Mosaic method and device for spaceborne hyperspectral remote sensing quantitative inversion products

Abstract

The embodiments of the present disclosure provide a mosaic method and device for spaceborne hyperspectral remote sensing quantitative inversion products, which are applied to the mosaic field. The method includes taking the spaceborne hyperspectral remote sensing quantitative inversion products to be processed that have spatial overlap with the reference benchmark products as the first set of products to be processed; inputting each product in the first set of products to be processed into a preset evaluation index model to output the total score corresponding to each product; sorting the products in the first set of products to be processed from high to low according to the total score of each product and mosaicking each product in turn until the mosaicking of each product in the first set of products to be processed is completed to obtain a new reference benchmark product, and taking the spaceborne hyperspectral remote sensing quantitative inversion products to be processed that have spatial overlap with the new reference benchmark product as the second set of products to be processed; repeating the above steps until all the spaceborne hyperspectral remote sensing quantitative inversion products to be processed are mosaicked. In this way, batch automated selection and mosaicking of inversion products are achieved.

Description

Technical Field

[0001] The present invention relates to the field of image processing technology, in particular to the field of automatic selection and mosaicking technology, and specifically to a mosaicking method and device for spaceborne hyperspectral remote sensing quantitative inversion products. Background Art

[0002] Satellite-borne hyperspectral data contains hundreds of spectral channels that are closely arranged, which can obtain more detailed spectral features and achieve a qualitative leap in the ability to identify and analyze land objects. Hyperspectral remote sensing technology can not only be used for land object classification and change detection, but also plays an important role in quantitative resource and environmental surveys.

[0003] In recent years, with the successive launch of hyperspectral satellites at home and abroad, the quantitative inversion technology of hyperspectral remote sensing has also developed rapidly, and the types and quantity of quantitative inversion products have become increasingly rich. However, most of the existing research on quantitative inversion products of spaceborne hyperspectral remote sensing focuses on the automated batch production of single-view products, and there is relatively little research on mosaic mapping of large-scale quantitative inversion products at regional scale.

[0004] At present, when producing regional-scale spaceborne hyperspectral remote sensing quantitative inversion products, most of them directly mosaic the hyperspectral images at the original product level, and then invert the mosaicked images to produce products. When mosaicking directly, a large amount of original information is often lost in order to achieve visual effects such as uniform color, resulting in reflectivity distortion and reducing the accuracy of the final product. This distortion not only affects the scientific value of the data, but also limits the application potential of hyperspectral data in precision agriculture, environmental monitoring and other fields. In addition, as the source of spaceborne hyperspectral data becomes increasingly abundant, the amount of data increases dramatically. Hyperspectral data usually contains hundreds of channels. Mosaicing these massive raw data not only results in a huge amount of data, but also has a lot of data redundancy, which brings challenges in terms of computational efficiency and storage space. When faced with such a large amount of data, current data processing methods are often inefficient and difficult to meet the needs of real-time or near-real-time processing. At the same time, they also put forward higher requirements for storage facilities.

[0005] Direct mosaicking of quantitative products can theoretically avoid information loss during the mosaicking process of raw data and maintain high precision and fidelity of data. However, due to large differences in band settings, spatial resolution, observation time, etc. of the original images, the quality of different products varies. In the mosaicking process, if affected by low-quality products, obvious stitching marks or data deviations may appear, reducing the accuracy of quantitative inversion products. Therefore, it is necessary to develop new mosaicking methods and devices to achieve seamless docking and high-precision fusion between different quantitative products, improve the mosaicking efficiency and accuracy of space-borne hyperspectral remote sensing quantitative inversion products, and give full play to the application potential of hyperspectral satellite data in various fields. Summary of the invention

[0006] The present invention provides a mosaic method and device for spaceborne hyperspectral remote sensing quantitative inversion products.

