Regional multi-season remote sensing feature set construction method and related device

Abstract

The invention provides a regional multi-season remote sensing feature set construction method and a related device, and relates to the field of remote sensing. Wherein the server divides a target geographic area into a plurality of sub-areas; respectively determining a vegetation growing season, a vegetation degeneration season, an earth surface low-temperature season and an earth surface high-temperature season of each sub-region; and according to the spectral characteristics of each sub-region in the vegetation growing season, the vegetation degeneration season, the surface low-temperature season and the surface high-temperature season, determining an optimal combination among the vegetation growing season, the vegetation degeneration season, the surface low-temperature season and the surface high-temperature season. Thus, the unique features of each region can be presented more accurately by constructing the seasonal features through the geographic space partitions, and the richness of information is increased by combining the multi-seasonal features, so that the information extraction effect is improved, and the feasibility of obtaining large-scale earth surface coverage is improved.

Description

Technical Field

[0001] The present application relates to the field of remote sensing, and in particular to a method for constructing a regional multi-season remote sensing feature set and related devices. Background Art

[0002] Interpreting land cover information based on remote sensing images is a complex but crucial task, which is particularly important for urban planning, agricultural management, forest monitoring and environmental protection. However, land surface information changes over time, especially in different seasons. In this regard, the accuracy of remote sensing data interpretation can be significantly improved through multi-source data fusion and multi-temporal feature synthesis technology. It can be understood that the extraction of seasonal features can help obtain data for a specific time period, and extract important information through feature synthesis methods, thereby making full use of the changes in the surface over time and the differences between objects and background, and improving the interpretation of surface cover.

[0003] In recent years, remote sensing interpretation technology has received increasing attention. How to effectively use multi-temporal remote sensing data to further improve the accuracy of interpretation and the reliability of information extraction has become a hot research issue. In the early days, land cover interpretation mainly relied on summer data because the surface differences in summer were large, especially vegetation-related information. However, using data from two seasons (such as summer and winter) can increase the diversity of features and thus improve interpretation accuracy. Studies have shown that features from different seasons have their own unique value. For example, winter data can improve the recognition accuracy of features such as water bodies and impervious surfaces.

[0004] However, in practice, it is found that there is great uncertainty in the definition of seasons within a large scale. In a small area, the seasonal information is relatively similar. However, in a large area, the similarity between seasonal dates and seasonal information is low, and it is impossible to use a unified feature construction standard to create a seasonal remote sensing feature set. In addition, seasonal dates are difficult to accurately determine at a large scale, which poses a challenge to the synthesis of seasonal remote sensing features. Although the method of obtaining seasonal information based on time series percentage is simple and direct, it needs to ensure the reliability of image quality, and this method is more suitable for classification tasks.

[0005] Therefore, how to utilize multi-temporal remote sensing data and combine the characteristics of large-scale partitioning to construct a remote sensing feature set with seasonal information is still a difficult problem that needs to be solved urgently. Summary of the invention

[0006] In order to overcome at least one of the shortcomings of the prior art, the present application provides a method for constructing a regional multi-season remote sensing feature set and a related device, specifically comprising:

[0007] In a first aspect, the present application provides a method for constructing a regional multi-season remote sensing feature set, the method comprising:

[0008] Divide the target geographic area into sub-regions;

[0009] Determine the vegetation growth season, vegetation degradation season, surface low temperature season and surface high temperature season of each sub-region respectively;

[0010] According to the spectral characteristics of each sub-region in the vegetation growth season, the vegetation degradation season, the low surface temperature season and the high surface temperature season, the optimal combination among the vegetation growth season, the vegetation degradation season, the low surface temperature season and the high surface temperature season is determined.

[0011] In a second aspect, the present application provides a device for constructing a regional multi-season remote sensing feature set, the device comprising:

[0012] A region division module, used for dividing a target geographical area into a plurality of sub-areas;

[0013] A season division module, used to respectively determine the vegetation growth season, vegetation degradation season, surface low temperature season and surface high temperature season of each sub-region;

[0014] The optimal combination module is used to determine the optimal combination between the vegetation growth season, the vegetation degradation season, the low surface temperature season and the high surface temperature season according to the spectral characteristics of each sub-area in the vegetation growth season, the vegetation degradation season, the low surface temperature season and the high surface temperature season.

[0015] In a third aspect, the present application provides a storage medium storing a computer program, which, when executed by a processor, implements the method for constructing a regional multi-seasonal remote sensing feature set.

[0016] In a fourth aspect, the present application provides an electronic device, comprising a processor and a memory, wherein the memory stores a computer program, and when the computer program is executed by the processor, the method for constructing a regional multi-season remote sensing feature set is implemented.

[0017] Compared with the prior art, this application has the following beneficial effects:

[0018] The present application provides a method for constructing a regional multi-season remote sensing feature set and a related device. Among them, the server divides the target geographical area into multiple sub-areas; determines the vegetation growth season, vegetation degradation season, surface low temperature season and surface high temperature season of each sub-area respectively; and determines the best combination of vegetation growth season, vegetation degradation season, surface low temperature season and surface high temperature season according to the spectral characteristics of each sub-area in the vegetation growth season, vegetation degradation season, surface low temperature season and surface high temperature season. In this way, constructing seasonal features through geographic space partitioning can more accurately present the unique characteristics of each region, and combining multi-season features increases the richness of information, thereby improving the information extraction effect and the feasibility of large-scale surface coverage acquisition. BRIEF DESCRIPTION OF THE DRAWINGS

[0019] In order to more clearly illustrate the technical solutions of the embodiments of the present application, the drawings required for use in the embodiments will be briefly introduced below. It should be understood that the following drawings only show certain embodiments of the present application and therefore should not be regarded as limiting the scope. For ordinary technicians in this field, other related drawings can be obtained based on these drawings without paying creative work.

