A data reconstruction method and device

By constructing a space-time window and weighted average method, using some clear sky observation data to reconstruct remote sensing optical data, the problems of high requirements for input data and low utilization in the prior art are solved, and efficient data arrangement and accuracy maintenance are achieved.

CN114663325BInactive Publication Date: 2024-11-22AEROSPACE INFORMATION RES INST CAS
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202210356976.7
Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
Filing Date
2022-04-06
Publication Date
2024-11-22
Estimated Expiration
Not applicable · inactive patent

AI Technical Summary

Technical Problem

The existing remote sensing optical data fusion algorithm has high requirements for input data and requires all clear sky observation data, resulting in large data preparation workload and low utilization rate of clear sky observation data.

Method used

A data reconstruction method is proposed, by obtaining remote sensing optical data of the first resolution and the second resolution, building a spatiotemporal window, determining the correlation coefficient and conversion coefficient, determining the simulation data based on the threshold, and performing weighted averages to reconstruct the high resolution Data, allowing the use of partial clear sky observation data, reducing the workload of data preparation.

Benefits of technology

It improves the utilization rate of clear sky observation data, reduces the data preparation workload, and can generate space-time continuous remote sensing optical data sets without reducing accuracy.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN114663325B_ABST
    Figure CN114663325B_ABST
Patent Text Reader

Abstract

The present application discloses a data reconstruction method and device, the method comprising: acquiring remote sensing optical data of a first resolution and remote sensing optical data of a second resolution, the first resolution being higher than the second resolution; constructing a space-time window for each pixel of the remote sensing optical data of the first resolution; selecting the remote sensing optical data of the second resolution within the space-time window, and determining the correlation coefficient and conversion coefficient between each pixel thereof and the central pixel of the remote sensing optical data of the first resolution within the space-time window; determining simulated remote sensing optical data of the first resolution according to the remote sensing optical data of the second resolution and the conversion coefficient corresponding to the correlation coefficient greater than a preset threshold; and determining reconstructed remote sensing optical data of the first resolution according to the remote sensing optical data of the first resolution and the simulated remote sensing optical data of the first resolution.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present invention relates to the technical field of remote sensing optical data processing, and in particular to a data reconstruction method and device. Background Art

[0002] At present, in remote sensing application fields such as crop growth analysis, land cover classification, vegetation phenology monitoring, and ecological environment assessment, continuous and seamless remote sensing optical data on both temporal and spatial scales are needed to provide high-precision observation, analysis, and evaluation. Global earth observation has entered the era of medium resolution. The representative satellites of this resolution, the Landsat series, Sentinel 2, and the domestically produced Gaofen-1 and Gaofen-6 satellites, can serve more detailed and accurate analysis and evaluation, especially in areas with strong spatial heterogeneity, such as cities, mountains, and mixed forests. However, the revisit period of remote sensing satellites at this spatial scale is low, and spatiotemporally continuous remote sensing optical data needs to be obtained through data fusion methods with high temporal resolution but low spatial resolution remote sensing optical data.

[0003] Commonly used data fusion algorithms include Spatial and Temporal Adaptive Reflectance Fusion Model (STARFM), Unmixing-Based Data Fusion (UBDF), Flexible Spatiotemporal Data Fusion (FSDAF), etc. The input of this type of algorithm usually includes two parts. The first part is all the medium-resolution and low-resolution remote sensing optical data acquired on the same day 1-2 days before or 1-2 days after a certain day to be reconstructed, which is used to construct the scale relationship between medium-resolution pixels and low-resolution pixels; the second part is all the low-resolution remote sensing optical data predicted for the above-mentioned day. The observation time of the first part of the input data and the second part of the input data needs to be close to ensure the accuracy of the algorithm. Repeat the above model operation at the time when there is a lack of remote sensing optical data, and use the simulated remote sensing optical data to supplement the remote sensing optical data without satellite observation or without clear satellite observation due to weather, so as to form a remote sensing optical data set that is continuous and seamless in time and space. This method has been widely used to improve the accuracy of land cover classification, vegetation phenology monitoring in heterogeneous areas, crop growth analysis, etc. Summary of the invention

[0004] The present application proposes a data reconstruction method and device for solving the problems that the commonly used data fusion algorithms have high requirements on input data, require all clear sky observation data, have a large workload for data preparation, and have low utilization rate of clear sky observation data.

[0005] In a first aspect, the present application proposes a data reconstruction method, comprising:

[0006] Acquire remote sensing optical data of a first resolution and remote sensing optical data of a second resolution, wherein the first resolution is higher than the second resolution;

[0007] Constructing a spatiotemporal window for each pixel of the remote sensing optical data of the first resolution;

[0008] Selecting the remote sensing optical data of the second resolution within the space-time window, and determining the correlation coefficient and conversion coefficient between each pixel thereof and the central pixel of the remote sensing optical data of the first resolution within the space-time window;

[0009] Determine simulated remote sensing optical data of the first resolution according to the remote sensing optical data of the second resolution corresponding to the correlation coefficient greater than a preset threshold and the conversion coefficient;

[0010] Reconstructed remote sensing optical data with the first resolution are determined according to the remote sensing optical data with the first resolution and the simulated remote sensing optical data with the first resolution.