[0007] According to a first aspect of the present disclosure, a mosaic method for spaceborne hyperspectral remote sensing quantitative inversion products is provided. The method comprises:

[0008] Acquire each spaceborne hyperspectral remote sensing quantitative inversion product to be processed and a reference benchmark product, and use the spaceborne hyperspectral remote sensing quantitative inversion product to be processed that has a spatial overlap with the reference benchmark product as a first set of products to be processed;

[0009] Input each product in the first set of products to be processed into a preset evaluation index model, and output a total score corresponding to each product;

[0010] Sort the products in the first set of products to be processed from high to low according to their total scores, and perform mosaic processing on the products in turn until the mosaic of the products in the first set of products to be processed is completed to obtain a new reference benchmark product, and use the space-borne hyperspectral remote sensing quantitative inversion products to be processed that have a spatial overlap with the new reference benchmark product as the second set of products to be processed;

[0011] Input each product in the second set of products to be processed into the preset evaluation index model, and output the total score corresponding to each product;

[0012] The products in the second set of products to be processed are sorted from high to low according to their total scores, and each product is mosaicked in turn until all products in the second set of products to be processed are mosaicked to obtain new reference benchmark products until all the space-borne hyperspectral remote sensing quantitative inversion products to be processed are mosaicked.

[0013] According to the above aspects and any possible implementation, an implementation is further provided, wherein the sequentially inlaying of each product includes:

[0014] Based on the preset mosaic parameters, the product with the highest total score is mosaicked with the reference benchmark product. The reference benchmark product is always placed on the top layer during mosaicking. The mosaicked new product is used as the new reference benchmark product to mosaic the next product until the mosaicking of all products in the product set to be processed is completed.

[0015] According to the above aspects and any possible implementation, an implementation is further provided, wherein the construction of the preset evaluation index model includes:

[0016] Obtain the index information of each spaceborne hyperspectral remote sensing quantitative inversion product to be processed and establish an index set;

[0017] Calculating the relative root mean square error between each product in the first set of products to be processed and the reference product in the spatial overlapping area;

[0018] Based on a preset impact estimation model, determine the impact of each indicator in the indicator set on the relative root mean square error, and score the impact;

[0019] The score of each indicator in the indicator set is divided by its maximum value, normalized, and converted into weights, sorted from high to low, and the indicators whose cumulative weights exceed the preset threshold are retained, and the ratio of the weight of each retained indicator to the total weight of all indicators is calculated as the final weight of each retained indicator;

[0020] According to the preset indicator scoring rules, score each retained indicator, multiply the score of each indicator by its corresponding weight, sum the weighted scores of all indicators, and obtain the total score of each product in the first set of products to be processed;

[0021] The effectiveness of the scoring rule is evaluated using the correlation index, and the preset threshold and the preset index scoring rule are adjusted until the preset evaluation index model is effective.

[0022] According to the above aspects and any possible implementation manner, an implementation manner is further provided, wherein the preset indicator scoring rule includes:

[0023] For numerical indicators, linear scoring functions are used for scoring; for text indicators, scoring is performed based on the scoring criteria determined according to the distribution of indicator values.

[0024] According to the aspects and any possible implementations described above, an implementation is further provided, wherein the reference benchmark product is a main phase and main payload spaceborne hyperspectral remote sensing quantitative inversion product.

[0025] According to the above aspects and any possible implementation manner, an implementation manner is further provided, wherein after obtaining the index information of each spaceborne hyperspectral remote sensing quantitative inversion product to be processed, the method further includes:

[0026] Generate a summary table of product information to be processed according to each of the spaceborne hyperspectral remote sensing quantitative inversion products to be processed, and pre-process the summary table of product information to be processed;

[0027] Based on the preset monitoring requirements and according to the summary table of preprocessed product information to be processed, the main phase and main payload spaceborne hyperspectral remote sensing quantitative inversion products are determined.

[0028] As described above and any possible implementation method, an implementation method is further provided, wherein the indicator information includes: satellite name, payload type, imaging time, scene number, product number, spatial resolution, satellite zenith angle, satellite azimuth angle, solar zenith angle, solar azimuth angle, cloud cover, radiation correction parameters, and longitude and latitude coordinates of the center point.