[0020] Figure 1 A schematic diagram of a flow chart of a method for constructing a regional multi-seasonal remote sensing feature set provided in an embodiment of the present application;

[0021] Figure 2 The detailed schematic diagram of the method for constructing a regional multi-seasonal remote sensing feature set provided for the embodiment of the present application is only;

[0022] Figure 3 A second detailed schematic diagram of the method for constructing a regional multi-seasonal remote sensing feature set provided in an embodiment of the present application;

[0023] Figure 4 A schematic diagram of a vegetation index sequence provided in an embodiment of the present application;

[0024] Figure 5 A schematic diagram of vegetation degradation and growing season provided for an embodiment of the present application;

[0025] Figure 6 A schematic diagram of a land surface temperature sequence provided in an embodiment of the present application;

[0026] Figure 7 A schematic diagram of low temperature and high temperature seasons on the ground provided in an embodiment of the present application;

[0027] Figure 8 A third detailed schematic diagram of the method for constructing a regional multi-seasonal remote sensing feature set provided in an embodiment of the present application;

[0028] Fig. 9A schematic diagram of the model training effect provided in the embodiment of the present application;

[0029] Fig.10 A schematic diagram of the combined effect of spectral features provided in an embodiment of the present application;

[0030] Fig.11 A schematic diagram of the structure of a device for constructing a domain multi-season remote sensing feature set provided in an embodiment of the present application;

[0031] Fig.12 A schematic diagram of the structure of an electronic device provided in an embodiment of the present application. DETAILED DESCRIPTION

[0032] In order to make the purpose, technical solution and advantages of the embodiments of the present application clearer, the technical solution in the embodiments of the present application will be clearly and completely described below in conjunction with the drawings in the embodiments of the present application. Obviously, the described embodiments are part of the embodiments of the present application, rather than all the embodiments. The components of the embodiments of the present application described and shown in the drawings here can be arranged and designed in various different configurations.

[0033] Therefore, the following detailed description of the embodiments of the present application provided in the accompanying drawings is not intended to limit the scope of the present application for which protection is sought, but merely represents selected embodiments of the present application. Based on the embodiments in the present application, all other embodiments obtained by ordinary technicians in the field without creative work are within the scope of protection of the present application.

[0034] It should be noted that similar reference numerals and letters denote similar items in the following drawings, and therefore, once an item is defined in one drawing, further definition and explanation thereof is not required in subsequent drawings.

[0035] In the description of the present application, it should be noted that the terms "first", "second", "third", etc. are only used to distinguish the description and cannot be understood as indicating or implying relative importance. In addition, the terms "comprise", "include" or any other variants thereof are intended to cover non-exclusive inclusion, so that a process, method, article or device including a series of elements includes not only those elements, but also includes other elements not explicitly listed, or also includes elements inherent to such process, method, article or device. In the absence of further restrictions, an element defined by the sentence "comprises a ..." does not exclude the presence of other identical elements in the process, method, article or device including the element.

[0036] Based on the above statement, as introduced in the background technology, how to utilize multi-temporal remote sensing data and combine the characteristics of large-scale partitioning to construct a remote sensing feature set with seasonal information is still a difficult problem to be solved.

[0037] Based on the discovery of the above technical problems, the inventors have proposed the following technical solutions to solve or improve the above problems through creative work. It should be noted that the defects in the solutions in the above prior art are the results obtained by the inventors after practice and careful research. Therefore, the discovery process of the above problems and the solutions proposed in the embodiments of the present application for the above problems below should all be the contributions made by the inventors to the present application in the process of invention and creation, and should not be understood as technical contents known to those skilled in the art.

[0038] In view of the above technical problems, this embodiment provides a method for constructing a regional multi-season remote sensing feature set. Figure 1 As shown, the method includes:

[0039] S1, divide the target geographical area into multiple sub-areas.

[0040] S2, determine the vegetation growth season, vegetation degradation season, surface low temperature season and surface high temperature season for each sub-region respectively.

[0041] S3, according to the spectral characteristics of each sub-region in the vegetation growth season, vegetation degradation season, surface low temperature season and surface high temperature season, determine the best combination among the vegetation growth season, vegetation degradation season, surface low temperature season and surface high temperature season.

[0042] In this way, constructing seasonal features through geographic spatial partitioning can more accurately present the unique characteristics of each region, and combining multi-seasonal features increases the richness of information, thereby improving the information extraction effect and the feasibility of large-scale surface cover acquisition.

[0043] In addition, it should be understood that the electronic device for implementing the above-mentioned regional multi-season remote sensing feature set construction method can be, but is not limited to, a mobile terminal, a tablet computer, a laptop computer, a desktop computer, and a server, etc. The server can be a single server or a server group. The server group can be centralized or distributed (for example, the server can be a distributed system). In some embodiments, the server can be local or remote relative to the user terminal. In some embodiments, the server can be implemented on a cloud platform; as an example only, the cloud platform can include a private cloud, a public cloud, a hybrid cloud, a community cloud (Community Cloud), a distributed cloud, an inter-cloud (Inter-Cloud), a multi-cloud (Multi-Cloud), etc., or any combination thereof. In some embodiments, the server can be implemented on an electronic device having one or more components.

