Data Processing Method and Storage Medium

Abstract

The present invention discloses a data processing method and a storage medium. Wherein, the method includes: obtaining a first target image and a second target image, where the first target image is used to represent the actual result obtained by detecting the weather before the target time period, and the second target image is used to represent the prediction result obtained by predicting the weather during the target time period; respectively converting the first target image and the second target image from the spatial domain to the frequency domain; reconstructing the second target image in the frequency domain based on the first target image in the frequency domain to obtain a third target image; converting the third target image from the frequency domain to the spatial domain to obtain a fourth target image. The present invention solves the technical problem of poor visualization effect of images.

Description

Technical Field

[0001] The present invention relates to the field of computers, and in particular, to a data processing method and a storage medium. Background Art

[0002] Currently, for extrapolated images, as the extrapolation time increases, the local pulsation intensity of the extrapolated images gradually weakens, and the situation of visual blurring of the images will be faced. For example, for radar echo extrapolation, as the prediction time increases, the conditional probabilities will continuously stack up, and the reflected output result is that the output image lacks details and the visual presentation is relatively blurred, thus there is a technical problem of poor visualization effect of the images.

[0003] Aiming at the above technical problem of poor visualization effect of the images, no effective solution has been proposed yet. Summary of the Invention

[0004] Embodiments of the present invention provide a data processing method and a storage medium to at least solve the technical problem of poor visualization effect of the images.

[0005] According to one aspect of the embodiments of the present invention, a data processing method is provided. The method may include: obtaining a first target image and a second target image, where the first target image is used to represent the actual result obtained by detecting the weather before a target time period, and the second target image is used to represent the prediction result obtained by predicting the weather during the target time period; respectively converting the first target image and the second target image from the spatial domain to the frequency domain; reconstructing the second target image in the frequency domain based on the first target image in the frequency domain to obtain a third target image; and converting the third target image from the frequency domain to the spatial domain to obtain a fourth target image.

[0006] According to one aspect of the embodiments of the present invention, another data processing method is further provided. The method may include: responding to an input operation instruction acting on an operation interface, and inputting a first target image and a second target image, where the first target image is used to represent the actual result obtained by detecting the weather before a target time period, and the second target image is used to represent the prediction result obtained by predicting the weather during the target time period; responding to a reconstruction operation instruction acting on the operation interface, and displaying a fourth target image on the operation interface, where the fourth target image is obtained by converting a third target image from the frequency domain to the spatial domain, and the third target image is obtained by reconstructing the second target image converted from the spatial domain to the frequency domain based on the first target image converted from the spatial domain to the frequency domain.

[0007] According to one aspect of an embodiment of the present invention, another data processing method is also provided. The method may include: obtaining a first target image and a second target image from a weather broadcast platform, and displaying the first target image and the second target image on an operation interface, where the first target image is used to represent the actual result obtained by detecting the weather before a target time period, and the second target image is used to represent the prediction result obtained by predicting the weather during the target time period; in response to a reconstruction operation instruction acting on the operation interface, displaying a fourth target image on the operation interface, where the fourth target image is obtained by converting a third target image from the frequency domain to the spatial domain, and the third target image is obtained by reconstructing the second target image converted from the spatial domain to the frequency domain based on the first target image converted from the spatial domain to the frequency domain; and returning the fourth target image to the weather broadcast platform.

[0008] According to one aspect of an embodiment of the present invention, another data processing method is also provided. The method may include: determining a target area where a target vegetation is located; obtaining a first target image and a second target image corresponding to the target area, where the first target image is used to represent the actual result obtained by detecting the weather in the target area before a target time period, and the second target image is used to represent the precipitation prediction result obtained by predicting the weather in the target area during the target time period; respectively converting the first target image and the second target image from the spatial domain to the frequency domain; reconstructing the second target image in the frequency domain based on the first target image in the frequency domain to obtain a third target image; converting the third target image from the frequency domain to the spatial domain to obtain a fourth target image; and determining whether to fertilize the target vegetation based on the fourth target image.

[0009] According to one aspect of an embodiment of the present invention, a data processing device is also provided. The device may include: a first acquisition unit, configured to acquire a first target image and a second target image, where the first target image is used to represent the actual result obtained by detecting the weather before a target time period, and the second target image is used to represent the prediction result obtained by predicting the weather during the target time period; a first conversion unit, configured to respectively convert the first target image and the second target image from the spatial domain to the frequency domain; a first reconstruction unit, configured to reconstruct the second target image in the frequency domain based on the first target image in the frequency domain to obtain a third target image; and a second conversion unit, configured to convert the third target image from the frequency domain to the spatial domain to obtain a fourth target image.

[0010] According to one aspect of an embodiment of the present invention, another data processing device is also provided. The device may include: an input unit, configured to input a first target image and a second target image in response to an input operation instruction acting on an operation interface, where the first target image is used to represent an actual result obtained by detecting the weather before a target time period, and the second target image is used to represent a prediction result obtained by predicting the weather during the target time period; a first display unit, configured to display a fourth target image on the operation interface in response to a reconstruction operation instruction acting on the operation interface, where the fourth target image is obtained by converting a third target image from the frequency domain to the spatial domain, and the third target image is obtained by reconstructing the second target image converted from the spatial domain to the frequency domain based on the first target image converted from the spatial domain to the frequency domain.

[0011] According to one aspect of an embodiment of the present invention, another data processing device is also provided. The device may include: a second acquisition unit, configured to acquire a first target image and a second target image from a weather broadcast platform and display the first target image and the second target image on an operation interface, where the first target image is used to represent an actual result obtained by detecting the weather before a target time period, and the second target image is used to represent a prediction result obtained by predicting the weather during the target time period; a second display unit, configured to display a fourth target image on the operation interface in response to a reconstruction operation instruction acting on the operation interface, where the fourth target image is obtained by converting a third target image from the frequency domain to the spatial domain, and the third target image is obtained by reconstructing the second target image converted from the spatial domain to the frequency domain based on the first target image converted from the spatial domain to the frequency domain; a return unit, configured to return the fourth target image to the weather broadcast platform.

[0012] According to one aspect of an embodiment of the present invention, another data processing device is also provided. The device may include: a first determination unit, configured to determine a target area where a target vegetation is located; a third acquisition unit, configured to acquire a first target image and a second target image corresponding to the target area, where the first target image is used to represent an actual result obtained by detecting the weather in the target area before a target time period, and the second target image is used to represent a precipitation prediction result obtained by predicting the weather in the target area during the target time period; a third conversion unit, configured to respectively convert the first target image and the second target image from the spatial domain to the frequency domain; a second reconstruction unit, configured to reconstruct the second target image in the frequency domain based on the first target image in the frequency domain to obtain a third target image; a fourth conversion unit, configured to convert the third target image from the frequency domain to the spatial domain to obtain a fourth target image; a second determination unit, configured to determine whether to fertilize the target vegetation based on the fourth target image.

[0013] An embodiment of the present invention also provides a computer-readable storage medium. The computer-readable storage medium includes a stored program, wherein when the program is run by a processor, it controls the device where the computer-readable storage medium is located to execute the data processing method of the embodiment of the present invention.

[0014] An embodiment of the present invention also provides a processor. The processor is used to run a program, wherein when the program runs, it executes the data processing method of the embodiment of the present invention.

[0015] An embodiment of the present invention also provides a data processing system. The system may include: a processor; a memory, connected to the processor, for providing instructions for the processor to perform the following processing steps: obtaining a first target image and a second target image, wherein the first target image is used to represent the actual result obtained by detecting the weather before a target time period, and the second target image is used to represent the prediction result obtained by predicting the weather during the target time period; respectively converting the first target image and the second target image from the spatial domain to the frequency domain; reconstructing the second target image in the frequency domain based on the first target image in the frequency domain to obtain a third target image; converting the third target image from the frequency domain to the spatial domain to obtain a fourth target image.

[0016] In the embodiment of the present invention, in the frequency domain space, based on the power spectrum of the style image, the radar echo extrapolation image is reconstructed. The overall echo form of the finally reconstructed image can be consistent with the radar echo extrapolation image, and the small-scale detail information of the style image is supplemented, improving the visual effect of the image and solving the technical problem of poor visualization effect of the image. BRIEF DESCRIPTION OF THE DRAWINGS

[0017] The drawings described herein are used to provide a further understanding of the present invention, form a part of this application, and the schematic embodiments of the present invention and their descriptions are used to explain the present invention and do not constitute an improper limitation to the present invention. In the drawings:

[0018] Figure 1 is a hardware structure block diagram of a computer terminal (or mobile device) for a data processing method according to an embodiment of the present invention;

[0019] Figure 2 is a flowchart of a data processing method according to an embodiment of the present invention;

[0020] Figure 3 is a flowchart of another data processing method according to an embodiment of the present invention;

[0021] Figure 4 is a flowchart of another data processing method according to an embodiment of the present invention;

[0022] Figure 5It is a flowchart of another data processing method according to an embodiment of the present invention;

[0023] Figure 6A It is a schematic diagram of a style image according to an embodiment of the present invention;

[0024] Figure 6B It is a schematic diagram of a content image according to an embodiment of the present invention;

[0025] Figure 6C It is a schematic diagram of the visualization effect of radar echo extrapolation and reconstruction according to an embodiment of the present invention;

[0026] Figure 7 It is a schematic diagram of a power spectrum according to an embodiment of the present invention;

[0027] Figure 8A It is a schematic diagram of a data processing scenario according to an embodiment of the present invention;

[0028] Figure 8B It is a schematic diagram of another data processing scenario according to an embodiment of the present invention;

[0029] Figure 9 It is a schematic diagram of a data processing device according to an embodiment of the present invention;

[0030] Figure 10 It is a schematic diagram of another data processing device according to an embodiment of the present invention;

[0031] Figure 11 It is a schematic diagram of another data processing device according to an embodiment of the present invention;

[0032] Figure 12 It is a schematic diagram of another data processing device according to an embodiment of the present invention;

[0033] Figure 13 It is a block diagram of the structure of a computer terminal according to an embodiment of the present invention. Detailed implementation manners

[0034] In order to enable those skilled in the art to better understand the solution of the present invention, the technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative efforts shall fall within the protection scope of the present invention.

[0035] It should be noted that the terms "first", "second", etc. in the description, claims and above-mentioned drawings of the present invention are used to distinguish similar objects, and do not necessarily need to be used to describe a specific order or sequence. It should be understood that the data used in this way can be interchanged under appropriate circumstances, so that the embodiments of the present invention described here can be implemented in an order other than those illustrated or described here. In addition, the terms "comprising" and "having" and any variations thereof are intended to cover non-exclusive inclusion. For example, a process, method, system, product or device comprising a series of steps or units does not necessarily have to be limited to those steps or units clearly listed, but may include other steps or units not clearly listed or inherent to these processes, methods, products or devices.

[0036] First, some nouns or terms that appear in the process of describing the embodiments of the present application are applicable to the following explanations:

[0037] Fourier transform, which converts a signal from the spatial domain to the frequency domain. For a two-dimensional spatial image, the signal after Fourier transform is a superposition of simple harmonic signals with different frequencies in the two-dimensional space. The transformed signal contains amplitude components and frequency components at each frequency. Fourier transform can be applied to the fields of engineering, science and mathematics;

[0038] Power spectrum, which is the abbreviation of the power spectral density function, is the probability density distribution of signal power in a frequency band, reflecting the distribution of signal power at different frequencies, and can be used for statistical signal processing;

[0039] Radar echo extrapolation refers to determining the intensity distribution of the echo, as well as the moving speed and direction of the echo body, based on the echo data detected by a weather radar, and forecasting the state of the radar echo after a period of time by performing time extrapolation on the echo signal. It can be used in scenarios such as severe convective weather analysis and short-term adjacent precipitation forecasting;

[0040] Radar extrapolation uses a historical radar echo sequence to predict / extrapolate a future radar sequence;

[0041] Masking is to locally block the image to be processed to control the area or process of image processing, and can be used in the field of digital image processing.

[0042] Embodiment 1

[0043] According to an embodiment of the present invention, an embodiment of a data processing method is also provided. It should be noted that the steps shown in the flowchart of the accompanying drawings can be executed in a computer system such as a set of computer-executable instructions, and although the logical order is shown in the flowchart, in some cases, the steps shown or described can be executed in a different order than here.

