intrusion data; inputting the raw data into a denoising autoencoder to output predicted data of the raw data through the denoising autoencoder; inputting the predicted data into an intrusion classifier to output first data and second data corresponding to the predicted data through the intrusion classifier; determining a first error based on the first data and the first preset data; determining a second error based on the second data and the second preset data; and determining that denoising of the raw data is completed when the first error is less than the first preset error and the second error is less than the second preset error.

[0057] In one embodiment, the method further includes: when the first error is greater than or equal to the first preset error, and / or the second error is greater than or equal to the second preset error, again executing the step of inputting the original data into the denoising autoencoder to output predicted data of the original data through the denoising autoencoder, until it is determined that the denoising of the original data is completed.

[0058] In one embodiment, determining the first error according to the first data and the first preset data includes: inputting the first data and the first preset data into a binary cross entropy loss function to output the first error through the binary cross entropy loss function.

[0059] In one embodiment, determining the second error according to the second data and the second preset data includes: inputting the second data and the second preset data into a binary cross entropy loss function to output the second error through the binary cross entropy loss function.

[0060] In one embodiment, the method also includes: after inputting the original data into the denoising autoencoder to output the predicted data of the original data through the denoising autoencoder, determining the target intrusion data according to the intrusion data; determining the target noise data according to the background noise data; determining the first prediction error of the denoising autoencoder according to the predicted data and the target intrusion data; and determining the second prediction error of the denoising autoencoder according to the predicted data and the target noise data.

[0061] In one embodiment, determining that the noise reduction of the original data is completed includes: when the first error is less than the first preset error and the second error is less than the second preset error, judging whether the first prediction error is less than the third preset error and whether the second prediction error is less than the fourth preset error; when the first prediction error is less than the third preset error and the second prediction error is less than the fourth preset error, determining that the noise reduction of the original data is completed.

[0062] In one embodiment, determining target intrusion data according to the intrusion data includes: processing the intrusion data by spectral subtraction to obtain the target intrusion data.

[0063] The present application also provides a computer program product, which, when executed on a data processing device, is suitable for executing a program for initializing steps of a method for reducing noise in fiber optic sensor waterfall diagram data.

[0064] Those skilled in the art will appreciate that the embodiments of the present application may be provided as methods, systems, or computer program products. Therefore, the present application may adopt the form of a complete hardware embodiment, a complete software embodiment, or an embodiment in combination with software and hardware. Moreover, the present application may adopt the form of a computer program product implemented in one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) that include computer-usable program code.

[0065] The present application is described with reference to the flowcharts and / or block diagrams of the methods, devices (systems), and computer program products according to the embodiments of the present application. It should be understood that each process and / or box in the flowchart and / or block diagram, as well as the combination of the processes and / or boxes in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing device to generate a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the processes in the flowchart and / or block diagram. Figure 1A process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.

[0066] These computer program instructions may also be stored in a computer-readable memory capable of directing a computer or other programmable data processing device to operate in a specific manner, so that the instructions stored in the computer-readable memory produce an article of manufacture comprising an instruction device, which implements the process Figure 1 A process or multiple processes and / or boxes Figure 1 A function specified in one or more boxes.

[0067] These computer program instructions can also be loaded onto a computer or other programmable data processing device so that a series of operating steps are executed on the computer or other programmable device to produce a computer-implemented process, thereby providing instructions for implementing the process. Figure 1 A process or multiple processes and / or boxes Figure 1 A step that specifies a function in one or more boxes.

[0068] In a typical configuration, a computing device includes one or more processors (CPU), input / output interfaces, network interfaces, and memory.

[0069] The memory may include non-permanent memory in a computer-readable medium, random access memory (RAM) and / or non-volatile memory in the form of read-only memory (ROM) or flash RAM. The memory is an example of a computer-readable medium.