[0044] To make the solution provided by this embodiment clearer, the following uses a server as an electronic device for implementing the method for constructing a regional multi-season remote sensing feature set. Figure 1 Each step in the method shown is described in detail. However, it should be understood that the operations of the flowchart may not be implemented in order, and steps without logical contextual relationships may be reversed or implemented simultaneously. In addition, those skilled in the art may add one or more other operations to the flowchart, or remove one or more operations from the flowchart, under the guidance of the content of this application. Based on the above statement, continue to refer to Figure 1 , the method comprising:

[0045] S1, divide the target geographical area into multiple sub-areas.

[0046] It should be understood that the target geographic area includes multiple original partitions. This embodiment obtains the vegetation index sequence and surface temperature sequence of each original partition to fully understand the dynamic changes of each partition in different time periods. And by using these sequences to cluster multiple original partitions, those partitions with similar vegetation growth patterns and temperature characteristics can be grouped into the same group. Therefore, as an optional implementation, Figure 2 As shown, for Figure 1 Step S1 in the embodiment may include the following implementations:

[0047] S1-1, obtain the vegetation index series and surface temperature series of each original partition.

[0048] S1-2, clustering multiple original partitions using the vegetation index sequences and surface temperature sequences of multiple original partitions to obtain multiple similar partition sets.

[0049] S1-3, merge the original partitions in each similar partition set to obtain multiple sub-regions.

[0050] Optionally, the server may perform hierarchical clustering on the multiple original partitions using the vegetation index sequences and the surface temperature sequences of the multiple original partitions to obtain multiple similar partition sets.

[0051] The so-called hierarchical clustering method refers to building a hierarchical structure by gradually merging or segmenting clusters. This process can be carried out by a bottom-up aggregation method (i.e. starting from a single sample and merging step by step) or a top-down decomposition method (i.e. starting from all samples as a cluster and splitting step by step). In the specific clustering process, the similarity or distance between each original partition is first calculated, and the commonly used metrics include Euclidean distance, Mahalanobis distance, etc. Then, the two partitions with the highest similarity are selected to merge to form a new cluster. This process can be repeated until a predetermined stopping condition is reached, such as reaching a specific number of clusters or a similarity threshold. Therefore, each sub-region in the present embodiment is also referred to as a reorganization partition.

[0052] Therefore, the server can first collect the vegetation index of each original partition at different time points, such as NDVI (Normalized Difference Vegetation Index) or EVI (Enhanced Vegetation Index) and surface temperature data series. These sequences reflect the vegetation health and temperature characteristics of each partition over time. Then, the server uses a clustering algorithm (such as K-means or hierarchical clustering, etc.) to process these data sequences and group them according to the similarity of vegetation index and surface temperature between partitions. In this way, partitions with similar vegetation index and surface temperature change trends will be classified into the same cluster, forming multiple similar partition sets.

[0053] In this way, the similar partition set obtained in this way can better reflect the common characteristics of vegetation growth and surface temperature in different geographical regions, thus helping to improve the accuracy and consistency of subsequent analysis and interpretation. In addition, the partitions in the similar partition set have more consistent surface characteristics, which helps to reduce the complexity of modeling, thus better capturing the changing patterns of vegetation and surface temperature in a specific area.

[0054] Based on the description of the multiple sub-regions in step S1 in the above embodiment, the following is a description of the multiple sub-regions in step S1. Figure 1 Step S1 in the following is described:

[0055] S2, determine the vegetation growth season, vegetation degradation season, surface low temperature season and surface high temperature season for each sub-region respectively.

[0056] The study found that by dividing a year into different seasons, namely vegetation growth season, vegetation degradation season, low surface temperature season and high surface temperature season, it is possible to more finely analyze the vegetation growth status and surface temperature changes of each sub-region in different time periods. Figure 3 As shown, the implementation of step S2 may include:

[0057] S2-1, for each sub-region, obtaining the enhanced vegetation index sequence and the surface temperature sequence of the sub-region.

[0058] S2-2, according to the vegetation index sequence, determine the vegetation growth season and vegetation degradation season of the sub-region.

[0059] Optionally, the server can obtain the vegetation index difference between the maximum vegetation index and the minimum vegetation index according to the vegetation index sequence; obtain the first index adjustment amount and the second index adjustment amount according to the vegetation index difference, and obtain the first index threshold and the second index threshold according to the minimum vegetation index, the first index adjustment amount and the second index adjustment amount; wherein the first index threshold is the sum of the minimum vegetation index and the first index adjustment amount, and the second index threshold is the sum of the minimum vegetation index and the second index adjustment amount; the time period corresponding to the vegetation index less than the first index threshold is regarded as the vegetation degradation season; and the time period corresponding to the vegetation index greater than the second index threshold is regarded as the vegetation growth season.

[0060] For the above implementation, it should be understood that the vegetation index sequence can reflect the seasonal changes of the surface, and the vegetation index sequence is used as a seasonal guidance feature to determine the seasonal date for each sub-region. Figure 4 As shown, the figure shows the vegetation index (EVI) series of two different sub-areas.

[0061] In the specific operation process, the spectral characteristics of different seasons can be obtained in the MOD09A1 dataset, and the effective pixels are set as pixels with vegetation index between 10% and 90%, and the mask method is used to screen and eliminate them.