[0044] The method embodiments provided in the first embodiment of this application can be executed on a mobile terminal, a computer terminal, or a similar computing device. Figure 1 is a hardware structure block diagram of a computer terminal (or mobile device) for a data processing method according to an embodiment of the present invention. As Figure 1 shown, the computer terminal 10 (or mobile device 10) may include one or more processors 102 (shown as 102a, 102b,..., 102n in the figure) (the processor 102 may include, but is not limited to, a processing device such as a microprocessor MCU or a programmable logic device FPGA), a memory 104 for storing data, and a transmission device 106 for communication functions. In addition, it may further include: a display, an input / output interface (I / O interface), a universal serial bus (USB) port (which may be included as one of the ports of the I / O interface), a network interface, a power supply, and / or a camera. Those of ordinary skill in the art can understand that Figure 1 the structure shown is only illustrative and does not limit the structure of the above-mentioned electronic device. For example, the computer terminal 10 may further include more or fewer components than Figure 1 shown, or have a different configuration from Figure 1 shown.

[0045] It should be noted that the above one or more processors 102 and / or other data processing circuits are generally referred to as "data processing circuits" in this article. The data processing circuit may be embodied in whole or in part as software, hardware, firmware, or any combination thereof. In addition, the data processing circuit may be a single independent processing module, or be incorporated in whole or in part into any one of the other elements in the computer terminal 10 (or mobile device). As involved in the embodiments of this application, the data processing circuit is a kind of processor control (such as the selection of a variable resistance terminal path connected to an interface).

[0046] The memory 104 can be used to store software programs and modules of application software, such as the program instructions / data storage devices corresponding to the data processing method in the embodiments of the present invention. The processor 102 executes various functional applications and data processing by running the software programs and modules stored in the memory 104, that is, implements the data processing method of the above application program. The memory 104 may include a high-speed random access memory, and may also include a non-volatile memory, such as one or more magnetic storage devices, flash memories, or other non-volatile solid-state memories. In some instances, the memory 104 may further include a memory remotely set relative to the processor 102, and these remote memories can be connected to the computer terminal 10 through a network. Examples of the above network include, but are not limited to, the Internet, an enterprise intranet, a local area network, a mobile communication network, and combinations thereof.

[0047] The transmission device 106 is used to receive or send data via a network. Specific examples of the above-mentioned network may include a wireless network provided by a communication provider of the computer terminal 10. In one example, the transmission device 106 includes a network adapter (Network Interface Controller, abbreviated as NIC), which can be connected to other network devices through a base station so as to communicate with the Internet. In one example, the transmission device 106 can be a Radio Frequency (RF) module, which is used to communicate with the Internet wirelessly.

[0048] The display can be, for example, a touch-screen liquid crystal display (LCD), which enables a user to interact with the user interface of the computer terminal 10 (or mobile device).

[0049] It should be noted here that, in some alternative embodiments, the above Figure 1 shown computer device (or mobile device) may include hardware elements (including circuits), software elements (including computer code stored on a computer-readable medium), or a combination of both hardware elements and software elements. It should be pointed out that Figure 1 is only an example of a specific specific instance and is intended to illustrate the types of components that may exist in the above computer device (or mobile device).

[0050] Figure 2 is a flowchart of a data processing method according to an embodiment of the present invention. As Figure 2 shown, the data processing method may include the following steps:

[0051] Step S202, obtain a first target image and a second target image, where the first target image is used to represent the actual result obtained by detecting the weather before the target time period, and the second target image is used to represent the prediction result obtained by predicting the weather during the target time period.

[0052] In the technical solution provided in step S202 of the present invention above, the weather before the target time period can be detected to obtain an actual result, and then a first target image can be obtained based on the actual result. Optionally, this embodiment can detect the weather before the target time period through a weather radar to obtain an actual result, and the actual result can include a real radar echo sequence, and then a first target image can be obtained based on the real radar echo sequence. Optionally, this embodiment can determine the last frame image of the input signal of the real radar echo sequence as the first target image, which can also be called a style image (style signal). Among them, the target time period can be a prediction time period.

[0053] In this embodiment, the weather during the target period can be predicted to obtain a prediction result. This can be achieved by using radar echo extrapolation to predict the weather during the target period. For example, based on the reflectivity echo data detected by the weather radar, the intensity distribution of the echo and the movement speed and direction of the echo body are determined, and the radar echo state during the target period is predicted. Optionally, this embodiment can calculate radar echo extrapolation using a neural network to predict the weather during the target period and obtain a prediction result. For example, the weather during the target period can be predicted using a machine learning model to obtain a prediction result. This can better predict the movement, intensity change, and disappearance of the echo. The machine learning model can be a neural network structure based on a recurrent neural network and a convolutional neural network based on the temporal and spatial properties unique to the radar echo extrapolation problem.

[0054] Optionally, the above-mentioned prediction result of this embodiment can be a radar echo extrapolation sequence, and then the second target image is obtained based on it. That is, the second target image of this embodiment can be obtained based on the radar echo extrapolation task. The second target image can be a radar echo extrapolation image (radar echo extrapolation image, radar extrapolation image, extrapolated image) that reflects the movement direction and trend of the echo signal in the overall sense, but lacks detailed information and can be recorded as a content image (content signal).

[0055] In this embodiment, the second target image may only have a good depiction of the motion characteristics of large-scale structures, but may be visually blurred in terms of small-scale structures, with a blurriness higher than the target threshold, where the target threshold is a critical threshold for measuring the visual blurriness of the second target image. In addition, the extrapolation loss based on deep learning can be defined as a pixel-level norm error, which corresponds to the maximum likelihood of the pixel value and thus tends to produce a unimodal distribution of the predicted results, which is significantly different from the multimodal distribution of the true results. Therefore, using the norm error as a loss function will cause the generated image to be distorted. However, this embodiment can reconstruct the second target image.

[0056] Step S204 : converting the first target image and the second target image from the spatial domain to the frequency domain respectively.

[0057] In the technical solution provided in the above step S204 of the present invention, after acquiring the first target image and the second target image, the first target image and the second target image can be converted from the spatial domain to the frequency domain respectively.

[0058] In this embodiment, the first target image and the second target image may be images in the spatial domain. In this embodiment, the first target image and the second target image can be respectively transformed from the spatial domain to the frequency domain, and the Fourier transform can be used to transform the first target image and the second target image from the spatial domain region to the frequency domain space, so as to optimize the second target image with the first target image in the frequency domain space.

[0059] Step S206: Reconstruct the second target image in the frequency domain based on the first target image in the frequency domain to obtain a third target image.

[0060] In the technical solution provided in step S206 of the present invention, after the first target image and the second target image are respectively transformed from the spatial domain to the frequency domain, the second target image in the frequency domain is reconstructed based on the first target image in the frequency domain to obtain a third target image. For example, the second target image in the frequency domain can be reconstructed based on the target power spectrum of the first target image in the frequency domain to obtain a third target image.

[0061] In this embodiment, the target power spectrum of the first target image in the frequency domain can be obtained. The target power spectrum can be a high wavenumber power spectrum, which reflects the probability density distribution of the signal power of the first target image in the frequency band and can reflect the distribution of the signal power of the first target image at different frequencies. Among them, both the horizontal axis and the vertical axis of the target power spectrum can use logarithmic coordinates, and the vertical axis scale can be the power multiplied by the wavenumber. In this way, the total pulsating energy of the signal power is positively correlated with the area enclosed by the target power spectrum diagram.

[0062] This embodiment can reconstruct the second target image in the frequency domain based on the target power spectrum of the first target image in the frequency domain to obtain a third target image. The third target image is a radar extrapolation reconstruction image of the second target image based on the power spectrum, that is, a frequency domain reconstruction image (reconstruction image) based on the target power spectrum, and its power spectrum can be consistent with the target power spectrum of the first target image. Optionally, this embodiment performs frequency domain energy reconstruction on the second target image in the frequency domain based on the target power spectrum of the first target image in the frequency domain to obtain a third target image.

[0063] It should be noted that the method of reconstructing the second target image in the frequency domain based on the target power spectrum of the first target image in the frequency domain in this embodiment to obtain a third target image is only a preferred implementation manner of the embodiment of the present invention, and is not limited to the method of obtaining the third target image in the embodiment of the present invention being only to reconstruct the second target image in the frequency domain based on the target power spectrum of the first target image in the frequency domain. Any statistical method that can be used to implement echo extrapolation and reconstruct the content image through the style image is acceptable, such as machine learning algorithms, which will not be listed one by one here.

[0064] Step S208: Convert the third target image from the frequency domain to the spatial domain to obtain the fourth target image.

[0065] In the technical solution provided in step S208 of the present invention, after reconstructing the second target image in the frequency domain based on the first target image in the frequency domain to obtain the third target image, the third target image is converted from the frequency domain to the spatial domain to obtain the fourth target image.

[0066] In this embodiment, for the above-mentioned third target image, it can be converted from the frequency domain to the spatial domain through inverse Fourier transform to obtain the fourth target image. The fourth target image can be the finally radar extrapolation reconstruction image based on the target power spectrum, that is, the spatial domain reconstruction image (the reconstructed echo image), which is used as the final extrapolation image.

[0067] In this embodiment, the energy of the signal of the fourth target image is consistent with that of the first target image on the high wavenumber power spectrum to ensure the richness of the small-scale details of the fourth target image. That is, in the high wavenumber range, the overall power spectrum curve of the fourth target image fits the power spectrum of the signal of the first target image.

[0068] In this embodiment, the overall echo form of the fourth target image can be consistent with that of the second target image, and the small-scale detail information of the first target image is supplemented to improve the overall visualization effect of the extrapolation image.

[0069] Through steps S202 to S208 of the present application, the first target image and the second target image are obtained. The first target image is used to represent the actual result obtained by detecting the weather before the target time period, and the second target image is used to represent the prediction result obtained by predicting the weather during the target time period. The first target image and the second target image are respectively converted from the spatial domain to the frequency domain. Based on the first target image in the frequency domain, the second target image in the frequency domain is reconstructed to obtain the third target image. The third target image is converted from the frequency domain to the spatial domain to obtain the fourth target image. That is to say, in the frequency domain space of the present application, based on the power spectrum of the style image, the radar echo extrapolation image is reconstructed. The overall echo form of the finally reconstructed image can be consistent with the radar echo extrapolation image, and the small-scale detail information of the style image is supplemented, improving the visual effect of the image and solving the technical problem of poor visualization effect of the image.

[0070] The above method of this embodiment will be further introduced below.

[0071] As an alternative implementation, the method further includes: obtaining a first frequency-domain signal of a first target image in the frequency domain, where the information refinement degree of the first frequency-domain signal is higher than a first threshold; and determining the power spectrum of the first frequency-domain signal as the target power spectrum of the first target image in the frequency domain.

[0072] In this embodiment, the full amount of frequency-domain signals of the first target image in the frequency domain can be obtained, and the first frequency-domain signal can be obtained by performing scale decomposition on the full amount of frequency-domain signals. For example, the full amount of frequency-domain signals of the first target image can be decomposed into the first frequency-domain signal by using a truncation scale (wave number), and the information refinement degree of the first frequency-domain signal is higher than the first threshold. That is, the first frequency-domain signal in this embodiment can be a small-scale signal, also referred to as a small-scale signal in the frequency domain of the style image, where the above truncation scale can be determined by using the missing range of the small-scale power spectrum.

[0073] In this embodiment, the probability density distribution of the signal power of the first target image in the frequency band can be performed to obtain the power spectrum of the first frequency-domain signal, which can also be referred to as the small-scale power spectrum. The horizontal axis and the vertical axis of the power spectrum diagram can use logarithmic coordinates, and the scale of the vertical axis is the power multiplied by the wave number. In this way, the total pulsating energy of the signal of the first target image and the area enclosed by the power spectrum diagram can show a positive correlation. This embodiment can determine the power spectrum of the first frequency-domain signal as the target power spectrum, and based on the target power spectrum of the first target image in the frequency domain, reconstruct the second target image in the frequency domain to obtain a third target image.

[0074] As an alternative implementation, reconstructing the second target image in the frequency domain based on the first target image in the frequency domain to obtain a third target image includes: determining the target phase component of the frequency-domain signal of the third target image based on the frequency-domain signal of the first target image and the frequency-domain signal of the second target image; determining the target amplitude component of the frequency-domain signal of the third target image based on the frequency-domain signal of the first target image; and generating the third target image based on the target phase component and the target amplitude component.