[0011] In one possible implementation, the remote sensing optical data of the first resolution and the remote sensing optical data of the second resolution respectively include clear sky observation data and / or non-clear sky observation data, the clear sky observation data include all clear sky observation data and / or the clear sky portion of partial clear sky observation data, and the non-clear sky observation data include the non-clear sky portion of partial clear sky observation data.

[0012] In a possible implementation, the remote sensing optical data with the first resolution and the remote sensing optical data with the second resolution are remote sensing optical data of a full time period prepared at one time.

[0013] In a possible implementation, constructing a spatiotemporal window for each pixel of the remote sensing optical data of the first resolution includes:

[0014] For each pixel of the remote sensing optical data of the first resolution, a spatial window with a spatial range of ±w*±w is selected with the pixel as the center, and a ±t translation is performed in the time range to obtain a (2w+1)*(2w+1)*(2t+1) space-time window, wherein w is the length from the center pixel in the spatial range, and t is the length from the center pixel in the time range.

[0015] In a possible implementation, determining the reconstructed remote sensing optical data with the first resolution according to the remote sensing optical data with the first resolution and the simulated remote sensing optical data with the first resolution includes:

[0016] The remote sensing optical data with the first resolution and the simulated remote sensing optical data with the first resolution are weighted averaged to obtain the reconstructed remote sensing optical data with the first resolution.

[0017] In a possible implementation, the step of weighted averaging the remote sensing optical data of the first resolution and the simulated remote sensing optical data of the first resolution to obtain the reconstructed remote sensing optical data of the first resolution includes:

[0018] If there is any pixel corresponding to the remote sensing optical data of the first resolution that is clear sky observation data, then determine the weight of the remote sensing optical data of the first resolution that is corresponding to the pixel as a first weight, wherein the first weight is higher than the weight corresponding to the simulated remote sensing optical data of the first resolution;

[0019] performing weighted summation of the remote sensing optical data of the first resolution corresponding to any pixel and the simulated remote sensing optical data of the first resolution corresponding to any pixel to obtain the reconstructed remote sensing optical data of the first resolution corresponding to any pixel; or

[0020] If the remote sensing optical data of the first resolution corresponding to any pixel is non-clear sky observation data, the weight of the remote sensing optical data of the first resolution corresponding to the pixel is determined to be 0;

[0021] The simulated remote sensing optical data of the first resolution corresponding to any one of the pixels is weighted to obtain the reconstructed remote sensing optical data of the first resolution corresponding to any one of the pixels.

[0022] In a second aspect, the present application proposes a data reconstruction device, comprising:

[0023] The acquisition unit is used to acquire remote sensing optical data of a first resolution and remote sensing optical data of a second resolution, wherein the first resolution is higher than the second resolution;

[0024] The processing unit is used to construct a spatiotemporal window for each pixel of the remote sensing optical data of the first resolution;

[0025] The processing unit is also used to select the remote sensing optical data of the second resolution within the space-time window, and determine the correlation coefficient and conversion coefficient between each pixel thereof and the central pixel of the remote sensing optical data of the first resolution within the space-time window;

[0026] The processing unit is further used to determine the simulated remote sensing optical data of the first resolution according to the remote sensing optical data of the second resolution corresponding to the correlation coefficient greater than a preset threshold and the conversion coefficient;

[0027] The processing unit is further configured to determine reconstructed remote sensing optical data with the first resolution according to the remote sensing optical data with the first resolution and the simulated remote sensing optical data with the first resolution.

[0028] In one possible implementation, the remote sensing optical data of the first resolution and the remote sensing optical data of the second resolution respectively include clear sky observation data and / or non-clear sky observation data, the clear sky observation data include all clear sky observation data and / or the clear sky portion of partial clear sky observation data, and the non-clear sky observation data include the non-clear sky portion of partial clear sky observation data.

[0029] In a possible implementation, the remote sensing optical data with the first resolution and the remote sensing optical data with the second resolution are remote sensing optical data of a full time period prepared at one time.

[0030] In a possible implementation, the processing unit is specifically used to select, for each pixel of the remote sensing optical data of the first resolution, a spatial window with a spatial range of ±w*±w with the pixel as the center, and simultaneously perform a ±t translation in the time range to obtain a (2w+1)*(2w+1)*(2t+1) space-time window, where w is the length from the center pixel in the spatial range, and t is the length from the center pixel in the time range.

[0031] In a possible implementation, the processing unit is specifically configured to perform a weighted average of the remote sensing optical data of the first resolution and the simulated remote sensing optical data of the first resolution to obtain the reconstructed remote sensing optical data of the first resolution.