[0029] According to a second aspect of the present disclosure, a mosaic device for spaceborne hyperspectral remote sensing quantitative inversion products is provided. The device comprises:

[0030] An acquisition module is used to acquire each spaceborne hyperspectral remote sensing quantitative inversion product to be processed and a reference benchmark product, and to use the spaceborne hyperspectral remote sensing quantitative inversion product to be processed that has a spatial overlap with the reference benchmark product as a first set of products to be processed;

[0031] A processing module, used for inputting each product in the first set of products to be processed into a preset evaluation index model, and outputting a total score corresponding to each product;

[0032] The processing module is further used to sort the products in the first set of products to be processed from high to low according to the total score of each product, and perform mosaic processing on each product in turn until the mosaic of each product in the first set of products to be processed is completed to obtain a new reference benchmark product, and the space-borne hyperspectral remote sensing quantitative inversion products to be processed that have a spatial overlap with the new reference benchmark product are used as the second set of products to be processed;

[0033] The processing module is further used to input each product in the second set of products to be processed into the preset evaluation index model and output the total score corresponding to each product;

[0034] The processing module is also used to sort the total scores of the products in the second set of products to be processed from high to low, and to mosaic each product in turn until the mosaic of each product in the second set of products to be processed is completed to obtain a new reference benchmark product until all the space-borne hyperspectral remote sensing quantitative inversion products to be processed are mosaicked.

[0035] According to a third aspect of the present disclosure, an electronic device is provided, which includes a memory and a processor, wherein a computer program is stored in the memory, and when the processor executes the program, the method described above is implemented.

[0036] According to a fourth aspect of the present disclosure, a computer-readable storage medium is provided, on which a computer program is stored, and when the program is executed by a processor, the method described above is implemented.

[0037] The embodiment of the present application provides a mosaic method and device for spaceborne hyperspectral remote sensing quantitative inversion products. The method can obtain each spaceborne hyperspectral remote sensing quantitative inversion product to be processed and a reference benchmark product, and use the spaceborne hyperspectral remote sensing quantitative inversion product to be processed that has a spatial overlap with the reference benchmark product as a first set of products to be processed; then input each product in the first set of products to be processed into a preset evaluation index model, and output the total score corresponding to each product; then sort the products in the first set of products to be processed from high to low according to the total score of each product, and mosaic the products in turn until the mosaic of each product in the first set of products to be processed is completed to obtain a new reference benchmark product, and use the spaceborne hyperspectral remote sensing quantitative inversion product to be processed that has a spatial overlap with the new reference benchmark product as a second set of products to be processed; then input each product in the second set of products to be processed into a preset evaluation index model, and output the total score corresponding to each product; then sort the products in the second set of products to be processed from high to low according to the total score of each product, and mosaic the products in turn until the mosaic of each product in the second set of products to be processed is completed. The mosaicking of products is completed and a new reference benchmark product is obtained until all the spaceborne hyperspectral remote sensing quantitative inversion products to be processed are mosaicked. Based on this, a method for automatic selection and mosaicking of spaceborne hyperspectral remote sensing quantitative inversion products can be proposed according to the characteristics of spaceborne hyperspectral remote sensing quantitative inversion products. After the data scoring index is determined in advance, the construction of the evaluation index model is completed, and the mosaicking order is automatically selected according to the comprehensive evaluation score. The specific implementation process and method of scoring rule formulation, automatic selection of mosaic product rules, and overall mosaicking order are given, which fully considers the data volume and data quality issues. It has the characteristics of simple operation, high degree of automation, and is suitable for large-scale and batch processing of spaceborne hyperspectral data. It solves the problem of consuming a lot of time in manual mosaicking, can perform automatic mosaicking, and select products with better quality for mosaicking according to automatic scoring to achieve the best mosaicking effect, thereby realizing the batch, large-scale, and automatic selection and mosaicking of spaceborne hyperspectral remote sensing quantitative inversion products, and providing important technical support for large-scale natural resource surveys and environmental monitoring of spaceborne hyperspectral data.