[0075] In this embodiment, in the frequency domain space, the signals in each frequency band are all complex numbers, having a phase component and an amplitude component. Then, for the third target image, the signals in each frequency band are also complex numbers and have a phase component and an amplitude component. Among them, the phase component is used to represent the circular frequency of the complex number, and the amplitude component is used to represent the modulus of the complex number. The third target image of this embodiment includes a target phase component and a target amplitude component in the frequency domain. When reconstructing the second target image in the frequency domain based on the first target image in the frequency domain to obtain the third target image, the frequency domain signal of the first target image in the frequency domain and the frequency domain signal of the second target image in the frequency domain can be obtained first, and then based on the frequency domain signal of the first target image and the frequency domain signal of the second target image, the target phase component of the frequency domain signal of the third target image is determined. This embodiment can also determine the target amplitude component of the frequency domain signal of the third target image based on the frequency domain signal of the first target image, and then jointly generate the third target image based on the target phase component and the target amplitude component.

[0076] As an alternative implementation, generating the third target image based on the target phase component and the target amplitude component includes: generating the third target image based on the target phase component, the target amplitude component, and the frequency domain signal of the second target image.

[0077] In this implementation, when generating the third target image based on the target phase component and the target amplitude component, the target phase component and the target amplitude component can be used to determine the small-scale part of the third target image, while for the large-scale part of the third target image, it can be realized through the frequency domain signal of the second target image. Thus, this embodiment can jointly generate the third target image based on the target phase component, the target amplitude component, and the frequency domain signal of the second target image.

[0078] As an alternative implementation, the ratio between the power spectrum of the frequency domain signal of the third target image and the target power spectrum of the first target image in the frequency domain is within the target threshold range.

[0079] This embodiment can determine the target amplitude component of the frequency domain signal of the third target image based on the frequency domain signal of the first target image in the frequency domain that has been obtained above. This target amplitude component is used to represent the small-scale energy component of the third target image, so that the ratio between the power spectrum of the frequency domain signal of the third target image and the target power spectrum of the first target image in the frequency domain is within the target threshold range, thereby achieving the purpose of making the fourth target image and the first target image consistent in the small-scale energy spectrum.

[0080] The above method of this embodiment will be further introduced below.

[0081] As an alternative implementation, based on the target power spectrum of the first target image in the frequency domain, reconstruct the second target image in the frequency domain to obtain a third target image, including: based on the second frequency-domain signal of the first target image in the frequency domain and the third frequency-domain signal of the second target image in the frequency domain, determine the target phase component of the fourth frequency-domain signal of the third target image, where the information refinement degree of the second frequency-domain signal is lower than the second threshold, the information refinement degree of the third frequency-domain signal is lower than the third threshold, and the information refinement degree of the fourth frequency-domain signal is higher than the fourth threshold; based on the first frequency-domain signal of the first target image in the frequency domain, determine the target amplitude component of the fourth frequency-domain signal of the third target image, where the ratio between the power spectrum of the fourth frequency-domain signal and the target power spectrum is within the target threshold range; based on the third frequency-domain signal of the second target image in the frequency domain, determine the fifth frequency-domain signal of the third target image, where the information refinement degree of the fifth frequency-domain signal is lower than the fifth threshold; based on the target phase component, the target amplitude component, and the fifth frequency-domain signal, generate the third target image.

[0082] In this embodiment, the second frequency-domain signal of the first target image in the frequency domain and the third frequency-domain signal of the second target image in the frequency domain can be obtained. The information refinement degree of the second frequency-domain signal is lower than the second threshold, that is, the second frequency-domain signal can be the large-scale signal of the first target image, and the third frequency-domain signal can be the large-scale signal of the second target image.

[0083] Optionally, this embodiment can obtain the full-frequency-domain signal of the first target image in the frequency domain. The above-mentioned second frequency-domain signal can be obtained by performing scale decomposition on the full-frequency-domain signal of the first target image. For example, the full-frequency-domain signal of the first target image can be decomposed into the above-mentioned second frequency-domain signal by using a truncation scale. Optionally, this embodiment can obtain the full-frequency-domain signal of the second target image in the frequency domain. The above-mentioned third frequency-domain signal can be obtained by performing scale decomposition on the full-frequency-domain signal of the second target image. For example, the full-frequency-domain signal of the second target image can be decomposed into the above-mentioned third frequency-domain signal by using a truncation scale.

[0084] In order to determine the target phase component of the third target image, after obtaining the second frequency-domain signal of the first target image in the frequency domain and the third frequency-domain signal of the second target image in the frequency domain in this embodiment, the target phase component of the fourth frequency-domain signal of the third target image can be determined based on the second frequency-domain signal and the third frequency-domain signal. The information refinement degree of the fourth frequency-domain signal is higher than the fourth threshold, that is, the fourth frequency-domain signal is the small-scale signal of the reconstructed image and can be used to provide small-scale detail information for echo extrapolation.

[0085] To determine the target amplitude component of the third target image, this embodiment can determine the target amplitude component of the fourth frequency-domain signal of the third target image based on the first frequency-domain signal of the first target image obtained above in the frequency domain, that is, determine the small-scale energy component of the reconstructed image, so that the ratio between the power spectrum of the fourth frequency-domain signal and the target power spectrum is within the target threshold range, thereby achieving the purpose of making the fourth target image and the first target image consistent in the small-scale energy spectrum.

[0086] This embodiment can also determine the fifth frequency-domain signal of the third target image based on the third frequency-domain signal of the second target image obtained above in the frequency domain. The information refinement degree of the fifth frequency-domain signal is lower than the fifth threshold, that is, the fifth frequency-domain signal is the large-scale signal of the third target image.

[0087] After determining the target phase component and the target amplitude component of the fourth frequency-domain signal of the third target image, and determining the fifth frequency-domain signal of the third target image, this embodiment can splice the target phase component and the target amplitude component of the fourth frequency-domain signal of the third target image, and the fifth frequency-domain signal of the third target image, and then obtain the complete third target image. That is, the small-scale signal of the third target image in this embodiment is the fourth frequency-domain signal determined by the target phase component and the target amplitude component, and the large-scale signal of the third target image is the fifth frequency-domain signal. Splice the large-scale signal and the small-scale signal in the frequency domain to obtain the complete frequency-domain reconstructed image.

[0088] As an optional implementation manner, based on the second frequency-domain signal of the first target image and the third frequency-domain signal of the second target image in the frequency domain, determining the target phase component of the fourth frequency-domain signal of the third target image includes: obtaining the phase difference between the third frequency-domain signal and the second frequency-domain signal; determining the target phase component based on the phase difference.

[0089] In this embodiment, since there is an overall movement of the second target image relative to the first target image, that is, there is a phase difference between the signal of the second target image and the signal of the first target image. To ensure the phase consistency between the fourth frequency-domain signal (small-scale signal) of the third target image and the phase of the overall image movement, this embodiment needs to calculate the target phase component of the fourth frequency-domain signal. Optionally, this embodiment can obtain the phase difference between the third frequency-domain signal and the second frequency-domain signal, and then determine the target phase component based on the phase difference, so that the target phase component of the third target image can be aligned with the large-scale phase angle. Optionally, the scale corresponding to the wave number 1 in the frequency domain is the full-image scale, and the corresponding phase difference can reflect the overall image movement. Therefore, the target phase component of the fourth frequency-domain signal of the third target image for each wave number can be the result of adding the phase component of the small-scale signal of the second target image to the product of the phase difference of wave number 1 and the corresponding reconstruction wave number.

[0090] As an alternative implementation, based on the first frequency-domain signal of the first target image in the frequency domain, determining the target amplitude component of the fourth frequency-domain signal of the third target image includes: determining the amplitude component of the first frequency-domain signal as the target amplitude component; and / or adjusting the amplitude component of the first frequency-domain signal based on the target coefficient to obtain the target amplitude component.

[0091] In this embodiment, the target amplitude component of the fourth frequency-domain signal of the third target image can be provided by the first target image. The amplitude component of the first frequency-domain signal can be determined as the target amplitude component, so that the target amplitude component of the third target image is aligned with the amplitude component of the first frequency-domain signal (target signal). In this embodiment, if the target amplitude component of the fourth frequency-domain signal of the third target image completely comes from the first target image, the consistency of the small-scale power spectra of the third target image and the first target image can be accurately ensured.

[0092] Optionally, considering that there will be a spatial masking operation on the third target image subsequently, its corresponding power spectrum will change. Therefore, this embodiment introduces a target coefficient, which can be an energy scaling coefficient and is used to represent the ratio of the small-scale total power of the first target image to the final fourth target image, so as to ensure the consistency of the small-scale power spectra of the final fourth target image and the first target image.

[0093] Optionally, the initial value of the above target coefficient can be set to 1.0, and this value can be updated through several rounds of iteration subsequently. The amplitude component of the small-scale signal of the final fourth target image can be the result of multiplying the amplitude component of the first frequency signal (small-scale signal) of the first target image by the above target coefficient.

[0094] In this embodiment, since the target coefficient represents the small-scale total power ratio of the first target image and the final fourth target image, this value needs to be close to 1.0 to ensure the overall consistency of the final fourth target image and the first target image in the small-scale energy spectrum. If the target coefficient differs significantly from 1.0, the following steps are re-executed: Obtain the first frequency-domain signal of the first target image in the frequency domain, and determine the power spectrum of the first frequency-domain signal as the target power spectrum of the first target image in the frequency domain; Based on the second frequency-domain signal of the first target image in the frequency domain and the third frequency-domain signal of the second target image in the frequency domain, determine the target phase component of the fourth frequency-domain signal of the third target image; Based on the first frequency-domain signal of the first target image in the frequency domain, determine the target amplitude component of the fourth frequency-domain signal of the third target image; Based on the third frequency-domain signal of the second target image in the frequency domain, determine the fifth frequency-domain signal of the third target image; Based on the target phase component, the target amplitude component, and the fifth frequency-domain signal, generate the third target image; Convert the third target image from the frequency domain to the spatial domain to obtain the fourth target image. If the target coefficient is relatively close to 1.0 after scaling, the iteration is stopped, and the power spectrum of the fourth target image can be ensured to be consistent with the power spectrum of the first target image through 3 to 5 iteration steps. After the iteration terminates, the final fourth target image can be output.

[0095] As an alternative implementation, based on the third frequency-domain signal of the second target image in the frequency domain, determining the fifth frequency-domain signal of the third target image includes: determining the third frequency-domain signal as the fifth frequency-domain signal.

[0096] In this embodiment, the third frequency-domain signal of the second target image can be directly determined as the fifth frequency-domain signal of the third target image, that is, the fifth frequency-domain signal of the third target image in this embodiment can completely originate from the third frequency-domain signal of the second target image, that is, for the frequency-domain signal of the final reconstructed image, its large-scale signal can completely originate from the large-scale signal of the content image.

[0097] As an alternative implementation, the method further includes: determining a target region in the second target image in the frequency domain, where the echo reflectivity of the target region is a target value; converting the third target image from the frequency domain to the spatial domain to obtain the fourth target image, including: converting the third target image from the frequency domain to the spatial domain, and filtering out the target region in the converted third target image to obtain the fourth target image.

[0098] In this embodiment, for the radar echo extrapolation task, in the generated second target image, there are usually a large number of target regions where the echo reflectivity is the target value. For example, when the target value is 0, there is no echo structure in such target regions, and small-scale energy reconstruction calculation is not required. Therefore, in this embodiment, a mask is used to shield the target regions (regions with a large number of echo reflectivities of 0). Among them, the reflectivity obtained from the radar echo can reflect the magnitude of the echo power of cloud and rain particles per unit volume within the detection range, and to a certain extent, reflects the distribution of the cloud and rain particle groups. In this embodiment, the third target image is transformed from the frequency domain to the spatial domain, and then the spatial mask of the second target image is loaded to filter out the target regions that do not need to be reconstructed, obtaining a complete fourth target image. In this embodiment, the target regions of the second target image are filtered by the target value, achieving the purpose of controlling the spatial region where the final frequency domain reconstruction needs to act.

[0099] As an alternative implementation, the prediction result includes a precipitation prediction result, and the method further includes: determining precipitation prediction results in a plurality of second target images respectively to obtain a plurality of precipitation prediction results; and hierarchically displaying each precipitation prediction result on the operation interface according to the precipitation level of each precipitation prediction result.