[0032] In a possible implementation, the processing unit is specifically used to determine, if any pixel has remote sensing optical data of the first resolution corresponding to it that is clear sky observation data, a weight of the remote sensing optical data of the first resolution corresponding to the pixel as a first weight, wherein the first weight is higher than a weight corresponding to the simulated remote sensing optical data of the first resolution;

[0033] The remote sensing optical data of the first resolution corresponding to any pixel and the simulated remote sensing optical data of the first resolution corresponding to any pixel are weighted and summed to obtain the reconstructed remote sensing optical data of the first resolution corresponding to any pixel; or if the remote sensing optical data of the first resolution corresponding to any pixel is non-clear sky observation data, the weight of the remote sensing optical data of the first resolution corresponding to the pixel is determined to be 0; the simulated remote sensing optical data of the first resolution corresponding to any pixel are weighted to obtain the reconstructed remote sensing optical data of the first resolution corresponding to any pixel.

[0034] In a third aspect, the present application further proposes a data reconstruction device, comprising at least one processor, wherein the processor is used to execute a program stored in a memory, and when the program is executed, the device executes the various steps in the first aspect and various possible implementations.

[0035] In a fourth aspect, the present application further proposes a non-transitory computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the various steps in the first aspect and various possible implementations.

[0036] It can be seen from the above technical scheme that the remote sensing optical data of the first resolution and the remote sensing optical data of the second resolution obtained in the present application can be all clear sky observation data and / or part of the clear sky observation data, which has low requirements for input data and improves the utilization rate of clear sky observation data; the remote sensing optical data of the first resolution and the remote sensing optical data of the second resolution are remote sensing optical data of the full time period prepared at one time, which reduces the workload of data preparation. BRIEF DESCRIPTION OF THE DRAWINGS

[0037] In order to more clearly illustrate the embodiments of the present application or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0038] Figure 1 A flowchart of a data reconstruction method provided in an embodiment of the present application;

[0039] Figure 2-4 The technical solutions provided in the embodiments of the present application are respectively used for Landsat 7ETM+, GF1 WFV, and Sentinel 2MSI in the case of satellite failure, satellite non-transit, and cloudy weather.

[0040] Figure 5 A schematic diagram of the structure of a data reconstruction device provided in an embodiment of the present application;

[0041] Figure 6 A schematic diagram of the structure of a data reconstruction device provided in an embodiment of the present application. DETAILED DESCRIPTION

[0042] The technical solutions in the embodiments of the present invention will be described below in conjunction with the accompanying drawings in the embodiments of the present invention. The following embodiments are only used to more clearly illustrate the technical solutions of the present invention, and cannot be used to limit the protection scope of the present invention.

[0043] It should be noted that the term "and / or" in this application is only a kind of association relationship describing the associated objects, indicating that there may be three relationships, for example, A and / or B, which can represent: A exists alone, A and B exist at the same time, and B exists alone. The terms "first" and "second" in the description and claims of the embodiments of this application are used to distinguish different objects, rather than to describe the specific order of objects. For example, the first resolution and the second resolution are used to distinguish different resolutions, rather than to describe the specific order of target objects. In addition, the first resolution, the second resolution and other resolutions mentioned in the embodiments of this application are all spatial resolutions. In the embodiments of this application, the words "exemplary", "for example" or "for example" are used to indicate examples, illustrations or explanations. Any embodiment or design described as "exemplary", "for example" or "for example" in the embodiments of this application should not be interpreted as having advantages over other embodiments or design schemes. Specifically, the use of words such as "exemplary" or "for example" is intended to present related concepts in a specific way. In the description of the embodiments of this application, unless otherwise specified, the meaning of "multiple" refers to two or more.

[0044] In a possible implementation, higher-resolution spatiotemporally continuous remote sensing optical data needs to be obtained by data fusion with lower-resolution remote sensing optical data. Commonly used data fusion algorithms include STARFM, UBDF, FSDAF, etc. The input of this type of algorithm usually includes two parts. The first part is all the medium-resolution and low-resolution remote sensing optical data acquired on the same day 1-2 days before or 1-2 days after a certain day to be reconstructed, which is used to construct the scale relationship between medium-resolution pixels and low-resolution pixels. The second part is all the predicted low-resolution remote sensing optical data for the above-mentioned day. In this possible implementation, there are the following technical problems:

[0045] 1) The remote sensing optical data in the first and second parts above need to be all clear sky observation data, which is difficult to meet in large areas and cloudy and rainy areas and seasons. In addition, if the observation time interval between the first and second part input data is long, the data fusion accuracy will be too low.

[0046] 2) The output of the commonly used data fusion algorithm is the fusion result of one day, which requires that the input data that meets the requirements be prepared for each day's fusion result, which increases the workload of data preparation.

[0047] 3) The input data of the commonly used data fusion algorithm is required to be all clear sky observation data. However, in reality, there is a large amount of partial clear sky observation data in remote sensing optical data, and its data volume may be larger than that of all clear sky observation data. Therefore, the partial clear sky observation data is not effectively used, and the utilization rate of clear sky observation data is low.