[0062] Then, for the vegetation index sequence of each sub-region, the threshold method can be used to perform seasonal segmentation. The server can obtain the maximum vegetation index, the minimum vegetation index, and the vegetation index difference between the maximum and minimum values ​​of the vegetation index sequence; and use 1 / 4 of the vegetation index difference as the first index adjustment amount, and 3 / 4 as the second index adjustment amount, and then add them to the minimum value to obtain the first index threshold and the second index threshold. The mathematical expression can be expressed as:

[0063] V EVI =Value EVImax -Value EVImin

[0064]

[0065] In the formula, Value represents the value in the vegetation index sequence, Value EVImax Indicates the maximum value in the vegetation index sequence, Value EVImin Represents the minimum value in the vegetation index sequence, represents the first exponential threshold, Indicates the second index threshold; Period drop and Period grow They are the periods corresponding to the vegetation degradation season and the vegetation growth season respectively.

[0066] Finally, the server determines the time period corresponding to the vegetation index less than the first index threshold as the vegetation degradation season, and determines the time period corresponding to the vegetation index greater than the second index threshold as the vegetation growth season.

[0067] like Figure 5 As shown, the figure shows the vegetation degradation season and vegetation growth season of multiple sub-regions obtained by the above implementation.

[0068] For ease of understanding, the following example illustrates the discrimination method of vegetation degradation season and vegetation growth season. Exemplarily, the difference between the maximum vegetation index and the minimum vegetation index in the vegetation index sequence is first calculated. Assuming that the maximum vegetation index is 0.8 and the minimum vegetation index is 0.4, the difference is 0.4. Next, the first index adjustment amount and the second index adjustment amount are defined. Assuming that the first index adjustment amount is 0.1 and the second index adjustment amount is 0.3. Then, the time point corresponding to the minimum vegetation index is found (assuming March 1), and the time point corresponding to the first index threshold (minimum vegetation index plus the first index adjustment amount, i.e. 0.5) is calculated (assuming February 1 and April 1), and the time point corresponding to the second index threshold (minimum vegetation index plus the second index adjustment amount, i.e. 0.7) (assuming July 1 and October 1), and the time point corresponding to the maximum vegetation index (assuming September 1).

[0069] Based on these time points, the period between February 1 and April 1 is considered as the vegetation degradation season, during which the vegetation index value is low. Then, the period between July 1 and October 1 is considered as the vegetation growth season, during which the vegetation index value is large. In this way, the vegetation growth season and decline season of each sub-region can be accurately defined according to the change pattern of the vegetation index, so as to better understand and manage the vegetation dynamics in these areas.

[0070] S2-3, according to the surface temperature sequence, determining the surface low temperature season and the surface high temperature season of the sub-region.

[0071] Optionally, the server may obtain the surface temperature difference between the maximum surface temperature and the minimum surface temperature according to the surface temperature sequence; obtain the first temperature adjustment amount and the second temperature adjustment amount according to the surface temperature difference, and obtain the first temperature threshold and the second temperature threshold according to the first temperature adjustment amount, the second temperature adjustment amount and the minimum surface temperature. The first temperature threshold is the sum of the minimum surface temperature and the first temperature adjustment amount, and the second temperature threshold is the sum of the minimum surface temperature and the second temperature adjustment amount; the time period corresponding to the surface temperature less than the first temperature threshold is regarded as the surface low temperature season; the time period corresponding to the surface temperature greater than the second temperature threshold is regarded as the surface high temperature season.

[0072] For the above embodiment, it should be understood that the surface temperature sequence can reflect the seasonal changes of the surface, and therefore can be used as a seasonal guidance feature to determine the seasonal date for each sub-region. Figure 6 As shown, the figure shows the land surface temperature (LST) series of two different sub-regions.

[0073] In the specific operation, the surface temperature characteristics of different seasons can be obtained in the MOD11A2 dataset. The effective pixels are set as pixels with surface temperatures between 10% and 90%, and the masks are used to screen and eliminate them.

[0074] Then, for each sub-region's surface temperature series, the threshold method is used to perform seasonal segmentation, and the maximum, minimum, and difference between the maximum and minimum values ​​of the LST curve are obtained.

[0075] Get each surface temperature sequence and use the threshold method to perform season segmentation. The server can obtain the maximum surface temperature, minimum surface temperature and the surface temperature difference between the two of each sub-region surface temperature sequence; and use the 1 / 4 of the surface temperature difference as the first temperature adjustment amount and the 3 / 4 as the second temperature adjustment amount, and then add them to the minimum surface temperature to obtain the first temperature threshold and the second temperature threshold. The mathematical expression can be expressed as:

[0076] V LST =Value LSTmax -Value LSTmin

[0077]

[0078] Where Value represents the value in the surface temperature series, Value LSTmax Indicates the maximum value in the surface temperature series, Value LSTmin represents the minimum value in the surface temperature series, represents the first temperature threshold, Indicates the second temperature threshold; Period low and Period high The distribution is the period corresponding to the surface low temperature season and the surface high temperature season.

[0079] Finally, the server regards the period corresponding to the surface temperature less than the first temperature threshold as the surface low temperature season; and regards the period corresponding to the surface temperature greater than the second temperature threshold as the surface high temperature season.

[0080] like Figure 7 As shown, the figure shows the surface low temperature season and the surface high temperature season of multiple sub-regions obtained by the above implementation.

[0081] For ease of understanding, the following is an example of how to distinguish between the low-temperature season and the high-temperature season on the surface. First, calculate the difference between the maximum and minimum surface temperatures in the surface temperature sequence. Assuming that the maximum surface temperature is 30°C and the minimum surface temperature is -10°C, the difference is 40°C. Next, define the first temperature adjustment and the second temperature adjustment, assuming that the first temperature adjustment is 10°C and the second temperature adjustment is 30°C. Then, find the time point corresponding to the minimum surface temperature (assuming February 1), calculate the time point corresponding to the first temperature threshold (minimum surface temperature plus the first temperature adjustment, i.e. 0°C) (assuming March 1 and November 1), the time point corresponding to the second temperature threshold (minimum surface temperature plus the second temperature adjustment, i.e. 20°C) (assuming May 1 and October 1), and the time point corresponding to the maximum surface temperature (assuming August 1).