[0038] It should be understood that the contents described in the summary of the invention are not intended to limit the key or important features of the embodiments of the present disclosure, nor are they intended to limit the scope of the present disclosure. Other features of the present disclosure will become easily understood through the following description. BRIEF DESCRIPTION OF THE DRAWINGS

[0039] The above and other features, advantages and aspects of the embodiments of the present disclosure will become more apparent with reference to the following detailed description in conjunction with the accompanying drawings. The accompanying drawings are used to better understand the present solution and do not constitute a limitation of the present disclosure. In the accompanying drawings, the same or similar reference numerals represent the same or similar elements, among which:

[0040] Figure 1A flow chart of a mosaic method for spaceborne hyperspectral remote sensing quantitative inversion products according to an embodiment of the present disclosure is shown;

[0041] Figure 2 A schematic diagram of a spaceborne hyperspectral remote sensing quantitative inversion product before mosaicking according to an embodiment of the present disclosure is shown;

[0042] Figure 3 A schematic diagram of a spaceborne hyperspectral remote sensing quantitative inversion product after mosaicking according to an embodiment of the present disclosure is shown;

[0043] Figure 4 A block diagram of a mosaicking device for spaceborne hyperspectral remote sensing quantitative inversion products according to an embodiment of the present disclosure is shown;

[0044] Figure 5 A block diagram of an exemplary electronic device capable of implementing embodiments of the present disclosure is shown. DETAILED DESCRIPTION

[0045] In order to make the purpose, technical solution and advantages of the embodiments of the present disclosure clearer, the technical solution in the embodiments of the present disclosure will be clearly and completely described below in conjunction with the drawings in the embodiments of the present disclosure. Obviously, the described embodiments are part of the embodiments of the present disclosure, not all of the embodiments. Based on the embodiments in the present disclosure, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present disclosure.

[0046] In addition, the term "and / or" in this article is only a description of the association relationship between the associated objects, indicating that there can be three relationships. For example, A and / or B can represent: A exists alone, A and B exist at the same time, and B exists alone. In addition, the character " / " in this article generally indicates that the associated objects before and after are in an "or" relationship.

[0047] In the present disclosure, a method for automatically selecting and mosaicking spaceborne hyperspectral remote sensing quantitative inversion products is proposed according to the characteristics of spaceborne hyperspectral remote sensing quantitative inversion products. After the data scoring index is determined in advance, the construction of the evaluation index model is completed, and the mosaicking order is automatically selected according to the comprehensive evaluation score. The specific implementation process and method of scoring rule formulation, automatic selection of mosaicking product rules, and overall mosaicking order are given, and the data volume and data quality issues are fully considered. The method has the characteristics of simple operation, high degree of automation, and suitability for large-scale and batch processing of spaceborne hyperspectral data. The problem of consuming a lot of time in manual mosaicking is solved, and automatic mosaicking can be performed. Products with better quality are selected for mosaicking according to the automatic scoring to achieve the best mosaicking effect, thereby realizing batch, large-scale, and automatic selection and mosaicking of spaceborne hyperspectral remote sensing quantitative inversion products, and providing important technical support for large-scale natural resource surveys and environmental monitoring of spaceborne hyperspectral data.

[0048] Figure 1 A flowchart of a mosaicking method 100 for spaceborne hyperspectral remote sensing quantitative inversion products according to an embodiment of the present disclosure is shown.

[0049] In block 110 , each spaceborne hyperspectral remote sensing quantitative inversion product to be processed and a reference benchmark product are obtained, and the spaceborne hyperspectral remote sensing quantitative inversion product to be processed that has a spatial overlap with the reference benchmark product is taken as a first set of products to be processed.

[0050] In box 120, each product in the first set of products to be processed is input into a preset evaluation index model, and the total score corresponding to each product is output.

[0051] In box 130, the products in the first set of products to be processed are sorted from high to low according to their total scores, and the products are mosaicked in turn until the mosaicking of the products in the first set of products to be processed is completed to obtain a new reference benchmark product, and the processed satellite-borne hyperspectral remote sensing quantitative inversion products that have spatially overlapping areas with the new reference benchmark products are used as the second set of products to be processed.

[0052] In block 140 , each product in the second set of products to be processed is input into a preset evaluation index model, and a total score corresponding to each product is output.

[0053] In box 150, the total scores of the products in the second set of products to be processed are sorted from high to low, and the products are mosaicked in turn until the mosaicking of the products in the second set of products to be processed is completed to obtain a new reference benchmark product, until all the space-borne hyperspectral remote sensing quantitative inversion products to be processed are mosaicked.