[0048] In order to solve the above technical problems, the present application embodiment proposes the following Figure 1 The flowchart of a data reconstruction method shown in FIG. 1 includes: S101 to S105, specifically including:

[0049] S101, acquiring remote sensing optical data of a first resolution and remote sensing optical data of a second resolution.

[0050] The first resolution is higher than the second resolution.

[0051] In one example, unlike the commonly used data fusion algorithm whose input data can only be all clear sky observation data, the remote sensing optical data of the first resolution and the remote sensing optical data of the second resolution in the embodiment of the present application can be all clear sky observation data and / or partial clear sky observation data, respectively. In addition, in the embodiment of the present application, the clear sky part of all clear sky observation data and / or partial clear sky observation data is referred to as clear sky observation data, and the non-clear sky part of partial clear sky observation data is referred to as non-clear sky observation data.

[0052] In one example, the remote sensing optical data of the first resolution and the remote sensing optical data of the second resolution are remote sensing optical data of a full time period prepared at one time. The full time period mentioned here can be a year or a growing season, with days as the time unit.

[0053] It is understandable that when using a commonly used data fusion algorithm to output a day's fusion result, it is necessary to prepare input data that meets the requirements for the fusion result of that day. That is to say, for each day's fusion result output by a commonly used data fusion algorithm, it is necessary to prepare the input data for each day. Exemplarily, taking Landsat 8 and 7 in the Beijing area in 2020 as an example, there are 11 scenes of all clear sky observation data throughout the year and 22 scenes of partial clear sky observation data. When forming continuous data with an interval of 4 days throughout the year, 365 / 4-11≈81 data preparations are required. In an embodiment of the present application, only one remote sensing optical data preparation for the entire time period is required, so the workload of data preparation is reduced by 80 times.

[0054] S102: constructing a spatiotemporal window for each pixel of the remote sensing optical data of the first resolution.

[0055] In one example, for each pixel of the remote sensing optical data of the first resolution, a spatial window with a spatial range of ±w*±w is selected with the pixel as the center, and a ±t translation is performed in the time range to obtain a (2w+1)*(2w+1)*(2t+1) space-time window, where w is the length from the center pixel in the spatial range, and t is the length from the center pixel in the time range.

[0056] S103, selecting remote sensing optical data of the second resolution within the spatiotemporal window, and determining the correlation coefficient and conversion coefficient between each pixel thereof and the central pixel of the remote sensing optical data of the first resolution within the spatiotemporal window.

[0057] Among them, the correlation coefficient represents the strength of the correlation between pixels, and the conversion coefficient represents the conversion multiple between pixels.

[0058] In one example, remote sensing optical data of the second resolution within the spatiotemporal window is selected, and correlation coefficients and conversion coefficients between each pixel thereof and a central pixel of remote sensing optical data of the first resolution within the spatiotemporal window are determined.

[0059] S104, determining simulated remote sensing optical data of the first resolution according to the remote sensing optical data of the second resolution corresponding to the correlation coefficient greater than a preset threshold and the conversion coefficient.

[0060] In one example, remote sensing optical data of the second resolution corresponding to a correlation coefficient greater than a preset threshold is selected, and the simulated remote sensing optical data of the first resolution is determined based on the second resolution remote sensing optical data and the conversion coefficient.

[0061] Exemplarily, the remote sensing optical data of the second resolution corresponding to the highest correlation coefficient is selected, and the simulated remote sensing optical data of the first resolution is calculated using formula (1). i,j,t represents the remote sensing optical data of the second resolution corresponding to the selected highest correlation coefficient, a i,j,t Represents the corresponding conversion factor, FS i,j,t represents the simulated remote sensing optical data of the first resolution. (i, j, t) represents the position of the pixel corresponding to the remote sensing optical data of the second resolution corresponding to the selected highest correlation coefficient in the space-time window (i, j) and time (t).

[0062] FS i,j,t =a i,j,t *C i,j,t (1)

[0063] S105 , determining reconstructed remote sensing optical data with the first resolution according to the remote sensing optical data with the first resolution and the simulated remote sensing optical data with the first resolution.

[0064] It can be understood that in order to eliminate observations such as clouds, shadows, thick aerosols, etc. that are not identified by the cloud mask that may exist in the remote sensing optical data of the first resolution and the simulated remote sensing optical data of the first resolution, outliers in the remote sensing optical data of the first resolution and the simulated remote sensing optical data of the first resolution are removed.

[0065] In one example, since the remote sensing optical data of the first resolution may have obvious seams or defects, the reconstructed remote sensing optical data of the first resolution is determined based on the remote sensing optical data of the first resolution with outliers removed and the simulated remote sensing optical data of the first resolution.