[0100] In this embodiment, the prediction result includes a precipitation prediction result, and the number of second target images can be multiple, that is, this embodiment has second target images for different target time periods. Thus, this embodiment can determine a plurality of precipitation prediction results corresponding to a plurality of different target time periods in the plurality of second target images, and can feedback the plurality of precipitation prediction results to the operation interface hierarchically. Optionally, this embodiment can determine the precipitation level of each precipitation prediction result, and on the operation interface, hierarchically display each precipitation prediction result according to the precipitation level of each precipitation prediction result, so that the user can clearly understand each precipitation prediction result.

[0101] As an alternative implementation, the method further includes: determining indication information for each precipitation prediction result, where the indication information is used to indicate the precipitation level of each precipitation prediction result; and hierarchically displaying each precipitation prediction result, including: hierarchically displaying each precipitation prediction result according to the indication information.

[0102] In this embodiment, each precipitation prediction result has indication information, and the indication information can be a gradient color, which is used to represent different precipitation levels. Thus, this embodiment can hierarchically display each precipitation prediction result on the operation interface according to the color corresponding to the precipitation level, that is, there is a corresponding relationship between the precipitation level and the color of each precipitation prediction result in this embodiment, so that the user can clearly understand each precipitation prediction result.

[0103] As an alternative embodiment, the method further includes: displaying each precipitation prediction result on the operation interface according to the region corresponding to each precipitation prediction result; and / or displaying the precipitation probability corresponding to the precipitation prediction result on the operation interface, and displaying a prompt message on the operation interface based on the precipitation probability, where the prompt message is used to prompt the travel strategy corresponding to the precipitation probability.

[0104] In this embodiment, the region (area) corresponding to each precipitation prediction result can be determined, that is, the precipitation prediction result is the result obtained by predicting the precipitation data of a certain region. Each precipitation prediction result can be displayed on the operation interface according to the region corresponding to each precipitation prediction result. For example, different display positions on the operation interface correspond to different regions, and the corresponding precipitation prediction result is displayed at the display position corresponding to the region, so as to achieve the purpose of displaying multiple precipitation prediction results in a partitioned manner. Optionally, for different regions, each precipitation prediction result can also be further displayed according to the precipitation level corresponding to each precipitation prediction result, so that the user can clearly understand each precipitation prediction result.

[0105] In this embodiment, the precipitation probability corresponding to the precipitation prediction result can be obtained and the precipitation probability can be displayed on the operation interface. Optionally, when the precipitation probability corresponding to each precipitation prediction result is greater than the target threshold, a prompt message can be displayed on the operation interface, and the prompt message is used to prompt the travel strategy corresponding to the precipitation probability to remind the user. For example, if the user needs to reach a region where the precipitation probability is greater than the target threshold, rain gear needs to be carried, etc.

[0106] As an alternative embodiment, a fourth target image is displayed on the operation interface; in response to a selection operation instruction acting on the fourth target image, a target image area is selected in the fourth target image; in response to an editing operation instruction acting on the target image area, an editing operation is performed on the target image area to obtain an editing result, and the editing result is displayed on the operation interface.

[0107] In this embodiment, after converting the third target image from the frequency domain to the spatial domain to obtain the fourth target image, the fourth target image can be displayed on the operation interface so that the user can perform further operations on the fourth target image. Optionally, this embodiment can receive a selection operation instruction on the operation interface. The selection operation instruction can be used to select a target image area in the fourth target image. Thus, in response to the selection operation instruction, a target image area is selected in the fourth target image. The target image area can be a local area in the entire area of the fourth target image, and the number thereof can be multiple. Furthermore, for the target image area, in response to an editing operation instruction acting on the target image area, an editing operation is performed on the target image area to obtain an editing result. For example, dragging at least one target image area, enlarging and shrinking at least one target image area, etc. Among them, the target image area can be an area representing the information of the extrapolated image through a certain color, that is, a color area. After performing an editing operation on the target image area to obtain an editing result, the editing result can be displayed on the operation interface, which can be to display at least one editing result corresponding to at least one target image area for the user to view and perform further analysis.

[0108] In this embodiment, an end-to-end method can be adopted to put the power spectrum as an error function into network training and perform multi-objective training together with other error functions to implement a radar echo extrapolation image reconstruction method based on the power spectrum.

[0109] The embodiment of the present invention also provides another data processing method from the human-computer interaction side.

[0110] Figure 3 It is a flowchart of another data processing method according to the embodiment of the present invention. As Figure 3 shown, the method can include the following steps:

[0111] Step S302, in response to an input operation instruction acting on the operation interface, input a first target image and a second target image, where the first target image is used to represent the actual result obtained by detecting the weather before the target time period, and the second target image is used to represent the prediction result obtained by predicting the weather during the target time period.

[0112] In the technical solution provided in step S302 of the present invention above, the operation interface can be a human-computer interaction interface on the front-end client. The user can trigger the operation interface to generate an input operation instruction, and the input operation instruction is used to input the first target image and the second target image. Thus, in response to the input operation instruction, the first target image and the second target image are obtained.

[0113] In this embodiment, the above-mentioned first target image of the input can be obtained by detecting the weather before the target time period and obtaining the actual result. Optionally, this embodiment can use the last frame image of the input signal of the real radar echo sequence as the first target image for input, and the input first target image can also be called the style image.

[0114] In this embodiment, the above-mentioned second target image of the input can be obtained by predicting the weather during the target time period and obtaining the prediction result. Optionally, this embodiment can calculate the radar echo extrapolation through a neural network method to predict the weather during the target time period and obtain the prediction result. The prediction result can be a radar echo extrapolation sequence, and then based on it, a second target image can be obtained. The second target image can be a radar echo extrapolation image reflecting the movement direction and trend in the overall sense of the echo signal, and it can be denoted as the content image.

[0115] In this embodiment, the above-mentioned second target image can only better depict the motion characteristics of large-scale structures, but there will be a situation where the image is visually blurred in terms of small-scale structures, and its blur degree is higher than the target threshold. And this embodiment can reconstruct the above-mentioned second target image.

[0116] Step S304, in response to the reconstruction operation instruction acting on the operation interface, display the fourth target image on the operation interface, where the fourth target image is obtained by converting the third target image from the frequency domain to the spatial domain, and the third target image is obtained by reconstructing the second target image converted from the spatial domain to the frequency domain based on the first target image converted from the spatial domain to the frequency domain.

[0117] In the technical solution provided in step S304 of the present invention above, the user can trigger the operation interface to generate a reconstruction operation instruction, and the reconstruction operation instruction is used to indicate reconstructing the second target image based on the first target image to obtain the fourth target image, and then display the fourth target image on the operation interface.

[0118] In this embodiment, the first target image and the second target image may be images in the spatial domain. In response to a reconstruction operation instruction on the operation interface, the first target image and the second target image can be respectively converted from the spatial domain to the frequency domain. The target power spectrum of the first target image in the frequency domain can be obtained. Based on the first target image in the frequency domain, the second target image in the frequency domain is reconstructed to obtain a third target image. The third target image is a radar extrapolation reconstruction image of the second target image based on the power spectrum, and its power spectrum can be consistent with the target power spectrum of the first target image. For the above-mentioned third target image, it can be converted from the frequency domain to the spatial domain through inverse Fourier transform to obtain a fourth target image. The fourth target image is the final radar extrapolation reconstruction image based on the target power spectrum, that is, the spatial domain reconstruction image. It is used as the final extrapolation image and then displayed on the operation interface.

[0119] In this embodiment, the fourth target image is consistent with the energy of the signal of the first target image in the high wavenumber power spectrum to ensure the richness of the small-scale details of the fourth target image. That is, in the high wavenumber range, the overall power spectrum curve of the fourth target image fits the power spectrum of the signal of the first target image.

[0120] In this embodiment, the overall echo shape of the fourth target image can be consistent with the second target image, and the small-scale detail information of the first target image is supplemented to improve the overall visualization effect of the extrapolation image.

[0121] The embodiment of the present invention also provides another data processing method from the meteorological broadcast scenario.

[0122] Figure 4 It is a flowchart of another data processing method according to the embodiment of the present invention. As Figure 4 shown, the method may include the following steps:

[0123] Step S402, obtain the first target image and the second target image from the meteorological broadcast platform, and display the first target image and the second target image on the operation interface, where the first target image is used to represent the actual result obtained by detecting the weather before the target time period, and the second target image is used to represent the prediction result obtained by predicting the weather during the target time period.

[0124] In the technical solution provided in step S402 of the present invention above, the first target image and the second target image may come from the meteorological broadcast platform. The meteorological broadcast platform can be used to broadcast predicted weather detection data and display images of actual results and prediction results, and can analyze weather detection data.

[0125] In this embodiment, the user can trigger the operation interface to generate an input operation instruction, and in response to the input operation instruction, obtain a first target image and a second target image from the weather broadcast platform.

[0126] In this embodiment, the above-mentioned input first target image may be the last frame image of the input signal of the real radar echo sequence, and the above-mentioned input second target image may be a radar echo extrapolation sequence.

[0127] Step S404: In response to the reconstruction operation instruction acting on the operation interface, display a fourth target image on the operation interface, where the fourth target image is obtained by converting the third target image from the frequency domain to the spatial domain, and the third target image is reconstructed based on the first target image converted from the spatial domain to the frequency domain and the second target image converted from the spatial domain to the frequency domain.

[0128] In the technical solution provided in step S404 of the present invention, the user can trigger the operation interface to generate a reconstruction operation instruction, which is used to indicate reconstructing the second target image based on the first target image to obtain a fourth target image, and then display the fourth target image on the operation interface.

[0129] As an optional implementation manner, returning the fourth target image to the weather broadcast platform includes: displaying the fourth target image on the operation interface of the weather broadcast platform; the method further includes: sending a selection operation instruction to the weather broadcast platform, where the selection operation instruction is used to select a target image area in the fourth target image; sending an editing operation instruction to the weather broadcast platform, where the editing operation instruction is used to perform an editing operation on the target image area and display the obtained editing result on the operation interface of the weather broadcast platform.

[0130] In this embodiment, after converting the third target image from the frequency domain to the spatial domain to obtain the fourth target image, the fourth target image is returned to the weather broadcast platform, and the fourth target image can be displayed on the operation interface of the weather broadcast platform so that the user can perform further operations on the fourth target image. Optionally, the weather broadcast platform in this embodiment can receive a selection operation instruction, which can be used to enable the weather broadcast platform to select a target image area in the fourth target image. The target image area can be a local area in the entire area of the fourth target image, and the number thereof can be multiple. Then, for the target image area, the weather broadcast platform can receive an editing operation instruction, which is used to enable the weather broadcast platform to perform an editing operation on the target image area to obtain an editing result. For example, dragging at least one target image area, enlarging and reducing at least one target image area, etc. Among them, the target image area can be an area that represents the information of the extrapolated image through a certain color, that is, a color area. After the weather broadcast platform performs an editing operation on the target image area to obtain an editing result, the editing result can be displayed on the operation interface of the weather broadcast platform, which can be to display at least one editing result corresponding to at least one target image area for the user to view and perform further analysis.

[0131] In this embodiment, the first target image and the second target image can be images in the spatial domain area. In response to a reconstruction operation instruction acting on the operation interface, the first target image and the second target image can be respectively converted from the spatial domain to the frequency domain. Based on the first target image in the frequency domain, the second target image in the frequency domain is reconstructed to obtain a third target image, which can be inverse Fourier transformed to convert it from the frequency domain to the spatial domain to obtain a fourth target image. The fourth target image is the final radar extrapolation reconstruction image based on the target power spectrum, and then it is displayed on the operation interface.

[0132] Step S406: Return the fourth target image to the weather broadcast platform.

[0133] In the technical solution provided in step S406 of the present invention above, after the fourth target image is displayed on the operation interface, the fourth target image can be returned to the weather broadcast platform.

[0134] This embodiment returns the fourth target image to the weather broadcast platform, can display the fourth target image on the weather broadcast platform, and perform further analysis on the fourth target image.

[0135] The embodiment of the present invention also provides another data processing method from an agricultural scenario.

[0136] Figure 5 is a flowchart of another data processing method according to an embodiment of the present invention. As Figure 5As shown, the method may include the following steps:

[0137] Step S502, determining a target area where the target vegetation is located.

[0138] In the technical solution provided in step S502 of the present invention, the method can be applied to a vegetation management scenario, for example, an agricultural scenario. Optionally, to determine the target area where the target vegetation is located, the target vegetation can be a crop.