[0070] Computer readable media include permanent and non-permanent, removable and non-removable media that can be implemented by any method or technology to store information. Information can be computer readable instructions, data structures, program modules or other data. Examples of computer storage media include, but are not limited to, phase change memory (PRAM), static random access memory (SRAM), dynamic random access memory (DRAM), other types of random access memory (RAM), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), flash memory or other memory technology, compact disk read-only memory (CD-ROM), digital versatile disk (DVD) or other optical storage, magnetic cassettes, magnetic tape magnetic disk storage or other magnetic storage devices or any other non-transmission media that can be used to store information that can be accessed by a computing device. As defined herein, computer readable media does not include temporary computer readable media (transitory media), such as modulated data signals and carrier waves.

[0071] It should also be noted that the terms "include", "comprises" or any other variations thereof are intended to cover non-exclusive inclusion, so that a process, method, commodity or device including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, commodity or device. In the absence of more restrictions, the elements defined by the sentence "comprises a ..." do not exclude the existence of other identical elements in the process, method, commodity or device including the elements.

[0072] The above are only embodiments of the present application and are not intended to limit the present application. For those skilled in the art, the present application may have various changes and variations. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present application should be included within the scope of the claims of the present application.

Claims

1. A method for reducing noise in optical fiber sensing waterfall diagram data, characterized in that: The method comprises: Acquire raw data of the optical fiber sensing waterfall diagram, wherein the raw data includes background noise data and intrusion data; Inputting the original data into a denoising autoencoder to output predicted data of the original data through the denoising autoencoder; Inputting the predicted data into an intrusion classifier, so that the intrusion classifier outputs first data and second data corresponding to the predicted data, wherein the first data is intrusion data with high purity included in the predicted data, and the second data is normal data with approximately no noise included in the predicted data; Determining a first error according to the first data and first preset data, wherein the first preset data is completely pure intrusion data; determining a second error according to the second data and second preset data, wherein the second preset data is normal data without noise; When the first error is smaller than a first preset error and the second error is smaller than a second preset error, determining that the original data denoising is completed; Wherein, determining the first error according to the first data and the first preset data comprises: Inputting the first data and the first preset data into a binary cross entropy loss function to output the first error through the binary cross entropy loss function; Wherein, determining the second error according to the second data and the second preset data comprises: The second data and the second preset data are input into a binary cross entropy loss function to output the second error through the binary cross entropy loss function.

2. The fiber optic sensing waterfall data denoising method according to claim 1, characterized in that: The method further comprises: When the first error is greater than or equal to the first preset error, and / or the second error is greater than or equal to the second preset error, the step of inputting the original data into the denoising autoencoder to output predicted data of the original data through the denoising autoencoder is performed again until it is determined that the denoising of the original data is completed.

3. The fiber optic sensing waterfall data denoising method according to claim 1, characterized in that: The method further comprises: After the original data is input into the denoising autoencoder so as to output the predicted data of the original data through the denoising autoencoder, target intrusion data is determined according to the intrusion data; Determining target noise data according to the background noise data; Determine a first prediction error of the denoising autoencoder according to the prediction data and the target intrusion data; A second prediction error of the denoising autoencoder is determined based on the prediction data and the target noise data.

4. The optical fiber sensing waterfall data denoising method according to claim 3, characterized in that: Determining that the original data denoising is completed includes: When the first error is smaller than the first preset error and the second error is smaller than the second preset error, determining whether the first prediction error is smaller than a third preset error and whether the second prediction error is smaller than a fourth preset error; When the first prediction error is smaller than the third preset error and the second prediction error is smaller than the fourth preset error, it is determined that the original data denoising is completed.

5. The optical fiber sensing waterfall diagram data denoising method according to claim 3, characterized in that: Determining target intrusion data according to the intrusion data includes: The intrusion data is processed by spectral subtraction to obtain the target intrusion data.

6. A processor, characterized in that: The method is configured to execute the fiber optic sensing waterfall diagram data noise reduction method according to any one of claims 1 to 5.

7. A fiber optic sensing waterfall data noise reduction device, characterized in that: The apparatus comprises a processor according to claim 6.

8. A machine-readable storage medium having instructions stored thereon, characterized in that: When the instruction is executed by a processor, the processor is configured to execute the fiber optic sensing waterfall diagram data noise reduction method according to any one of claims 1 to 5.

Images