[0082] Based on these time points, the period between January 1 and March 1 and the period between November 1 and December 31 are defined as the low-temperature surface season, during which the surface temperature is low, reflecting the low temperature state of the surface. Then, the period between May 1 and October 1 is defined as the high-temperature surface season, during which the surface temperature is high, reflecting the high temperature state of the surface. In this way, the low-temperature surface season and the high-temperature surface season of each sub-region can be accurately defined according to the variation pattern of the surface temperature, so as to better understand and manage the surface temperature dynamics of these regions.

[0083] Based on the description of the vegetation growth season, vegetation degradation season, surface low temperature season and surface high temperature season in the above embodiment, the following will continue to describe Figure 1 Step S3 in the following is described:

[0084] S3, according to the spectral characteristics of each sub-region in the vegetation growth season, vegetation degradation season, surface low temperature season and surface high temperature season, determine the best combination among the vegetation growth season, vegetation degradation season, surface low temperature season and surface high temperature season.

[0085] Based on the above seasonal division, the server can use the multi-temporal feature fusion algorithm to generate a remote sensing feature set that includes the spectral feature set of the vegetation growth season and the vegetation degradation season, as well as the temperature feature set of the surface high temperature season and the surface low temperature season. Multi-temporal data include multi-temporal MODIS vegetation index (MOD13Q1), multi-temporal MODIS surface temperature (MOD11A2) and multi-temporal MODIS surface reflectance (MOD09A1). MODIS refers to the sensor on the Aqua satellite, which can provide a series of comprehensive data on the earth's surface and atmosphere. MOD13Q1, MOD11A2 and MOD09A1 represent different types of sensors.

[0086] Specifically, for each sub-region, the server constructs a multi-temporal feature fusion algorithm at the partition level, decomposes the multi-temporal remote sensing data into seasonal scales using regional seasonal information, and synthesizes remote sensing features with seasonal differences. As shown in the table below, the vegetation growth season and vegetation degradation season are used to extract the multi-temporal MOD09A1 data features of each partition, and the mean synthesis method is used to generate seasonal B1 to B7 spectral features; the surface high temperature season and surface low temperature season are used to extract the multi-temporal MOD11A2 data features of each partition, and the mean synthesis method is used to generate seasonal daytime and nighttime LST temperature features.

[0087] Band i,growPeriod = MEAN[Band i,growPeriod ]

[0088] Band i,dropPeriod = MEAN[Band i,dropPeriod ]

[0089] LST j,lowPeriod = MEAN[LST j,lowPeriod ]

[0090] LST j,highPeriod = MEAN[LST j,highPeriod ]

[0091] Wherein, i represents the i band of MOD09A1 data, i=1,...,7, and j represents the j band of MOD11A2 data, j=daytime,nighttime.

[0092] Feature Set describe Vegetation growing season spectral feature set <![CDATA[Band i,growPeriod ,i=1,2,3,4,5,6,7]]> Spectral feature set of vegetation degradation season <![CDATA[Band i，dropPeriod ,i=1,2,3,4,5,6,7]]> Surface temperature characteristics in hot season <![CDATA[LST j,highPeriod ,j=daytime,ighttime]]> Surface low temperature season characteristic set <![CDATA[LST j，lowPeriod ,j=daytime,nighttime]]>

[0093] Based on the above analysis of spectral characteristics in different seasons, the study found that different seasonal combinations can capture the surface characteristics of different time periods. For example, the spectral characteristics of the vegetation growth season and the vegetation degradation season may be very different from the spectral characteristics of the surface low temperature season and the surface high temperature season. By considering multiple seasonal combinations, these characteristics can be analyzed comprehensively to ensure that no important seasonal information is missed. In addition, by training models for each seasonal combination, the ability of different seasonal combinations to explain changes in surface cover can be evaluated. For example, some seasonal combinations may perform better in certain sub-regions, while other combinations may be more effective in other sub-regions. By training multiple models, we can more comprehensively evaluate the effects of different seasonal combinations. Therefore, Figure 8 As shown, as an optional implementation, step S3 may include:

[0094] S3-1, obtain various seasonal combinations among vegetation growth season, vegetation degradation season, surface low temperature season and surface high temperature season.

[0095] Exemplarily, this embodiment provides the following combinations:

[0096]

[0097] S3-2, for each seasonal combination, the spectral features of each sub-region corresponding to the seasonal combination are used to perform model training to obtain a test model corresponding to the seasonal combination.

[0098] S3-3, select the target model with the best performance from multiple test models corresponding to multiple seasonal combinations.

[0099] S3-4, determining the season combination corresponding to the target model as the best combination.

[0100] Exemplarily, the spectral features in the above example are used as training samples to verify the above combination. This embodiment adopts two models: a partition model and a global model. The global model uses samples from the entire study area to build a model; the partition model uses samples within each geographical partition to build a model. Both models use random forest regression models, and the number of trees is set to 100.

[0101] The zoning model is constructed based on five sub-regions. The input sample uses the percentage of impervious surface as the training target of the model, because the percentage of impervious surface is one of the most representative indicators of surface cover types. The ratio of training samples to test samples is 8:2, and the input features are the spectral feature set of the vegetation growing season.