[0054] In block 110 , each spaceborne hyperspectral remote sensing quantitative inversion product to be processed may be a spaceborne hyperspectral remote sensing quantitative inversion product in a study area searched and acquired according to user requirements and its corresponding metadata file and coverage vector file.

[0055] In some embodiments, the spaceborne hyperspectral remote sensing quantitative inversion product may include the following: surface reflectance product or surface parameter inversion product, product vector range file, metadata xml file, etc. Among them, the xml file can display the satellite name (SatelliteID), payload type (SensorID), imaging time (StartTime, EndTime), orbit number (OrbitID), scene number (SceneID), product number (ProductID), spatial resolution (ImageGSD), satellite zenith angle (SatelliteZenith), satellite azimuth angle (SatelliteAzimuth), solar zenith angle (SolarZenith), solar azimuth angle (SolarAzimuth), cloud cover (CloudCover), radiation correction parameters (GainMode, RadiometricParam) and center point longitude and latitude (CenterLatitude, CenterLongitude) and other information corresponding to each product, as shown in Table 1.

[0056] Table 1: Metadata XML file example

[0057]

[0058] In some embodiments, the first set of products to be processed may be determined by determining the degree of spatial overlap between the space-borne hyperspectral remote sensing quantitative inversion products to be processed and the reference benchmark products.

[0059] Specifically, if the spatial overlap between the to-be-processed spaceborne hyperspectral remote sensing quantitative inversion product and the reference benchmark product is greater than or equal to the preset spatial overlap threshold, it is determined that the to-be-processed spaceborne hyperspectral remote sensing quantitative inversion product and the reference benchmark product have an overlapping area, and the to-be-processed spaceborne hyperspectral remote sensing quantitative inversion product that has a spatial overlap area with the reference benchmark product is taken as the first to-be-processed product set A. The preset spatial overlap threshold can be set according to the actual needs of the user.

[0060] In some embodiments, before determining whether the degree of spatial overlap between the spaceborne hyperspectral remote sensing quantitative inversion product to be processed and the reference benchmark product is greater than or equal to a preset spatial overlap threshold, Python code can be used to complete the hard mosaic of the reference benchmark product to obtain the vector coverage of the reference benchmark product, and Python code can be used to read the vector coverage file of the product to be processed, and the Geopandas or Shapely library is used to calculate the spatial relationship between the vector coverage of the product to be processed and the vector coverage of the reference benchmark product, and the spaceborne hyperspectral remote sensing quantitative inversion product to be processed that has a spatial overlap area with the reference benchmark product is used as the first set of products A to be processed.

[0061] Specifically, based on the phase and load of the reference benchmark product, Python code is used to complete hard mosaicking, and the reference benchmark product hard mosaicking results and the corresponding vector coverage range files are obtained. For each spaceborne hyperspectral remote sensing quantitative inversion product to be processed, the vector coverage range of the product to be processed is judged with the vector coverage range of the reference benchmark product to obtain the spaceborne hyperspectral remote sensing quantitative inversion product to be processed that has an overlapping area with the reference benchmark product, and the reference benchmark product and these spaceborne hyperspectral remote sensing quantitative inversion products are cropped to obtain the reference benchmark product and spaceborne hyperspectral remote sensing quantitative inversion product in the overlapping area.

[0062] In some embodiments, the phase and payload of the reference benchmark product can be used as evaluation criteria for each satellite-borne hyperspectral remote sensing quantitative inversion product to be processed, that is, each satellite-borne hyperspectral remote sensing quantitative inversion product to be processed can also be input into a preset evaluation index model, and the total score corresponding to each product can be output, and all main phase and main payload data can be automatically screened out from the scored satellite-borne hyperspectral remote sensing quantitative inversion products to be processed in descending order according to the score. The main phase can be divided into main time point data and main time period data. First, Python is used to automatically hard-mosaic the main time point data, and then Python is used to automatically hard-mosaic the main time period that does not overlap with the main time point data, and Python code is used to merge the vector data that has been hard-mosaiced simultaneously. This process completely realizes Python code automation.

[0063] In block 120 , the preset evaluation index model may be determined according to the actual needs of the user, based on the quantitative inversion product of spaceborne hyperspectral remote sensing and in combination with a human-computer interaction method.