[0139] Step S504, obtaining a first target image and a second target image corresponding to the target area, where the first target image is used to represent the actual result obtained by detecting the weather in the target area before the target time period, and the second target image is used to represent the precipitation prediction result obtained by predicting the weather in the target area during the target time period.

[0140] In the technical solution provided in step S504 of the present invention, the input first target image can be the last frame image of the input signal of the real radar echo sequence, and the input second target image can be the radar echo extrapolation sequence.

[0141] Step S506, respectively converting the first target image and the second target image from the spatial domain to the frequency domain.

[0142] In the technical solution provided in step S506 of the present invention, the first target image and the second target image can be images in the spatial domain area, and the first target image and the second target image can be respectively converted from the spatial domain to the frequency domain.

[0143] Step S508, reconstructing the second target image in the frequency domain based on the first target image in the frequency domain to obtain a third target image.

[0144] In the technical solution provided in step S508 of the present invention, obtain the target power spectrum of the first target image in the frequency domain, and reconstruct the second target image in the frequency domain based on the target power spectrum to obtain a third target image.

[0145] Step S510, converting the third target image from the frequency domain to the spatial domain to obtain a fourth target image.

[0146] In the technical solution provided in step S510 of the present invention, the third target image can be inversely Fourier-transformed to convert it from the frequency domain to the spatial domain to obtain a fourth target image. The fourth target image is the final radar extrapolation reconstruction image based on the target power spectrum, and then it is displayed on the operation interface.

[0147] Step S512, determining whether to fertilize the target vegetation based on the fourth target image.

[0148] In the technical solution provided in step S512 of the present invention, it is possible to determine whether to fertilize the target vegetation based on the precipitation prediction result of the fourth target image. For example, it is possible to determine whether the precipitation data in the precipitation prediction result is greater than a threshold value. If so, it can be determined that fertilization of the target vegetation is required.

[0149] Optionally, through the above method, this embodiment can predict the precipitation in the target time period in the future in the area where the target vegetation is located to determine whether fertilization of the target vegetation is required at present.

[0150] This embodiment proposes a method for reconstructing radar echo extrapolation images based on the power spectrum. The second target image to be reconstructed can be transformed into the frequency domain. The target amplitude component of the small-scale signal of the reconstructed image can be provided by the first target image, and the target phase component of the small-scale signal of the reconstructed image can be obtained from the large-scale phase difference between the second target image and the first target image. Finally, the inverse Fourier transform is used to transform it from the frequency domain to the spatial domain to obtain the final reconstructed image. The final reconstructed image obtained in this embodiment can maintain the same energy as the first target image on the high wavenumber power spectrum, thus ensuring the richness of the extrapolated small-scale details and solving the problem of blurriness existing in radar extrapolation images.

[0151] Embodiment 2

[0152] Next, a preferred implementation manner of the above method in this embodiment will be further introduced.

[0153] Radar echo extrapolation technology is the main method for nowcasting. It determines the moving speed and direction of the echo body based on the echo data detected by weather radar. The reflectivity obtained from the radar echo can reflect the magnitude of the echo power of cloud and rain particles within the detection range per unit volume, and to a certain extent, reflects the distribution of the cloud and rain particle group. Since the scattering cross-section of cloud and rain particles increases with the scale and quantity of (cloud and rain) particles, the reflectivity can have a positive correlation with the scale or quantity of precipitation particles in the meteorological target.

[0154] Radar echo extrapolation means that based on the reflectivity echo data detected by weather radar, the intensity distribution of the echo, the moving speed and direction of the echo body can be determined, and the radar echo state after a certain time period can be predicted. Deep learning methods have achieved rapid development in the problem of radar echo extrapolation. Based on the unique time and space attributes of the radar echo extrapolation problem, a neural network structure based on recurrent neural networks and convolutional neural networks can be constructed.

[0155] In the above method, machine learning is used to calculate the extrapolation of radar echoes, which can better predict the movement, intensity change and disappearance of echoes. However, the extrapolated image usually only has a good description of the motion characteristics of large-scale structures, and usually faces the situation of blurred image vision in small-scale structures. As the extrapolation time increases, the extrapolated radar image will gradually become blurred, and the local pulsation intensity will gradually weaken. For the problem of radar echo extrapolation, as the prediction time increases, the conditional probability will continuously stack up, and the result reflected in the output is that the output image lacks details and the visual presentation is relatively blurred. In addition, the extrapolation loss based on deep learning is usually defined as the norm error at the pixel level, which corresponds to the maximum likelihood of pixel values, and thus tends to produce a unimodal distribution of predicted values, which is quite different from the multimodal distribution of the real results. Therefore, using the norm error as the loss function will cause the generated image to be distorted.

[0156] Furthermore, for radar extrapolation visualization, in related technologies, the optical flow method can be used. It is assumed that the radar signal satisfies the Lagrangian conservation within a short time, and spatial continuity is introduced to solve the optical flow field. It completes the mapping of the echo image of the next frame by calculating the motion vectors between two frames. The defect of this method is that the assumption condition of the optical flow method is that the relative displacement of the same target between different frames is small and the intensity does not change significantly. In fact, the radar echo usually accompanies generation, development, weakening, and disappearance over time, that is, the brightness of the target point is constantly changing. Therefore, the optical flow method usually cannot meet complex business scenarios such as radar echo extrapolation and has gradually been replaced by machine learning algorithms.

[0157] In another related technology, a generative adversarial network can be used to achieve radar extrapolation visualization. Among them, the generative adversarial network model mainly includes a generative model and a discriminative model, and the two play against each other. Among them, the task of the generative model is to maximize the error probability of the discriminator, and the goal of the discriminative model is to distinguish the generated samples and the real samples as much as possible. The adversarial generation scheme usually encounters the problem of mode collapse. The generator is easy to generate high-probability samples and makes the distribution of the generated data deviate from the distribution of the real data. For the radar echo extrapolation task, the image generated by the adversarial generation method is likely to show structures that frequently appear in the training set, resulting in the lack of rationality in its extrapolation structure.

[0158] To address the above problems, this embodiment proposes a radar echo extrapolation image reconstruction scheme based on the power spectrum, which can transform the image to be reconstructed into the frequency domain. For small-scale signals, align its amplitude component to the target signal and its phase component to the large-scale phase angle, and finally obtain the reconstructed image through inverse Fourier transform to the spatial domain. The reconstructed image obtained by using this embodiment can keep the energy consistent with the target signal on the high-wave number power spectrum, thereby filling in the small-scale information missing in the blurred image and ensuring the richness of the extrapolated small-scale details.

[0159] In this embodiment, the information sources for image reconstruction come from two parts: one part is the real radar echo sequence before the prediction moment. Here, the last frame of its input signal can be used as the main source for reconstructing the small-scale signal, which can be denoted as the style image; the other part is the blurred radar echo extrapolation image generated by the machine learning model. This part of the image can generally reflect the movement direction and trend of the echo signal in the overall sense, but lacks detailed information, and is denoted as the content image here.

[0160] In this embodiment, the content image and the style image can be transformed to the frequency domain space through Fourier transform, and the large-scale signal and the small-scale signal can be obtained through scale decomposition; the large-scale signal of the reconstructed image comes from the large-scale part of the content image to obtain the macroscopic large-scale motion characteristics output by the prediction model; the small-scale signal of the reconstructed image is used to provide the detailed information of echo extrapolation. Its phase component is obtained from the phase difference between the large-scale signals of the content image and the style image, and its amplitude component is provided by the style image, and the energy scaling coefficient is used to ensure the consistency of the small-scale energy spectra of the final reconstructed image and the style image. Finally, for the frequency domain reconstructed image, it is transformed back to the spatial domain through inverse Fourier transform to obtain the final radar extrapolation reconstructed image based on the power spectrum. The following will introduce it in further detail.

[0161] S1, Transformation from spatial domain to frequency domain and scale decomposition.

[0162] In this embodiment, the content image and the style image can be transformed from the spatial domain to the frequency domain space through Fourier transform, and the large-scale signal and the small-scale signal of the content image and the style image can be obtained through scale decomposition.

[0163] In this embodiment, the style image and the content image can be transformed from the spatial domain to the frequency domain by using Fourier transform. In the frequency domain, the scale can be truncated to decompose each signal into a large-scale signal and a small-scale signal. In the frequency domain space, the signals on each frequency band are all complex numbers, with a phase component and an amplitude component. Among them, the phase component is used to represent the circular frequency of the complex number, and the amplitude component is used to represent the modulus of the complex number. In this embodiment, the truncation scale (wave number) can be used to decompose the full-scale frequency domain signal into its large-scale signal and small-scale signal. Among them, for the radar echo extrapolation problem, the truncation scale can be determined according to the missing range of the small-scale power spectrum.

[0164] S2, Acquisition of content image mask.

[0165] In this embodiment, the role of the image mask is to screen the regions of the content image that need to be reconstructed. For the radar echo extrapolation task, the generated extrapolated image (i.e., the content image here) usually has a large number of regions where the echo reflectivity is 0. There are no echo structures in such regions, and small-scale energy reconstruction calculations are not required. Therefore, a mask can be used to shield such regions (regions with a large number of echo reflectivities of 0). This step obtains the mask through threshold filtering to control the spatial region where the final frequency-domain reconstruction needs to act.

[0166] S3. Calculate the small-scale power spectrum of the style image.

[0167] In this embodiment, for the small-scale signals of the style image obtained from S1 in the frequency domain, the probability density distribution of its signal power is made in the frequency band to obtain the small-scale power spectrum. Both the horizontal and vertical axes of the power spectrum diagram use logarithmic coordinates, and the vertical axis scale is the power multiplied by the wave number. In this way, there is a positive correlation between the total pulsating energy of the signal and the area enclosed by the power spectrum diagram.

[0168] S4. Synthesize the small-scale phase components after reconstruction.

[0169] In this embodiment, the phase component of the small-scale signal of the reconstructed image can be obtained from the phase difference between the large-scale signals of the content image and the style image.

[0170] In this embodiment, the small-scale signal of the reconstructed image is used to provide the small-scale detail information for echo extrapolation. Since there is an overall shift of the content image relative to the style image, that is, there is a phase difference in the signals. To ensure the phase consistency of the small-scale signal with the overall shift, it is necessary to calculate the phase component of the small-scale signal. The scale corresponding to the wave number 1 in the frequency domain is the full-image scale, and the corresponding phase difference can reflect the overall shift of the image. Therefore, the phase component of the small-scale signal of the reconstructed image at each wave number can be the result obtained by superimposing the phase component of the small-scale signal of the content image and the phase difference of wave number 1 multiplied by the corresponding reconstructed wave number.

[0171] S5. Synthesize the small-scale energy components after reconstruction.

[0172] In this embodiment, the amplitude component of the small-scale signal of the reconstructed image can be provided by the style image, and an energy scaling factor is used to ensure the consistency of the final reconstructed image and the style image in the small-scale energy spectrum.

[0173] In this embodiment, if the small-scale amplitude component of the reconstructed image completely originates from the style image, the consistency of the small-scale power spectra of the two can be accurately guaranteed. However, considering the subsequent spatial masking operation on the reconstructed image, its corresponding power spectrum will change. Therefore, this embodiment introduces an energy scaling factor, which can represent the ratio of the total small-scale power of the style image to the final reconstructed image, to ensure the consistency of the small-scale energy spectra of the final reconstructed image and the style image. Optionally, the initial value of the energy scaling factor in this embodiment can be set to 1.0, and this value can be updated through several rounds of iteration later. The amplitude component of the small-scale signal of the final reconstructed image can be obtained by multiplying the amplitude component of the small-scale signal of the style image by the energy scaling factor.

[0174] S6. Obtain the frequency-domain reconstructed image.

[0175] In this embodiment, for the frequency-domain signal of the reconstructed image, its large-scale signal can completely originate from the large-scale signal of the content image. For the small-scale signal of the reconstructed image, its phase component is obtained by S4, and the amplitude component can be obtained by S5.

[0176] This embodiment splices the upscaled signal and the small-scale signal in the frequency domain to obtain a complete frequency-domain reconstructed image.

[0177] S7. Obtain the spatial-domain reconstructed image.

[0178] In this embodiment, the frequency-domain reconstructed image can be transformed to the spatial domain through inverse Fourier transform to obtain the final radar extrapolation reconstructed image based on the power spectrum.

[0179] In this embodiment, through inverse Fourier transform, the frequency-domain reconstructed image obtained in S6 is transformed to the spatial domain. Subsequently, the spatial mask of the content image obtained in S2 is loaded to filter out the areas that do not need to be reconstructed, and a complete spatial-domain reconstructed image is obtained, which is also the final reconstructed image.