[0102] The server obtains multiple season combinations in the above table, including different combinations of vegetation growth season, vegetation degradation season, surface low temperature season and surface high temperature season. For each season combination, the spectral characteristics of each geographical region in the season combination are used to train the model and obtain the test model corresponding to the season combination. From the multiple test models corresponding to the multiple season combinations, the target model with the best performance is selected. Finally, the season combination corresponding to the target model is determined as the best combination.

[0103] like Fig. 9 As shown in the figure, 4 out of 5 partition models have better goodness of fit than the global model. The average goodness of fit of the partition model is 0.7469, while the goodness of fit of the global model is 0.7281. Overall, the partition model is better than the global model.

[0104] In addition, in order to explore the effectiveness of the feature combination scheme, this embodiment also applies different feature combination schemes to the partition model. Fig.10It can be seen that with the addition of surface temperature features, the goodness of fit of each partition model has improved. When both the dual-season spectral features and the surface temperature features are included in the partition model, the goodness of fit of the partition model reaches a maximum of 0.7811. Therefore, overall, the addition of seasonal spectral features and surface temperature features can significantly improve the accuracy of remote sensing interpretation.

[0105] Based on the same inventive concept as the method for constructing a regional multi-seasonal remote sensing feature set provided in this embodiment, this embodiment also provides a device for constructing a regional multi-seasonal remote sensing feature set, which includes at least one software function module that can be stored in the memory 120 or fixed in the electronic device in the form of software. The processor 130 in the electronic device is used to execute the executable module stored in the memory 120. For example, the software function modules and computer programs included in the device. Please refer to Fig.11 , functionally speaking, the device may include:

[0106] A region division module 11, used for dividing a target geographical region into a plurality of sub-regions;

[0107] A season division module 12 is used to respectively determine the vegetation growth season, vegetation degradation season, surface low temperature season and surface high temperature season of each sub-region;

[0108] The best combination module 13 is used to determine the best combination between the vegetation growth season, the vegetation degradation season, the surface low temperature season and the surface high temperature season according to the spectral characteristics of each sub-region in the vegetation growth season, the vegetation degradation season, the surface low temperature season and the surface high temperature season.

[0109] In this embodiment, the above-mentioned area division module 11 is used to implement Figure 1 In step S1, the seasonal division module 12 is used to implement Figure 1 In step S2, the best combination module 13 is used to implement Figure 1 Therefore, for the detailed description of the above modules, please refer to the specific implementation of the corresponding steps.

[0110] In addition, since the invention concept is the same as that of the above-mentioned method for constructing a regional multi-season remote sensing feature set, the device for constructing a regional multi-season remote sensing feature set can also implement other steps or sub-steps of the method through the above-mentioned modules.

[0111] Optionally, the optimal combination module 13 is further specifically used for:

[0112] Obtain multiple seasonal combinations among vegetation growth season, vegetation degradation season, surface low temperature season and surface high temperature season;

[0113] For each seasonal combination, the spectral features of each sub-region corresponding to the seasonal combination are used for model training to obtain a test model corresponding to the seasonal combination;

[0114] Select the target model with the best performance from multiple test models corresponding to multiple seasonal combinations;

[0115] The season combination corresponding to the target model is determined as the best combination.

[0116] Optionally, the season division module 12 is further specifically used for:

[0117] For each sub-region, obtain the enhanced vegetation index series and the surface temperature series of the sub-region;

[0118] According to the vegetation index sequence, the vegetation growth season and vegetation degradation season of the sub-region are determined;

[0119] According to the surface temperature sequence, the surface low temperature season and surface high temperature season of the sub-region are determined.

[0120] Optionally, the season division module 12 is further specifically used for:

[0121] According to the vegetation index sequence, the vegetation index difference between the maximum vegetation index and the minimum vegetation index is obtained;

[0122] According to the vegetation index difference, a first index adjustment amount and a second index adjustment amount are obtained, wherein the second index adjustment amount is greater than the first index adjustment amount;

[0123] According to the minimum vegetation index, the first index adjustment amount and the second index adjustment amount, a first index threshold and a second index threshold are obtained, wherein the first index threshold is the sum of the minimum vegetation index and the first index adjustment amount, and the second index threshold is the sum of the minimum vegetation index and the second index adjustment amount;

[0124] The period corresponding to the vegetation index less than the first index threshold is regarded as the vegetation degradation season; the period corresponding to the vegetation index greater than the second index threshold is regarded as the vegetation growth season.

[0125] Optionally, the season division module 12 is further specifically used for:

[0126] According to the surface temperature sequence, the surface temperature difference between the maximum surface temperature and the minimum surface temperature is obtained;

[0127] According to the surface temperature difference, a first temperature adjustment amount and a second temperature adjustment amount are obtained, wherein the second temperature adjustment amount is greater than the first temperature adjustment amount;

[0128] According to the minimum surface temperature, the second temperature adjustment amount and the first temperature adjustment amount, a first temperature threshold and a second temperature threshold are obtained, wherein the first temperature threshold is the sum of the minimum surface temperature and the first temperature adjustment amount, and the second temperature threshold is the sum of the minimum surface temperature and the second temperature adjustment amount;

[0129] The time period corresponding to the surface temperature being less than the first temperature threshold is regarded as the surface low temperature season; the time period corresponding to the surface temperature being greater than the second temperature threshold is regarded as the surface high temperature season.

[0130] Optionally, the target geographical area includes a plurality of original partitions, and the area division module 11:

[0131] Obtain the vegetation index series and surface temperature series of each original partition;

[0132] The vegetation index sequences and surface temperature sequences of multiple original partitions are used to cluster multiple original partitions to obtain multiple similar partition sets.