[0066] Specifically, if there is any pixel whose corresponding remote sensing optical data of the first resolution is clear sky observation data, the weight of the remote sensing optical data of the first resolution corresponding to the pixel is determined to be a first weight, and the first weight is higher than the weight corresponding to the simulated remote sensing optical data of the first resolution;

[0067] performing weighted summation of the remote sensing optical data of the first resolution corresponding to any pixel and the simulated remote sensing optical data of the first resolution corresponding to any pixel to obtain the reconstructed remote sensing optical data of the first resolution corresponding to any pixel; or

[0068] If the remote sensing optical data of the first resolution corresponding to any pixel is non-clear sky observation data, the weight of the remote sensing optical data of the first resolution corresponding to the pixel is determined to be 0;

[0069] The simulated remote sensing optical data of the first resolution corresponding to any one of the pixels is weighted to obtain the reconstructed remote sensing optical data of the first resolution corresponding to any one of the pixels.

[0070] For example, if the remote sensing optical data of the first resolution corresponding to any pixel on a certain day is clear sky observation data F F , then it is determined that the weight of the remote sensing optical data of the first resolution corresponding to the pixel is the highest, which is the first weight r F , the first weight is higher than the weight corresponding to the simulated remote sensing optical data of the first resolution. The remote sensing optical data of the first resolution corresponding to any pixel and the simulated remote sensing optical data of the first resolution corresponding to any pixel are weighted and summed by formula (2) to obtain the reconstructed remote sensing optical data of the first resolution corresponding to any pixel FS A If the remote sensing optical data of the first resolution corresponding to any pixel on a certain day is non-clear sky observation data, the weight of the remote sensing optical data of the first resolution corresponding to the pixel is determined to be 0. The simulated remote sensing optical data of the first resolution corresponding to any pixel is weighted by formula (2) to obtain the reconstructed remote sensing optical data of the first resolution corresponding to any pixel FS A Among them, r * Represents weights, whose sum is 1.

[0071]

[0072] After performing the above operations on each pixel, complete clear sky observation data is formed. Outputting the image of each day is a spatially seamless remote sensing optical dataset for the whole year or the entire growing season.

[0073] It should be noted that the number of remote sensing optical data of the first resolution for the whole year corresponding to any one pixel in the embodiment of the present application should be greater than 5, and the number of missing remote sensing optical data of the second resolution corresponding to the pixel should be relatively small.

[0074] In actual application scenarios, taking Landsat 8 and 7 in the Beijing area in 2020 as an example, the number of pixels of all clear sky observation data accounts for 40% of the number of pixels of clear sky observation data. Therefore, the data utilization rate has increased by more than 2 times.

[0075] The technical solution provided in this application is applicable to the reconstruction of data missing caused by different satellites (Landsat series, GF multispectral satellite, Sentinel2 satellite) and different situations (satellite failure, satellite non-transit, cloudy weather, etc.). Figure 2-4 The technical solutions provided in this application are used for the data reconstruction results of Landsat 7ETM+, GF1 WFV, and Sentinel 2MSI under the conditions of satellite failure, satellite non-passing, and cloudy weather. The reconstructed index uses the Enhanced Vegetation Index (EVI).

[0076] See also Figure 2 , Figure 2 The left picture shows remote sensing optical data observed by satellite. The slashed part represents the missing data caused by satellite failure. Figure 2 The right image shows the reconstructed result. Figure 3 , Figure 3 The picture on the left shows remote sensing optical data observed by satellite. The black triangles in the picture represent missing data due to the satellite not passing through. Figure 3 The right image shows the reconstructed result. Figure 4 , Figure 4 The inside of the box on the left side of the figure is divided into shadows that are not identified in the cloud mask algorithm during cloudy weather, and the result after reconstruction, that is, Figure 4 In the middle right figure, this part of the shadow is well corrected. It can be seen that the reconstructed result fills in the missing data due to satellite failure, satellite non-passing, cloudy weather, etc. The reconstructed data has no obvious seams. In the calculation, the outlier deletion process is introduced to eliminate the influence of shadows that are not identified in the medium-resolution data.

[0077] In actual application scenarios, the accuracy evaluation of the technical solution provided in the embodiment of the present application selects the partial clear sky observation data of a single scene in a year as the benchmark, and the number of clear sky pixels in the partial clear sky observation data is greater than 80%. The benchmark data is removed in the calculation, and then the reconstructed data is compared with the benchmark data. Three indicators, namely Pearson's r, root mean square error RMSE, and bias BIAS, are selected to evaluate the performance of the technical solution of the present application. Table 1 shows that the performance of the three indicators of the algorithm TSFM of the technical solution of the present application on the 104th day and the 152nd day is similar to the performance of the three indicators of the commonly used data fusion algorithm ESTARFM.

[0078] Table 1

[0079]

[0080] Figure 5 A structural diagram 500 of a data reconstruction device provided in the present application includes: an acquisition unit 501 and a processing unit 502;

[0081] The acquisition unit 501 is used to acquire remote sensing optical data of a first resolution and remote sensing optical data of a second resolution, wherein the first resolution is higher than the second resolution;

[0082] The processing unit 502 is used to construct a spatiotemporal window for each pixel of the remote sensing optical data of the first resolution;

[0083] The processing unit 502 is further used to select the remote sensing optical data of the second resolution within the space-time window, and determine the correlation coefficient and conversion coefficient between each pixel thereof and the central pixel of the remote sensing optical data of the first resolution within the space-time window;

[0084] The processing unit 502 is further configured to determine simulated remote sensing optical data of the first resolution according to the remote sensing optical data of the second resolution corresponding to the correlation coefficient greater than a preset threshold and the conversion coefficient;

[0085] The processing unit 502 is further configured to determine reconstructed remote sensing optical data with the first resolution according to the remote sensing optical data with the first resolution and the simulated remote sensing optical data with the first resolution.