[0180] S8. Calculate the ratio of the small-scale power of the final reconstructed image to the style image.

[0181] In this embodiment, the energy scaling factor represents the ratio of the total small-scale power of the style image to the final reconstructed image. This value needs to be close to 1.0 to ensure the overall consistency of the small-scale energy spectra of the final reconstructed image and the style image. If the energy scaling factor differs significantly from 1.0, S3 to S7 can be repeated. If the energy scaling factor is relatively close to 1.0, the iteration can be stopped. Usually, 3 to 5 iteration steps are required to ensure the consistency of the small-scale power spectrum of the final reconstructed image and the style power spectrum. After the iteration terminates, the spatial-domain reconstructed image obtained in S7 is the final output.

[0182] Figure 6AIt is a schematic diagram of a style image according to an embodiment of the present invention. As Figure 6A shown, the style image can be taken from the last frame of the real radar echo sequence before the prediction moment. Figure 6B It is a schematic diagram of a content image according to an embodiment of the present invention. As Figure 6B shown, the content image can be a radar echo extrapolation image obtained 1.5 hours in advance by using a network algorithm. Figure 6C It is a schematic diagram of the visualization effect of radar echo extrapolation reconstruction according to an embodiment of the present invention. As Figure 6C shown, after the frequency-domain energy reconstruction is performed by the above method of this embodiment, the overall shape of the echo can be made to be consistent with the content image, and the small-scale detail information of the style image is supplemented, and the overall visualization effect is improved.

[0183] Figure 7 It is a schematic diagram of a power spectrum according to an embodiment of the present invention. As Figure 7 shown, it shows the power spectra corresponding to the style image, the content image, and the reconstructed image. For the content image, the energy outside the cut-off wave number is much smaller than the style signal, that is, there is a significant lack of small-scale energy, which will cause the problem of visualization blur. After adopting the method for reconstructing small-scale energy of this embodiment, in the high wave number range, its overall power spectrum curve fits with the power spectrum of the style signal.

[0184] In this embodiment, a two-stage method is adopted. In the first stage, a relatively blurred extrapolation image is generated by a machine learning model; in the second stage, a radar echo extrapolation image reconstruction scheme based on the power spectrum is used as visualization enhancement. Optionally, this embodiment can also adopt an end-to-end method, putting the power spectrum as an error function into network training and performing multi-objective training together with other error functions.

[0185] Figure 8A It is a schematic diagram of a data processing scenario according to an embodiment of the present invention. As Figure 8A shown, a first target image and a second target image are input to a computing device. The first target image can be the last frame image of the input signal of the real radar echo sequence, and the second target image can be a radar echo extrapolation sequence. The first target image and the second target image can be images in the spatial domain. The first target image and the second target image can be respectively converted from the spatial domain to the frequency domain. Based on the target power spectrum of the first target image in the frequency domain, the second target image in the frequency domain is reconstructed to obtain a third target image. It can be subjected to inverse Fourier transform to convert it from the frequency domain to the spatial domain to obtain a fourth target image. This fourth target image is the final radar extrapolation reconstruction image based on the target power spectrum, and then it is displayed on the operation interface.

[0186] Figure 8BIt is a schematic diagram of another data processing scenario according to an embodiment of the present invention. As Figure 8B shown, the operation interface can be a human-computer interaction interface on the front-end client. The user can trigger the operation interface to generate an input operation instruction, and in response to the input operation instruction, obtain a first target image and a second target image. In response to the reconstruction operation instruction triggered by the user on the operation interface, a fourth target image is displayed on the operation interface, where the fourth target image is obtained by converting the third target image from the frequency domain to the spatial domain, and the third target image is obtained by reconstructing the second target image converted from the spatial domain to the frequency domain based on the first target image converted from the spatial domain to the frequency domain.

[0187] This embodiment proposes a radar echo extrapolation image reconstruction scheme based on the power spectrum, which can transform the signal to be reconstructed into the frequency domain. For small-scale signals, align their amplitude components to the target signal (the amplitude components of the small-scale signals in the reconstructed image are provided by the style image), and align their phase components to the large-scale phase angle (the phase components of the small-scale signals in the reconstructed image are obtained from the phase difference between the large-scale signals of the content image and the style image), and finally obtain the final reconstructed image by inverse Fourier transform to the spatial domain. The reconstructed image of this embodiment can maintain consistency with the target signal energy in the high wavenumber power spectrum, thus ensuring the richness of extrapolated small-scale details and solving the problem of blurriness in radar extrapolation images.

[0188] In the radar extrapolation scheme in the related art, the assumption condition of the optical flow method is that the relative displacement of the same target between different frames is small and the intensity does not change significantly, and this constraint is too strong. And the generative adversarial method has the problem of mode collapse, which easily makes the generated high wavenumber information biased.

[0189] However, in this application, optimizing the frequency domain space can make the large-scale signals of the reconstructed image come from the output of the neural network model, while the small-scale signals match the style image in the power spectrum. This does not have the assumption of target motion compared with the optical flow method, and compared with the adversarial generation method, it can ensure that the generated high wavenumber pulsation information is unbiased.

[0190] It should be noted that for the foregoing method embodiments, for the sake of simple description, they are all expressed as a series of action combinations. However, those skilled in the art should know that the present invention is not limited by the described action sequence, because according to the present invention, certain steps can be performed in other sequences or simultaneously. Secondly, those skilled in the art should also know that the embodiments described in the specification are all preferred embodiments, and the actions and modules involved are not necessarily essential to the present invention.

[0191] Through the description of the above embodiments, those skilled in the art can clearly understand that the method according to the above embodiments can be implemented by means of software plus a necessary general hardware platform. Of course, it can also be implemented by hardware, but in many cases, the former is a better implementation. Based on such an understanding, the technical solution of the present invention, in essence, or the part that contributes to the prior art, can be embodied in the form of a software product. This computer software product is stored in a storage medium (such as ROM / RAM, magnetic disk, optical disk) and includes several instructions to enable a terminal device (which can be a mobile phone, computer, server, or network device, etc.) to execute the methods described in various embodiments of the present invention.

[0192] Embodiment 3

[0193] According to an embodiment of the present invention, there is also provided a data processing device for implementing the above Figure 2 data processing method shown.

[0194] Figure 9 is a schematic diagram of a data processing device according to an embodiment of the present invention. As Figure 9 shown, the data processing device 90 may include: a first acquisition unit 91, a first conversion unit 92, a first reconstruction unit 93, and a second conversion unit 94.

[0195] The first acquisition unit 91 is configured to acquire a first target image and a second target image, where the first target image is used to represent the actual result obtained by detecting the weather before the target time period, and the second target image is used to represent the predicted result obtained by predicting the weather during the target time period.

[0196] The first conversion unit 92 is configured to respectively convert the first target image and the second target image from the spatial domain to the frequency domain.

[0197] The first reconstruction unit 93 is configured to reconstruct the second target image in the frequency domain based on the first target image in the frequency domain to obtain a third target image.

[0198] The second conversion unit 94 is configured to convert the third target image from the frequency domain to the spatial domain to obtain a fourth target image.

[0199] It should be noted here that the above first acquisition unit 91, first conversion unit 92, first reconstruction unit 93, and second conversion unit 94 correspond to steps S202 to S208 in Embodiment 1. The functions of the four units are the same as those of the corresponding steps in terms of implementation examples and application scenarios, but are not limited to the content disclosed in Embodiment 1 above. It should be noted that the above units, as part of the device, can run in the computer terminal 10 provided in Embodiment 1.

[0200] According to an embodiment of the present invention, there is also provided a data processing apparatus for implementing the above-mentioned Figure 3 data processing method shown.

[0201] Figure 10 is a schematic diagram of another data processing apparatus according to an embodiment of the present invention. As Figure 10 shown, the data processing apparatus 100 may include: an input unit 101 and a first display unit 102.

[0202] The input unit 101 is configured to input a first target image and a second target image in response to an input operation instruction acting on an operation interface, where the first target image is used to represent an actual result obtained by detecting the weather before a target time period, and the second target image is used to represent a prediction result obtained by predicting the weather during the target time period.

[0203] The first display unit 102 is configured to display a fourth target image on the operation interface in response to a reconstruction operation instruction acting on the operation interface, where the fourth target image is obtained by converting a third target image from a frequency domain to a spatial domain, and the third target image is obtained by reconstructing the second target image converted from a spatial domain to a frequency domain based on the first target image converted from a spatial domain to a frequency domain.

[0204] It should be noted here that the above-mentioned input unit 101 and the first display unit 102 correspond to steps S302 to S304 in Embodiment 1. The functions of the two units are the same as those of the corresponding steps in terms of the implemented examples and application scenarios, but are not limited to the content disclosed in the above-mentioned Embodiment 1. It should be noted that the above units, as part of the apparatus, can run in the computer terminal 10 provided in Embodiment 1.

[0205] According to an embodiment of the present invention, there is also provided a data processing apparatus for implementing the above-mentioned Figure 3 data processing method shown.

[0206] Figure 11 is a schematic diagram of another data processing apparatus according to an embodiment of the present invention. As Figure 11 shown, the data processing apparatus 110 may include: a second acquisition unit 111, a second display unit 112, and a return unit 113.

[0207] The second acquisition unit 111 is configured to acquire a first target image and a second target image from a meteorological broadcast platform and display the first target image and the second target image on the operation interface, where the first target image is used to represent an actual result obtained by detecting the weather before a target time period, and the second target image is used to represent a prediction result obtained by predicting the weather during the target time period.

[0208] A second display unit 112, configured to display a fourth target image on the operation interface in response to a reconstruction operation instruction acting on the operation interface, where the fourth target image is obtained by converting a third target image from the frequency domain to the spatial domain, and the third target image is obtained by reconstructing a second target image converted from the spatial domain to the frequency domain based on a first target image converted from the spatial domain to the frequency domain.

[0209] A return unit 113, configured to return the fourth target image to the weather broadcast platform.

[0210] It should be noted here that the above-mentioned second acquisition unit 111, second display unit 112, and return unit 113 correspond to steps S402 to S406 in Embodiment 1. The functions of the three units and the corresponding steps are the same in terms of implementation examples and application scenarios, but are not limited to the content disclosed in the above-mentioned Embodiment 1. It should be noted that the above units, as part of the device, can run in the computer terminal 10 provided in Embodiment 1.

[0211] According to an embodiment of the present invention, there is also provided a data processing device for implementing the above Figure 4 data processing method shown.

[0212] Figure 12 is a schematic diagram of another data processing device according to an embodiment of the present invention. As Figure 12 shown, the data processing device 120 may include: a first determination unit 121, a third acquisition unit 122, a third conversion unit 123, a second reconstruction unit 124, a fourth conversion unit 125, and a second determination unit 126.

[0213] The first determination unit 121 is configured to determine a target area where the target vegetation is located.

[0214] The third acquisition unit 122 is configured to acquire a first target image and a second target image corresponding to the target area, where the first target image is used to represent the actual result obtained by detecting the weather in the target area before the target time period, and the second target image is used to represent the precipitation prediction result obtained by predicting the weather in the target area during the target time period.

[0215] The third conversion unit 123 is configured to respectively convert the first target image and the second target image from the spatial domain to the frequency domain.

[0216] The second reconstruction unit 124 is configured to reconstruct the second target image in the frequency domain based on the first target image in the frequency domain to obtain a third target image.

[0217] The fourth conversion unit twelve5 is configured to convert the third target image from the frequency domain to the spatial domain to obtain a fourth target image.

[0218] A second determination unit 126, configured to determine whether to fertilize the target vegetation based on a fourth target image.

[0219] It should be noted here that the above first determination unit 121, third acquisition unit 122, third conversion unit 123, second reconstruction unit 124, fourth conversion unit 125, and second determination unit 126 correspond to steps S502 to S512 in Embodiment 1. The instances and application scenarios implemented by the six units and the corresponding steps are the same, but are not limited to the content disclosed in the above Embodiment 1. It should be noted that the above units, as part of the device, can run in the computer terminal 10 provided in Embodiment 1.

[0220] In the data processing device of this embodiment, based on the power spectrum of the style image, the radar echo extrapolation image is reconstructed. The overall echo shape of the obtained final reconstructed image can be consistent with the radar echo extrapolation image, and the small-scale detail information of the style image is supplemented, improving the visual effect of the image and solving the technical problem of poor visualization effect of the image.