[0064] In some embodiments, the construction of the above-mentioned preset evaluation index model includes:

[0065] Obtain the index information of each spaceborne hyperspectral remote sensing quantitative inversion product to be processed and establish an index set;

[0066] Calculate the relative root mean square error between each product in the first set of products to be processed and the reference benchmark product in the spatial overlapping area;

[0067] Based on the preset impact estimation model, determine the impact of each indicator in the indicator set on the relative root mean square error and score the impact;

[0068] Divide the score of each indicator in the indicator set by its maximum value, perform normalization, convert the score into weight, sort it from high to low, and retain the indicators whose cumulative weight exceeds the preset threshold. Calculate the ratio of the weight of each retained indicator to the total weight of all indicators as the final weight of each retained indicator;

[0069] According to the preset indicator scoring rules, score each retained indicator, multiply the score of each indicator by its corresponding weight, sum the weighted scores of all indicators, and obtain the total score of each product in the first set of products to be processed;

[0070] Use correlation indicators to evaluate the effectiveness of the scoring rules, and adjust the preset thresholds and preset indicator scoring rules until the preset evaluation indicator model is effective.

[0071] In some embodiments, the metadata file (xml format) of each spaceborne hyperspectral remote sensing quantitative inversion product to be processed can be read by using Python code to comprehensively obtain the index information of the spaceborne hyperspectral remote sensing quantitative inversion product to be processed, and establish the index set Z.

[0072] In some embodiments, the above index information includes: satellite name, payload type, imaging time, scene number, product number, spatial resolution, satellite zenith angle, satellite azimuth angle, solar zenith angle, solar azimuth angle, cloud cover, radiation correction parameters and longitude and latitude coordinates of the center point.

[0073] In some embodiments, after obtaining the index information of each spaceborne hyperspectral remote sensing quantitative inversion product to be processed, the method further includes:

[0074] Generate a summary table of product information to be processed based on each spaceborne hyperspectral remote sensing quantitative inversion product to be processed, and pre-process the summary table of product information to be processed;

[0075] Based on the preset monitoring requirements and the summary table of pre-processed product information to be processed, the main phase and main payload spaceborne hyperspectral remote sensing quantitative inversion products are determined.

[0076] In some embodiments, preprocessing the summary table of product information to be processed may include formatting dates, encoding text information, etc., so as to facilitate subsequent mosaic processing and thereby improve mosaic efficiency.

[0077] In some embodiments, based on the summary table of product information to be processed, the user determines the phase and load requirements of the final quantitative inversion product according to monitoring requirements, and sets them as the main phase and main load respectively, and uses the main phase and main load satellite-borne hyperspectral remote sensing quantitative inversion products as reference benchmark products S.

[0078] In some embodiments, the similarity between each product in set A and the reference benchmark product S in the overlapping area is calculated respectively. Specifically, the relative root mean square error RRMSE can be used as the similarity evaluation index. For each product in set A, the RRMSE calculation formula is:

[0079] ;

[0080] in, and Respectively represent the i-th pixel value of the product and S in the overlapping area, represents the mean value of all pixels of the reference benchmark product S in the overlapping area, n Indicates the total number of pixels in the overlapping area. Calculate the RRMSE of all products in set A and the reference benchmark product S in the overlapping area. The smaller the RRMSE value, the higher the similarity between the product and the reference benchmark product S in the overlapping area.

[0081] In some embodiments, the preset impact estimation model can be set according to the actual needs of the user, such as a machine learning model (random forest, gradient boosting machine, etc.). The machine learning model algorithm is used to evaluate the impact of each indicator in the indicator set Z on the RRMSE, and the impact is scored. The score of each indicator is divided by its maximum value, normalized, and converted into a weight, and the indicators whose cumulative weight exceeds a specific threshold, that is, the preset threshold, are retained. The preset threshold can be set as needed, such as 90%, 95%, 98%, 99%, etc. For the retained indicators, the ratio of the weight of each indicator to the total weight of all indicators is calculated as the final weight of each indicator.

[0082] In some embodiments, the preset indicator scoring rules can be set according to the actual needs of the user.