[0086] In one possible implementation, the remote sensing optical data of the first resolution and the remote sensing optical data of the second resolution respectively include clear sky observation data and / or non-clear sky observation data, the clear sky observation data include all clear sky observation data and / or the clear sky portion of partial clear sky observation data, and the non-clear sky observation data include the non-clear sky portion of partial clear sky observation data.

[0087] In a possible implementation, the remote sensing optical data with the first resolution and the remote sensing optical data with the second resolution are remote sensing optical data of a full time period prepared at one time.

[0088] In a possible implementation, the processing unit 502 is specifically used to select a spatial window with a spatial range of ±w*±w with each pixel of the remote sensing optical data of the first resolution as the center, and simultaneously perform a ±t translation in the time range to obtain a (2w+1)*(2w+1)*(2t+1) space-time window, wherein w is the length from the center pixel in the spatial range, and t is the length from the center pixel in the time range.

[0089] In a possible implementation, the processing unit 502 is specifically configured to perform a weighted average of the remote sensing optical data with the first resolution and the simulated remote sensing optical data with the first resolution to obtain the reconstructed remote sensing optical data with the first resolution.

[0090] In a possible implementation, the processing unit 502 is specifically configured to determine, if any pixel has remote sensing optical data of the first resolution corresponding to it that is clear sky observation data, a weight of the remote sensing optical data of the first resolution corresponding to the pixel as a first weight, wherein the first weight is higher than a weight corresponding to the simulated remote sensing optical data of the first resolution;

[0091] The remote sensing optical data of the first resolution corresponding to any pixel and the simulated remote sensing optical data of the first resolution corresponding to any pixel are weighted and summed to obtain the reconstructed remote sensing optical data of the first resolution corresponding to any pixel; or if the remote sensing optical data of the first resolution corresponding to any pixel is non-clear sky observation data, the weight of the remote sensing optical data of the first resolution corresponding to the pixel is determined to be 0; the simulated remote sensing optical data of the first resolution corresponding to any pixel are weighted to obtain the reconstructed remote sensing optical data of the first resolution corresponding to any pixel.

[0092] See also Figure 6 , the embodiment of the present application also provides a data reconstruction device 600 for implementing the functions of the above-mentioned method embodiment. The data reconstruction device 600 can be a chip system. In the embodiment of the present application, the chip system can be composed of chips, or can include chips and other discrete devices. The data reconstruction device 600 includes at least one processor 610, which is used to implement the functions of the method embodiment provided by the embodiment of the present application. The data reconstruction device 600 may also include a communication interface 620. In the embodiment of the present application, the communication interface may be a transceiver, circuit, bus, module or other type of communication interface, which is used to communicate with other devices through a transmission medium. For example, the communication interface 620 is used for the device in the data reconstruction device 600 to communicate with other devices.

[0093] The processor 610 can execute the functions executed by the processing unit 502 in the data reconstruction device 500 ; the communication interface 620 can be used to execute the functions executed by the acquisition unit 501 in the data reconstruction device 500 .

[0094] When the data reconstruction device 600 is used to execute the data reconstruction method (eg Figure 1 The method shown in the figure) is as follows: the communication interface 620 is used to obtain remote sensing optical data of a first resolution and remote sensing optical data of a second resolution, the first resolution being higher than the second resolution; the processor 610 is used to construct a space-time window for each pixel of the remote sensing optical data of the first resolution; the remote sensing optical data of the second resolution within the space-time window is selected, and the correlation coefficient and conversion coefficient between each pixel thereof and the central pixel of the remote sensing optical data of the first resolution within the space-time window are determined; according to the remote sensing optical data of the second resolution corresponding to the correlation coefficient greater than a preset threshold and the conversion coefficient, the simulated remote sensing optical data of the first resolution is determined; according to the remote sensing optical data of the first resolution and the simulated remote sensing optical data of the first resolution, the reconstructed remote sensing optical data of the first resolution is determined.

[0095] The communication interface 620 is also used to perform other receiving or sending steps or operations involved in the above method embodiment. The processor 610 can also be used to perform other corresponding steps or operations in the above method embodiment except for sending and receiving, which will not be described one by one here.

[0096] The data reconstruction device 600 may also include at least one memory 630 for storing program instructions and / or data. The memory 630 is coupled to the processor 610. The coupling in the embodiment of the present application is an indirect coupling or communication connection between devices, units or modules, which may be electrical, mechanical or other forms, and is used for information exchange between devices, units or modules. The processor 620 may operate in conjunction with the memory 630. The processor 610 may execute program instructions stored in the memory 630. In one possible implementation, at least one of the at least one memory may be integrated with the processor. In another possible implementation, the memory 630 is located outside the data reconstruction device 600.