[0133] The original partitions in each similar partition set are merged to obtain multiple sub-regions.

[0134] Optionally, the area division module 11 is further specifically used for:

[0135] The vegetation index sequences and surface temperature sequences of multiple original partitions are used to perform hierarchical clustering on multiple original partitions to obtain multiple similar partition sets.

[0136] In addition, the functional modules in the various embodiments of the present application may be integrated together to form an independent part, or each module may exist separately, or two or more modules may be integrated to form an independent part.

[0137] It should also be understood that if the above implementation is implemented in the form of a software function module 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 application, or the part that contributes to the prior art, or the part of the technical solution, can be embodied in the form of a software product, which is stored in a storage medium and includes several instructions for a computer device (which can be a personal computer, a server, or a network device, etc.) to perform all or part of the steps of the method described in each embodiment of the present application.

[0138] Therefore, this embodiment also provides a storage medium, which is a computer-readable storage medium. The storage medium stores a computer program, and when the computer program is executed by a processor, the method for constructing a regional multi-seasonal remote sensing feature set provided in this embodiment is implemented. Among them, the storage medium can be a U disk, a mobile hard disk, a read-only memory (ROM, Read-Only Memory), a random access memory (RAM, Random Access Memory), a disk or an optical disk, etc., which can store program codes.

[0139] This embodiment provides an electronic device for implementing a method for constructing a regional multi-season remote sensing feature set. Fig.12 As shown, the electronic device may include a processor 22 and a memory 21. In addition, the memory 21 stores a computer program, and the processor 22 implements the method for constructing a regional multi-seasonal remote sensing feature set provided in this embodiment by reading and executing the computer program corresponding to the above implementation in the memory 21.

[0140] Continue to see Fig.12 The electronic device further includes a communication unit 23. The memory 21, the processor 22 and the communication unit 23 are electrically connected to each other directly or indirectly through a system bus 24 to achieve data transmission or interaction.

[0141] The memory 21 may be an information recording device based on any electronic, magnetic, optical or other physical principle, used to record execution instructions, data, etc. In some embodiments, the memory 21 may be, but is not limited to, a volatile memory, a non-volatile memory, a storage drive, etc.

[0142] In some embodiments, the volatile memory may be a random access memory (RAM); in some embodiments, the non-volatile memory may be a read-only memory (ROM), a programmable read-only memory (PROM), an erasable programmable read-only memory (EPROM), an electrically erasable read-only memory (EEPROM), a flash memory, etc.; in some embodiments, the storage drive may be a disk drive, a solid-state drive, any type of storage disk (such as a CD, a DVD, etc.), or a similar storage medium, or a combination thereof, etc.

[0143] The communication unit 23 is used to send and receive data through a network. In some embodiments, the network may include a wired network, a wireless network, a fiber optic network, a telecommunication network, an intranet, the Internet, a local area network (LAN), a wide area network (WAN), a wireless local area network (WLAN), a metropolitan area network (MAN), a wide area network (WAN), a public switched telephone network (PSTN), a Bluetooth network, a ZigBee network, or a near field communication (NFC) network, or any combination thereof. In some embodiments, the network may include one or more network access points. For example, the network may include a wired or wireless network access point, such as a base station and / or a network switching node, through which one or more components of the service request processing system may be connected to the network to exchange data and / or information.

[0144] The processor 22 may be an integrated circuit chip having a signal processing capability, and the processor 22 may include one or more processing cores (e.g., a single-core processor or a multi-core processor). By way of example only, the processor 22 may include a central processing unit (CPU), an application-specific integrated circuit (ASIC), an application-specific instruction set processor (ASIP), a graphics processing unit (GPU), a physical processing unit (PPU), a digital signal processor (DSP), a field programmable gate array (FPGA), a programmable logic device (PLD), a controller, a microcontroller unit, a reduced instruction set computer (RISC), or a microprocessor, or any combination thereof.

[0145] Understandably, Fig.12The structure shown is for illustration only. The electronic device may also have Fig.12 More or fewer components than shown, or with Fig.12 Different configurations are shown. Fig.12 The components shown may be implemented in hardware, software or a combination thereof.

[0146] It should be understood that the apparatus and method disclosed in the above-mentioned embodiments can also be implemented in other ways. The apparatus embodiments described above are merely schematic. For example, the flowcharts and block diagrams in the accompanying drawings show the possible architecture, functions and operations of the apparatus, methods and computer program products according to the multiple embodiments of the present application. In this regard, each box in the flowchart or block diagram can represent a module, a program segment or a part of a code, and the module, a program segment or a part of a code contains one or more executable instructions for implementing the specified logical function. It should also be noted that in some alternative implementations, the functions marked in the box can also occur in a different order from the order marked in the accompanying drawings. For example, two consecutive boxes can actually be executed substantially in parallel, and they can sometimes be executed in the opposite order, depending on the functions involved. It should also be noted that each box in the block diagram and / or the flowchart, and the combination of boxes in the block diagram and / or the flowchart can be implemented with a dedicated hardware-based system that performs a specified function or action, or can be implemented with a combination of dedicated hardware and computer instructions.

[0147] The above are only various implementations of the present application, but the protection scope of the present application is not limited thereto. Any person skilled in the art who is familiar with the present technical field can easily think of changes or substitutions within the technical scope disclosed in the present application, which should be included in the protection scope of the present application. Therefore, the protection scope of the present application should be based on the protection scope of the claims.