[0221] Embodiment 4

[0222] An embodiment of the present invention can provide a data processing system. The data processing system may include a computer terminal, and the computer terminal may be any computer terminal device in a computer terminal group. Optionally, in this embodiment, the above computer terminal may also be replaced with a terminal device such as a mobile terminal.

[0223] Optionally, in this embodiment, the above computer terminal may be located in at least one network device among multiple network devices of a computer network.

[0224] In this embodiment, the above computer terminal may execute program codes of the following steps in the data processing method of the application program: acquire a first target image and a second target image, where the first target image is used to represent the actual result obtained by detecting the weather before a target time period, and the second target image is used to represent the prediction result obtained by predicting the weather during the target time period; respectively convert the first target image and the second target image from the spatial domain to the frequency domain; based on the first target image in the frequency domain, reconstruct the second target image in the frequency domain to obtain a third target image; convert the third target image from the frequency domain to the spatial domain to obtain a fourth target image.

[0225] Optionally, Figure 13 is a structural block diagram of a computer terminal according to an embodiment of the present invention. As Figure 13 shown, the computer terminal A may include: one or more (only one is shown in the figure): a processor 1302, a memory 1304, and a transmission device 1306.

[0226] Among them, the memory can be used to store software programs and modules, such as the program instructions / modules corresponding to the data processing method and device in the embodiments of the present invention. The processor executes various functional applications and data processing by running the software programs and modules stored in the memory, that is, to implement the above-mentioned data processing method. The memory may include a high-speed random access memory, and may also include a non-volatile memory, such as one or more magnetic storage devices, flash memory, or other non-volatile solid-state memories. In some instances, the memory may further include a memory remotely set relative to the processor, and these remote memories can be connected to the computer terminal A through a network. Examples of the above network include but are not limited to the Internet, enterprise intranet, local area network, mobile communication network, and combinations thereof.

[0227] The processor can call the information and application programs stored in the memory through the transmission device to execute the following steps: obtain a first target image and a second target image, where the first target image is used to represent the actual result obtained by detecting the weather before the target time period, and the second target image is used to represent the prediction result obtained by predicting the weather during the target time period; respectively convert the first target image and the second target image from the spatial domain to the frequency domain; based on the first target image in the frequency domain, reconstruct the second target image in the frequency domain to obtain a third target image; convert the third target image from the frequency domain to the spatial domain to obtain a fourth target image.

[0228] Optionally, the above processor can also execute the program code of the following steps: reconstruct the second target image in the frequency domain based on the first target image in the frequency domain to obtain a third target image, including: determining the target phase component of the frequency domain signal of the third target image based on the frequency domain signal of the first target image and the frequency domain signal of the second target image; determining the target amplitude component of the frequency domain signal of the third target image based on the frequency domain signal of the first target image; generating the third target image based on the target phase component and the target amplitude component.

[0229] Optionally, the above processor can also execute the program code of the following steps: generate the third target image based on the target phase component and the target amplitude component, including: generating the third target image based on the target phase component, the target amplitude component, and the frequency domain signal of the second target image.

[0230] Optionally, the above processor can also execute the program code of the following steps: obtain the first frequency domain signal of the first target image in the frequency domain, where the information refinement degree of the first frequency domain signal is higher than the first threshold; determine the target power spectrum of the first target image in the frequency domain based on the power spectrum of the first frequency domain signal.

[0231] Optionally, the processor may also execute the program code of the following steps: determining the target phase component of the fourth frequency domain signal of the third target image based on the second frequency domain signal of the first target image in the frequency domain and the third frequency domain signal of the second target image in the frequency domain, wherein the information refinement degree of the second frequency domain signal is lower than the second threshold, the information refinement degree of the third frequency domain signal is lower than the third threshold, and the information refinement degree of the fourth frequency domain signal is higher than the fourth threshold; determining the target amplitude component of the fourth frequency domain signal of the third target image based on the first frequency domain signal of the first target image in the frequency domain, wherein the ratio between the power spectrum of the fourth frequency domain signal and the target power spectrum is within the target threshold range; determining the fifth frequency domain signal of the third target image based on the third frequency domain signal of the second target image in the frequency domain, wherein the information refinement degree of the fifth frequency domain signal is lower than the fifth threshold; generating the third target image based on the target phase component, the target amplitude component and the fifth frequency domain signal.

[0232] Optionally, the processor may further execute program code of the following steps: obtaining a phase difference between the third frequency domain signal and the second frequency domain signal; and determining a target phase component based on the phase difference.

[0233] Optionally, the processor may further execute program code of the following steps: determining the amplitude component of the first frequency domain signal as the target amplitude component; and / or adjusting the amplitude component of the first frequency domain signal based on the target coefficient to obtain the target amplitude component.

[0234] Optionally, the processor may further execute a program code of the following step: determining the third frequency domain signal as the fifth frequency domain signal.

[0235] Optionally, the processor may also execute program code for the following steps: determining a target area in a second target image in the frequency domain, wherein the echo reflectivity of the target area is a target value; converting the third target image from the frequency domain to the spatial domain to obtain a fourth target image, including: converting the third target image from the frequency domain to the spatial domain, and filtering out the target area in the converted third target image to obtain a fourth target image.

[0236] Optionally, the processor may further execute program code of the following steps: determining precipitation prediction results in multiple second target images respectively to obtain multiple precipitation prediction results; and displaying each precipitation prediction result in a graded manner on the operation interface according to the precipitation level of each precipitation prediction result.

[0237] Optionally, the processor may also execute program code for the following steps: determining indication information for each precipitation forecast result, wherein the indication information is used to indicate the precipitation level of each precipitation forecast result; and displaying each precipitation forecast result in a graded manner, including: displaying each precipitation forecast result in a graded manner according to the indication information.

[0238] Optionally, the above-mentioned processor may also execute the program code of the following steps: display each precipitation prediction result on the operation interface according to the area corresponding to each precipitation prediction result; and / or display the precipitation probability corresponding to the precipitation prediction result on the operation interface, and display a prompt message on the operation interface based on the precipitation probability, where the prompt message is used to prompt the travel strategy corresponding to the precipitation probability.

[0239] Optionally, the above-mentioned processor may also execute the program code of the following steps: display a fourth target image on the operation interface; respond to a selection operation instruction on the fourth target image, select a target image area in the fourth target image; respond to an editing operation instruction on the target image area, perform an editing operation on the target image area to obtain an editing result, and display the editing result on the operation interface.

[0240] As an optional example, the processor may also call the information and application programs stored in the memory through the transmission device to execute the following steps: respond to an input operation instruction on the operation interface, input a first target image and a second target image, where the first target image is used to represent the actual result obtained by detecting the weather before the target time period, and the second target image is used to represent the prediction result obtained by predicting the weather during the target time period; respond to a reconstruction operation instruction on the operation interface, display a fourth target image on the operation interface, where the fourth target image is obtained by converting the third target image from the frequency domain to the spatial domain, and the third target image is reconstructed based on the first target image converted from the spatial domain to the frequency domain and the second target image converted from the spatial domain to the frequency domain.

[0241] As an optional example, the processor may also call the information and application programs stored in the memory through the transmission device to execute the following steps: obtain a first target image and a second target image from the meteorological broadcast platform, and display the first target image and the second target image on the operation interface, where the first target image is used to represent the actual result obtained by detecting the weather before the target time period, and the second target image is used to represent the prediction result obtained by predicting the weather during the target time period; respond to a reconstruction operation instruction on the operation interface, display a fourth target image on the operation interface, where the fourth target image is obtained by converting the third target image from the frequency domain to the spatial domain, and the third target image is reconstructed based on the first target image converted from the spatial domain to the frequency domain and the second target image converted from the spatial domain to the frequency domain; return the fourth target image to the meteorological broadcast platform.

[0242] Optionally, the above-mentioned processor may also execute the program code of the following steps: display the fourth target image on the operation interface of the weather broadcast platform; send a selection operation instruction to the weather broadcast platform, where the selection operation instruction is used to select a target image area in the fourth target image; send an editing operation instruction to the weather broadcast platform, where the editing operation instruction is used to perform an editing operation on the target image area and display the obtained editing result on the operation interface of the weather broadcast platform.

[0243] As an optional example, the processor may also call the information and application programs stored in the memory through the transmission device to execute the following steps: determine the target area where the target vegetation is located; obtain a first target image and a second target image corresponding to the target area, where the first target image is used to represent the actual result obtained by detecting the weather in the target area before the target time period, and the second target image is used to represent the precipitation prediction result obtained by predicting the weather in the target area during the target time period; respectively convert the first target image and the second target image from the spatial domain to the frequency domain; based on the first target image in the frequency domain, reconstruct the second target image in the frequency domain to obtain a third target image; convert the third target image from the frequency domain to the spatial domain to obtain a fourth target image; determine whether to fertilize the target vegetation based on the fourth target image.

[0244] By adopting the embodiment of the present invention, a data processing method is provided. Obtain a first target image and a second target image, where the first target image is used to represent the actual result obtained by detecting the weather before the target time period, and the second target image is used to represent the prediction result obtained by predicting the weather during the target time period; respectively convert the first target image and the second target image from the spatial domain to the frequency domain; based on the first target image in the frequency domain, reconstruct the second target image in the frequency domain to obtain a third target image; convert the third target image from the frequency domain to the spatial domain to obtain a fourth target image. That is to say, in the frequency domain space of the present application, based on the power spectrum of the style image, the radar echo extrapolation image is reconstructed, and the overall echo form of the finally reconstructed image can be consistent with the radar echo extrapolation image, and the small-scale detail information of the style image is supplemented, improving the visual effect of the image and solving the technical problem of poor visualization effect of the image.

[0245] Those of ordinary skill in the art can understand that Figure 13 The structure shown is only for illustration, and the computer terminal A may also be a smart phone (such as an Android phone, an iOS phone, etc.), a tablet computer, a palm computer, and a mobile Internet device (abbreviated as MID), a PAD and other terminal devices. Figure 13 It does not limit the structure of the above-mentioned computer terminal A. For example, the computer terminal A may further include moreFigure 13 more or fewer components (such as network interfaces, display devices, etc.) shown therein, or having a configuration different from that Figure 13 shown.

[0246] Those of ordinary skill in the art can understand that all or part of the steps in the various methods of the above embodiments can be completed by instructing the hardware related to the terminal device through a program, and the program can be stored in a computer-readable storage medium. The storage medium may include: a flash drive, a read-only memory (ROM), a random access memory (RAM), a magnetic disk, an optical disc, etc.

[0247] Embodiment 5

[0248] An embodiment of the present invention further provides a computer-readable storage medium. Optionally, in this embodiment, the above computer-readable storage medium may be used to store the program code executed by the data processing method provided in the above Embodiment 1.

[0249] Optionally, in this embodiment, the above computer-readable storage medium may be located in any one of the computer terminals in a computer terminal group in a computer network, or in any one of the mobile terminals in a mobile terminal group.

[0250] Optionally, in this embodiment, the computer-readable storage medium is set to store program code for performing the following steps: obtaining a first target image and a second target image, where the first target image is used to represent the actual result obtained by detecting the weather before a target time period, and the second target image is used to represent the prediction result obtained by predicting the weather during the target time period; respectively converting the first target image and the second target image from the spatial domain to the frequency domain; reconstructing the second target image in the frequency domain based on the first target image in the frequency domain to obtain a third target image; converting the third target image from the frequency domain to the spatial domain to obtain a fourth target image.

[0251] Optionally, the computer-readable storage medium is further set to store program code for performing the following steps: reconstructing the second target image in the frequency domain based on the first target image in the frequency domain to obtain a third target image, including: determining the target phase component of the frequency domain signal of the third target image based on the frequency domain signal of the first target image and the frequency domain signal of the second target image; determining the target amplitude component of the frequency domain signal of the third target image based on the frequency domain signal of the first target image; generating the third target image based on the target phase component and the target amplitude component.

[0252] Optionally, the computer-readable storage medium is further configured to store program code for performing the following steps: generating a third target image based on the target phase component and the target amplitude component, including: generating a third target image based on the target phase component, the target amplitude component, and the frequency-domain signal of the second target image.

[0253] Optionally, the computer-readable storage medium is further configured to store program code for performing the following steps: obtaining a first frequency-domain signal of a first target image in the frequency domain, where the information refinement degree of the first frequency-domain signal is higher than a first threshold; determining the power spectrum of the first frequency-domain signal as the target power spectrum of the first target image in the frequency domain.