[0097] The specific connection medium between the communication interface 620, the processor 610 and the memory 630 is not limited in the embodiment of the present application. Figure 6 In the embodiment, the memory 630, the processor 610 and the communication interface 620 are connected via a bus 640. Figure 6The connections between the other components are shown in bold lines, which are only for illustration and are not intended to be limiting. The bus can be divided into an address bus, a data bus, a control bus, etc. For ease of representation, Figure 6 Only one thick line is used in the diagram, but this does not mean that there is only one bus or only one type of bus.

[0098] It should be noted that the processor 610 may be one or more central processing units (CPUs). When the processor 610 is a CPU, the CPU may be a single-core CPU or a multi-core CPU. The processor 610 may be a general-purpose processor, a digital signal processor, an application-specific integrated circuit, a field programmable gate array or other programmable logic device, a discrete gate or transistor logic device, or a discrete hardware component, and may implement or execute the methods, steps, and logic block diagrams disclosed in the embodiments of the present application. A general-purpose processor may be a microprocessor or any conventional processor, etc. The steps of the method disclosed in conjunction with the embodiments of the present application may be directly embodied as being executed by a hardware processor, or may be executed by a combination of hardware and software modules in the processor.

[0099] In the embodiment of the present application, the memory 630 may include, but is not limited to, non-volatile memory such as a hard disk drive (HDD) or a solid-state drive (SSD), a random access memory (RAM), an erasable programmable read-only memory (EPROM), a read-only memory (ROM), or a portable read-only memory (CD-ROM). The memory is any other medium that can be used to carry or store a desired program code in the form of an instruction or data structure and can be accessed by a computer, but is not limited thereto. The memory in the embodiment of the present application can also be a circuit or any other device that can realize a storage function, for storing program instructions and / or data.

[0100] The present application provides a non-transitory computer-readable storage medium having a computer program stored thereon. When the computer program is executed by a processor, the following steps are implemented:

[0101] Acquire remote sensing optical data of a first resolution and remote sensing optical data of a second resolution, wherein the first resolution is higher than the second resolution;

[0102] Constructing a spatiotemporal window for each pixel of the remote sensing optical data of the first resolution;

[0103] Selecting the remote sensing optical data of the second resolution within the space-time window, and determining the correlation coefficient and conversion coefficient between each pixel thereof and the central pixel of the remote sensing optical data of the first resolution within the space-time window;

[0104] Determine simulated remote sensing optical data of the first resolution according to the remote sensing optical data of the second resolution corresponding to the correlation coefficient greater than a preset threshold and the conversion coefficient;

[0105] Reconstructed remote sensing optical data with the first resolution are determined according to the remote sensing optical data with the first resolution and the simulated remote sensing optical data with the first resolution.

[0106] The device embodiments described above are merely illustrative, wherein the units described as separate components may or may not be physically separated, and the components displayed as units may or may not be physical units, that is, they may be located in one place, or they may be distributed on multiple network units. Some or all of the modules may be selected according to actual needs to achieve the purpose of the scheme of this embodiment. Ordinary technicians in this field can understand and implement it without paying creative labor.

[0107] Through the description of the above implementation methods, those skilled in the art can clearly understand that each implementation method can be implemented by means of software plus a necessary general hardware platform, and of course, it can also be implemented by hardware. Based on this understanding, the above technical solution is essentially or the part that contributes to the prior art can be embodied in the form of a software product, and the computer software product can be stored in a computer-readable storage medium, such as ROM / RAM, a disk, an optical disk, etc., including a number of instructions for a computer device (which can be a personal computer, a server, or a network device, etc.) to execute the methods described in each embodiment or some parts of the embodiments.

[0108] It should be noted that the above embodiments are only used to illustrate the technical solutions of the present application, rather than to limit it. Although the present application has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or make equivalent replacements for some of the technical features therein. However, these modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the embodiments of the present application.