Claims

1. A method for constructing a regional multi-season remote sensing feature set, characterized in that: The method comprises: Divide the target geographic area into sub-regions; Determine the vegetation growth season, vegetation degradation season, surface low temperature season and surface high temperature season of each sub-region respectively; According to the spectral characteristics of each sub-region in the vegetation growth season, the vegetation degradation season, the low surface temperature season and the high surface temperature season, the optimal combination among the vegetation growth season, the vegetation degradation season, the low surface temperature season and the high surface temperature season is determined.

2. The method for constructing a regional multi-seasonal remote sensing feature set according to claim 1, characterized in that: Determining the best combination of the vegetation growth season, the vegetation degradation season, the surface low temperature season and the surface high temperature season according to the spectral characteristics of each of the sub-regions in the vegetation growth season, the vegetation degradation season, the surface low temperature season and the surface high temperature season, including: Acquire multiple seasonal combinations among the vegetation growth season, the vegetation degradation season, the surface low temperature season and the surface high temperature season; For each of the seasonal combinations, model training is performed using the spectral features of each of the sub-regions corresponding to the seasonal combination to obtain a test model corresponding to the seasonal combination; Selecting a target model with the best performance from a plurality of test models corresponding to the plurality of season combinations; The season combination corresponding to the target model is determined as the optimal combination.

3. The method for constructing a regional multi-seasonal remote sensing feature set according to claim 1, characterized in that: The vegetation growth season, vegetation degradation season, surface low temperature season and surface high temperature season of each sub-region are determined respectively, including: For each of the sub-regions, obtaining an enhanced vegetation index sequence and a surface temperature sequence of the sub-region; Determining the vegetation growth season and the vegetation degradation season of the sub-region according to the vegetation index sequence; According to the surface temperature sequence, the surface low temperature season and the surface high temperature season of the sub-region are determined.

4. The method for constructing a regional multi-seasonal remote sensing feature set according to claim 3, characterized in that: Determining the vegetation growth season and vegetation degradation season of the sub-region according to the vegetation index sequence includes: According to the vegetation index sequence, a vegetation index difference between a maximum vegetation index and a minimum vegetation index is obtained; According to the vegetation index difference, a first index adjustment amount and a second index adjustment amount are obtained, wherein the second index adjustment amount is greater than the first index adjustment amount; According to the minimum vegetation index, the first index adjustment amount and the second index adjustment amount, a first index threshold and a second index threshold are obtained, wherein the first index threshold is the sum of the minimum vegetation index and the first index adjustment amount, and the second index threshold is the sum of the minimum vegetation index and the second index adjustment amount; The time period corresponding to the vegetation index less than the first index threshold is regarded as the vegetation degradation season; and the time period corresponding to the vegetation index greater than the second index threshold is regarded as the vegetation growth season.

5. The method for constructing a regional multi-seasonal remote sensing feature set according to claim 3, characterized in that: Determining the surface low temperature season and the surface high temperature season of the sub-region according to the surface temperature sequence includes: According to the surface temperature sequence, a surface temperature difference between a maximum surface temperature and a minimum surface temperature is obtained; Obtaining a first temperature adjustment amount and a second temperature adjustment amount according to the surface temperature difference, wherein the second temperature adjustment amount is greater than the first temperature adjustment amount; Obtaining a first temperature threshold and a second temperature threshold according to the minimum ground surface temperature, the second temperature adjustment amount, and the first temperature adjustment amount, wherein the first temperature threshold is the sum of the minimum ground surface temperature and the first temperature adjustment amount, and the second temperature threshold is the sum of the minimum ground surface temperature and the second temperature adjustment amount; The time period corresponding to the surface temperature less than the first temperature threshold is regarded as the surface low temperature season; the time period corresponding to the surface temperature greater than the second temperature threshold is regarded as the surface high temperature season.

6. The method for constructing a regional multi-seasonal remote sensing feature set according to claim 1, characterized in that: The target geographical area includes a plurality of original partitions, and the target geographical area is divided into a plurality of sub-areas, including: Obtaining a vegetation index sequence and a surface temperature sequence for each of the original partitions; Clustering the multiple original partitions using the vegetation index sequences and the surface temperature sequences of the multiple original partitions to obtain multiple similar partition sets; The original partitions in each of the similar partition sets are merged to obtain the multiple sub-regions.

7. The method for constructing a regional multi-seasonal remote sensing feature set according to claim 6, characterized in that: The plurality of original partitions are clustered using the vegetation index sequences and the surface temperature sequences of the plurality of original partitions to obtain a plurality of similar partition sets, including: The multiple original partitions are hierarchically clustered using the vegetation index sequences and the surface temperature sequences of the multiple original partitions to obtain multiple similar partition sets.

8. A device for constructing a regional multi-season remote sensing feature set, characterized in that: The device comprises: A region division module, used for dividing a target geographical area into a plurality of sub-areas; A season division module, used to respectively determine the vegetation growth season, vegetation degradation season, surface low temperature season and surface high temperature season of each sub-region; The optimal combination module is used to determine the optimal combination between the vegetation growth season, the vegetation degradation season, the low surface temperature season and the high surface temperature season according to the spectral characteristics of each sub-area in the vegetation growth season, the vegetation degradation season, the low surface temperature season and the high surface temperature season.

9. A storage medium, characterized in that: The storage medium stores a computer program, and when the computer program is executed by a processor, the method for constructing a regional multi-seasonal remote sensing feature set according to any one of claims 1 to 7 is implemented.

10. An electronic device, characterized in that: The electronic device includes a processor and a memory, wherein the memory stores a computer program, and when the computer program is executed by the processor, the method for constructing a regional multi-seasonal remote sensing feature set as described in any one of claims 1 to 7 is implemented.

Images