[0254] Optionally, the computer-readable storage medium is further configured to store program code for performing the following steps: determining a target phase component of a fourth frequency-domain signal of a third target image based on a second frequency-domain signal of the first target image in the frequency domain and a third frequency-domain signal of the second target image in the frequency domain, where the information refinement degree of the second frequency-domain signal is lower than a second threshold, the information refinement degree of the third frequency-domain signal is lower than a third threshold, and the information refinement degree of the fourth frequency-domain signal is higher than a fourth threshold; determining a target amplitude component of the fourth frequency-domain signal of the third target image based on the first frequency-domain signal of the first target image in the frequency domain, where the ratio between the power spectrum of the fourth frequency-domain signal and the target power spectrum is within a target threshold range; determining a fifth frequency-domain signal of the third target image based on the third frequency-domain signal of the second target image in the frequency domain, where the information refinement degree of the fifth frequency-domain signal is lower than a fifth threshold; generating a third target image based on the target phase component, the target amplitude component, and the fifth frequency-domain signal.

[0255] Optionally, the computer-readable storage medium is further configured to store program code for performing the following steps: obtaining a phase difference between the third frequency-domain signal and the second frequency-domain signal; determining the target phase component based on the phase difference.

[0256] Optionally, the computer-readable storage medium is further configured to store program code for performing the following steps: determining the amplitude component of the first frequency-domain signal as the target amplitude component; and / or adjusting the amplitude component of the first frequency-domain signal based on a target coefficient to obtain the target amplitude component.

[0257] Optionally, the computer-readable storage medium is further configured to store program code for performing the following steps: determining the third frequency-domain signal as the fifth frequency-domain signal.

[0258] Optionally, the computer-readable storage medium is further configured to store program code for performing the following steps: determining a target region in a second target image in the frequency domain, where the echo reflectivity of the target region is a target value; converting a third target image from the frequency domain to the spatial domain to obtain a fourth target image, including: converting the third target image from the frequency domain to the spatial domain and filtering out the target region in the converted third target image to obtain the fourth target image.

[0259] Optionally, the computer-readable storage medium is further configured to store program code for performing the following steps: respectively determining precipitation prediction results in a plurality of second target images to obtain a plurality of precipitation prediction results; and hierarchically displaying each precipitation prediction result on an operation interface according to the precipitation level of each precipitation prediction result.

[0260] Optionally, the computer-readable storage medium is further configured to store program code for performing the following steps: determining indication information for each precipitation prediction result, where the indication information is used to indicate the precipitation level of each precipitation prediction result; hierarchically displaying each precipitation prediction result, including: hierarchically displaying each precipitation prediction result according to the indication information.

[0261] Optionally, the computer-readable storage medium is further configured to store program code for performing the following steps: displaying each precipitation prediction result on an operation interface according to the region corresponding to each precipitation prediction result; and / or displaying the precipitation probability corresponding to the precipitation prediction result on the operation interface, and displaying a prompt message on the operation interface based on the precipitation probability, where the prompt message is used to prompt a travel strategy corresponding to the precipitation probability.

[0262] Optionally, the computer-readable storage medium is further configured to store program code for performing the following steps: displaying a fourth target image on an operation interface; responding to a selection operation instruction acting on the fourth target image, selecting a target image region in the fourth target image; responding to an editing operation instruction acting on the target image region, performing an editing operation on the target image region to obtain an editing result, and displaying the editing result on the operation interface.

[0263] As an alternative example, the computer-readable storage medium is configured to store program code for performing the following steps: in response to an input operation instruction acting on an operation interface, input a first target image and a second target image, where the first target image is used to represent the actual result obtained by detecting the weather before a target time period, and the second target image is used to represent the prediction result obtained by predicting the weather during the target time period; in response to a reconstruction operation instruction acting on the operation interface, display a fourth target image on the operation interface, where the fourth target image is obtained by converting a third target image from the frequency domain to the spatial domain, and the third target image is obtained by reconstructing the second target image converted from the spatial domain to the frequency domain based on the first target image converted from the spatial domain to the frequency domain.

[0264] As an alternative example, the computer-readable storage medium is configured to store program code for performing the following steps: obtain a first target image and a second target image from a weather broadcast platform, and display the first target image and the second target image on an operation interface, where the first target image is used to represent the actual result obtained by detecting the weather before a target time period, and the second target image is used to represent the prediction result obtained by predicting the weather during the target time period; in response to a reconstruction operation instruction acting on the operation interface, display a fourth target image on the operation interface, where the fourth target image is obtained by converting a third target image from the frequency domain to the spatial domain, and the third target image is obtained by reconstructing the second target image converted from the spatial domain to the frequency domain based on the first target image converted from the spatial domain to the frequency domain; return the fourth target image to the weather broadcast platform.

[0265] Optionally, the computer-readable storage medium is further configured to store program code for performing the following steps: display the fourth target image on the operation interface of the weather broadcast platform; send a selection operation instruction to the weather broadcast platform, where the selection operation instruction is used to select a target image area in the fourth target image; send an edit operation instruction to the weather broadcast platform, where the edit operation instruction is used to perform an edit operation on the target image area, and display the obtained edit result on the operation interface of the weather broadcast platform.

[0266] As an alternative example, a computer-readable storage medium is configured to store program code for performing the following steps: determining a target area where the target vegetation is located; obtaining a first target image and a second target image corresponding to the target area, wherein the first target image is used to represent the actual result obtained by detecting the weather of the target area before a target time period, and the second target image is used to represent the precipitation prediction result obtained by predicting the weather of the target area during the target time period; respectively converting the first target image and the second target image from the spatial domain to the frequency domain; reconstructing the second target image in the frequency domain based on the first target image in the frequency domain to obtain a third target image; converting the third target image from the frequency domain to the spatial domain to obtain a fourth target image; and determining whether to fertilize the target vegetation based on the fourth target image.

[0267] The serial numbers of the above embodiments of the present invention are only for description and do not represent the advantages or disadvantages of the embodiments.

[0268] In the above embodiments of the present invention, the descriptions of the respective embodiments have their own emphases. For the parts not detailed in a certain embodiment, reference may be made to the relevant descriptions of other embodiments.

[0269] In several embodiments provided in the present application, it should be understood that the disclosed technical content can be implemented in other ways. Among them, the device embodiments described above are only illustrative. For example, the division of the units is only a logical function division. In actual implementation, there may be other division methods. For example, multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the displayed or discussed couplings or direct couplings or communication connections to each other can be through some interfaces. The indirect couplings or communication connections of the units or modules can be in electrical or other forms.

[0270] The units described as separate components may or may not be physically separated. The components displayed as units may or may not be physical units, that is, they may be located in one place, or may be distributed to multiple network units. Some or all of the units can be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0271] In addition, the functional units in the various embodiments of the present invention can be integrated into one processing unit, or each unit can exist physically alone, or two or more units can be integrated into one unit. The above integrated units can be implemented in the form of hardware or in the form of software functional units.

[0272] When the integrated unit is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention, in essence, or the part that contributes to the prior art, or all or part of the technical solution, can be embodied in the form of a software product. The computer software product is stored in a storage medium and includes several instructions for causing a computer device (which may be a personal computer, a server, or a network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of the present invention. The aforementioned storage medium includes: various media such as USB flash drives, read-only memories (ROMs), random access memories (RAMs), mobile hard disks, magnetic disks, or optical discs that can store program codes.

[0273] The above are only the preferred embodiments of the present invention. It should be noted that for those of ordinary skill in the art, without departing from the principle of the present invention, several improvements and refinements can be made, and these improvements and refinements should also be regarded as the protection scope of the present invention.

Claims

1. A data processing method, characterized in that, including: Obtain a first target image and a second target image, where the first target image is used to represent the actual result obtained by detecting the weather before a target time period, and the second target image is used to represent the prediction result obtained by predicting the weather during the target time period; Convert the first target image and the second target image from the spatial domain to the frequency domain respectively; Based on the first target image in the frequency domain, reconstruct the second target image in the frequency domain to obtain a third target image; Convert the third target image from the frequency domain to the spatial domain to obtain a fourth target image.

2. The method according to claim 1, wherein Based on the first target image in the frequency domain, reconstruct the second target image in the frequency domain to obtain a third target image, including: Based on the frequency domain signal of the first target image and the frequency domain signal of the second target image, determine the target phase component of the frequency domain signal of the third target image; Based on the frequency domain signal of the first target image, determine the target amplitude component of the frequency domain signal of the third target image; Generate the third target image based on the target phase component and the target amplitude component.

3. The method according to claim 2, wherein Generate the third target image based on the target phase component and the target amplitude component, including: Generate the third target image based on the target phase component, the target amplitude component, and the frequency domain signal of the second target image.

4. The method according to claim 2, wherein The ratio between the power spectrum of the frequency domain signal of the third target image and the power spectrum of the first target image in the frequency domain is within a target threshold range.

5. The method according to claim 1, characterized in that The method further includes: Determine a target region in the second target image in the frequency domain, where the echo reflectivity of the target region is a target value; Convert the third target image from the frequency domain to the spatial domain to obtain a fourth target image, including: Convert the third target image from the frequency domain to the spatial domain, and filter out the target region in the converted third target image to obtain the fourth target image.

6. The method according to claim 1, wherein The prediction result includes a precipitation prediction result, and the method further includes: Determine the precipitation prediction result in multiple second target images respectively to obtain multiple precipitation prediction results; On the operation interface, display each precipitation prediction result according to the precipitation level of each precipitation prediction result.

7. The method according to claim 6, wherein: The method further includes: determining the indication information of each precipitation prediction result, where the indication information is used to indicate the precipitation level of each precipitation prediction result; Displaying each precipitation prediction result in a hierarchical manner includes: displaying each precipitation prediction result in a hierarchical manner according to the indication information.

8. The method according to claim 7, wherein The method further includes: On the operation interface, display each precipitation prediction result according to the corresponding region of each precipitation prediction result; and / or On the operation interface, display the precipitation probability corresponding to the precipitation prediction result, and display a prompt message on the operation interface based on the precipitation probability, where the prompt message is used to prompt the travel strategy corresponding to the precipitation probability.

9. The method according to claim 1, wherein The method further includes: Display the fourth target image on the operation interface; In response to a selection operation instruction acting on the fourth target image, select a target image area in the fourth target image; In response to an editing operation instruction acting on the target image area, perform an editing operation on the target image area to obtain an editing result, and display the editing result on the operation interface.

10. A data processing method, characterized in that, It includes: In response to an input operation instruction acting on the operation interface, input a first target image and a second target image, where the first target image is used to represent the actual result obtained by detecting the weather before the target time period, and the second target image is used to represent the prediction result obtained by predicting the weather during the target time period; In response to a reconstruction operation instruction acting on the operation interface, display a fourth target image on the operation interface, where the fourth target image is obtained by converting a third target image from the frequency domain to the spatial domain, and the third target image is reconstructed based on the first target image converted from the spatial domain to the frequency domain and the second target image converted from the spatial domain to the frequency domain.

11. A data processing method, characterized in that, It includes: Obtain a first target image and a second target image from a weather broadcast platform, and display the first target image and the second target image on the operation interface, where the first target image is used to represent the actual result obtained by detecting the weather before the target time period, and the second target image is used to represent the prediction result obtained by predicting the weather during the target time period; In response to a reconstruction operation instruction acting on the operation interface, display a fourth target image on the operation interface, where the fourth target image is obtained by converting a third target image from the frequency domain to the spatial domain, and the third target image is reconstructed based on the first target image converted from the spatial domain to the frequency domain and the second target image converted from the spatial domain to the frequency domain; Return the fourth target image to the weather broadcast platform.

12. The method according to claim 11, wherein Returning the fourth target image to the weather broadcast platform includes: displaying the fourth target image on the operation interface of the weather broadcast platform; The method further includes: sending a selection operation instruction to the weather broadcast platform, where the selection operation instruction is used to select a target image area in the fourth target image; sending an editing operation instruction to the weather broadcast platform, where the editing operation instruction is used to perform an editing operation on the target image area and display the obtained editing result on the operation interface of the weather broadcast platform.

13. A computer-readable storage medium, characterized in that, The computer-readable storage medium includes a stored program, wherein when the program is run by a processor, it controls the device where the computer-readable storage medium is located to execute the method according to any one of claims 1 to 12.

Images