Claims

1. A data reconstruction method, It is characterized in that include: Acquire remote sensing optical data of a first resolution and remote sensing optical data of a second resolution, wherein the first resolution is higher than the second resolution; Constructing a spatiotemporal window for each pixel of the remote sensing optical data of the first resolution; Selecting the remote sensing optical data of the second resolution within the space-time window, and determining the correlation coefficient and conversion coefficient between each pixel thereof and the central pixel of the remote sensing optical data of the first resolution within the space-time window; Determine simulated remote sensing optical data of the first resolution according to the remote sensing optical data of the second resolution corresponding to the correlation coefficient greater than a preset threshold and the conversion coefficient; Determining reconstructed remote sensing optical data of the first resolution according to the remote sensing optical data of the first resolution and the simulated remote sensing optical data of the first resolution; The remote sensing optical data of the first resolution and the remote sensing optical data of the second resolution respectively include clear sky observation data and / or non-clear sky observation data, the clear sky observation data include all clear sky observation data and / or the clear sky part of part of the clear sky observation data, and the non-clear sky observation data include the non-clear sky part of the part of the clear sky observation data; The step of determining the reconstructed remote sensing optical data of the first resolution according to the remote sensing optical data of the first resolution and the simulated remote sensing optical data of the first resolution comprises: performing weighted averaging of the remote sensing optical data of the first resolution and the simulated remote sensing optical data of the first resolution to obtain the reconstructed remote sensing optical data of the first resolution; The step of weighted averaging the remote sensing optical data of the first resolution and the simulated remote sensing optical data of the first resolution to obtain the reconstructed remote sensing optical data of the first resolution comprises: If there is any pixel corresponding to the remote sensing optical data of the first resolution that is clear sky observation data, then determine the weight of the remote sensing optical data of the first resolution that is corresponding to the pixel as a first weight, wherein the first weight is higher than the weight corresponding to the simulated remote sensing optical data of the first resolution; The remote sensing optical data of the first resolution corresponding to any pixel and the simulated remote sensing optical data of the first resolution corresponding to any pixel are weightedly summed to obtain the reconstructed remote sensing optical data of the first resolution corresponding to any pixel; or If the remote sensing optical data of the first resolution corresponding to any pixel is non-clear sky observation data, the weight of the remote sensing optical data of the first resolution corresponding to the pixel is determined to be 0; The simulated remote sensing optical data of the first resolution corresponding to any one of the pixels is weighted to obtain the reconstructed remote sensing optical data of the first resolution corresponding to any one of the pixels.

2. The method according to claim 1, It is characterized in that The remote sensing optical data of the first resolution and the remote sensing optical data of the second resolution are remote sensing optical data of a full time period prepared at one time.

3. The method according to claim 1, It is characterized in that The step of constructing a spatiotemporal window for each pixel of the remote sensing optical data of the first resolution includes: For each pixel of the remote sensing optical data of the first resolution, a spatial window with a spatial range of ±w*±w is selected with the pixel as the center, and a ±t translation is performed in the time range to obtain a (2w+1)*(2w+1)*(2t+1) space-time window, wherein w is the length from the center pixel in the spatial range, and t is the length from the center pixel in the time range.

4. A data reconstruction device, It is characterized in that include: An acquisition unit, configured to acquire remote sensing optical data of a first resolution and remote sensing optical data of a second resolution, wherein the first resolution is higher than the second resolution; A processing unit, configured to construct a spatiotemporal window for each pixel of the remote sensing optical data of the first resolution; The processing unit is further used to select the remote sensing optical data of the second resolution within the space-time window, and determine the correlation coefficient and conversion coefficient between each pixel thereof and the central pixel of the remote sensing optical data of the first resolution within the space-time window; The processing unit is further used to determine the simulated remote sensing optical data of the first resolution according to the remote sensing optical data of the second resolution corresponding to the correlation coefficient greater than a preset threshold and the conversion coefficient; The processing unit is further used to determine reconstructed remote sensing optical data of the first resolution based on the remote sensing optical data of the first resolution and the simulated remote sensing optical data of the first resolution; The remote sensing optical data of the first resolution and the remote sensing optical data of the second resolution respectively include clear sky observation data and / or non-clear sky observation data, the clear sky observation data include all clear sky observation data and / or the clear sky part of part of the clear sky observation data, and the non-clear sky observation data include the non-clear sky part of the part of the clear sky observation data; The processing unit is specifically used to perform weighted averaging on the remote sensing optical data of the first resolution and the simulated remote sensing optical data of the first resolution to obtain the reconstructed remote sensing optical data of the first resolution; The processing unit is specifically configured to determine, if any pixel has remote sensing optical data of the first resolution corresponding to it that is clear sky observation data, a weight of the remote sensing optical data of the first resolution corresponding to the pixel as a first weight, wherein the first weight is higher than a weight corresponding to the simulated remote sensing optical data of the first resolution; performing weighted summation of the remote sensing optical data of the first resolution corresponding to any pixel and the simulated remote sensing optical data of the first resolution corresponding to any pixel to obtain the reconstructed remote sensing optical data of the first resolution corresponding to any pixel; or, If the remote sensing optical data of the first resolution corresponding to any pixel is non-clear sky observation data, the weight of the remote sensing optical data of the first resolution corresponding to the pixel is determined to be 0; The simulated remote sensing optical data of the first resolution corresponding to any one of the pixels is weighted to obtain the reconstructed remote sensing optical data of the first resolution corresponding to any one of the pixels.

5. A data reconstruction device, It is characterized in that The device comprises at least one processor, wherein the processor is configured to execute a program stored in a memory, and when the program is executed, the device executes: The method according to any one of claims 1 to 3.

6. A non-transitory computer-readable storage medium having a computer program stored thereon, It is characterized in that When the computer program is executed by a processor, the method according to any one of claims 1 to 3 is implemented.