[0083] In some embodiments, the above preset indicator scoring rules include:

[0084] For numerical indicators, linear scoring functions are used for scoring; for text indicators, scoring is performed based on the scoring criteria determined according to the distribution of indicator values.

[0085] In some embodiments, when designing the scoring rules for each indicator, for numerical indicators, a linear scoring function is used for scoring, and for text indicators, a scoring standard is determined according to the distribution of the indicator value, such as threshold, classification, level, etc., and the highest score for each indicator is 100 points, and the lowest score is 0. The score of each indicator is multiplied by its corresponding weight, and then the weighted scores of all indicators are summed to obtain the total score of the product to be processed.

[0086] In some embodiments, the total score of each product in set A can be calculated, and the relationship between the total score of each product and the RRMSE of the product can be analyzed, such as using correlation indicators such as the Pearson correlation coefficient to evaluate the effectiveness of the scoring rules. When the correlation coefficient reaches 0.5, the scoring model can be considered valid, that is, the preset evaluation index model is valid. The higher the correlation coefficient, the better the effectiveness of the scoring model; when the correlation coefficient is lower than 0.5, it is necessary to return to the previous steps to modify the threshold or scoring rules until the scoring model is valid, so that after verifying that the scoring model is valid, the Python code can be used to realize the automatic reading and scoring of the summary table of product information to be processed.

[0087] In some embodiments, the scoring index P determined in the preset evaluation index model i (i=1,2,…,5), P1 to P5 correspond to the imaging year, imaging date, satellite name, orbit number and cloud cover information respectively. According to the metadata xml file of the spaceborne hyperspectral remote sensing quantitative inversion product acquired in advance, the name of each product xml file and the internal scoring index information are read through Python code, and the information of each product read is summarized into a data summary table, showing the data distribution of different loads and phases, as shown in Table 2.

[0088] Table 2: Examples of data distribution for different loads and time phases

[0089]

[0090] It should be noted that in the process of constructing the preset evaluation index model, the retained scoring indicators can also be determined based on the space-borne hyperspectral remote sensing quantitative inversion products to be processed and the reference benchmark products, so as to perform scoring in the hard mosaic stage.

[0091] In some embodiments, during the hard mosaic stage, the main time point is accurate year, month, and day information, such as April 15, 2021. The main time period is a buffer zone set for the main time point. If the time range is set to one month, the main time period of the main time point is April 1, 2021 to April 30, 2021. Multiple main orbits can be determined based on the orbital distribution of the main time phase data. The total score of each product is obtained by comprehensive scoring of multiple indicators, which is used to automatically select the optimal product mosaic order. Combined with the information of each indicator in the summary table of the product information to be processed, the degree of influence of each indicator on the RRMSE is evaluated, and the degree of influence is scored. The score of each indicator is normalized and converted into a weight. The indicators whose cumulative weight exceeds a specific threshold are retained. For the retained indicators, the ratio of the weight of each indicator to the total weight of all indicators is calculated to determine the scoring indicator P. i The weight w i . Each scoring index P i The weight w i The diagram is shown in Table 3. Among them, the main season is April, which belongs to spring. The corresponding track is regarded as the main track, and the cloud cover less than 30% is regarded as meaningful.

[0092] Table 3: Scoring indicators P i The weight w i Hint

[0093]

[0094] .

[0095] In some embodiments, the main phase can be mosaicked according to the main time point and the main time period respectively, and Python code is used to automatically screen out all the main phase and main load products, and sort them from large to small according to the scores. First, the main time point product is read from the screened main phase and main load product list, and the product with cloud cover less than 10% is hard mosaicked using Python code to obtain the main time point hard mosaic product and its corresponding vector range; then Python code is used to determine whether there is an overlapping area between each scene product of the main time period and the hard mosaicked product. If there is no overlapping area and the product cloud cover is less than 10%, the product of the main time period is The product is hard-mosaiced with the hard-mosaiced product of the main time point. If it exists, it is skipped to obtain the hard-mosaiced product of the main time phase. The vector range of the hard-mosaiced product in the main time period is synchronously merged with the hard-mosaiced vector range of the main time point to obtain the vector coverage range corresponding to the hard-mosaic product of the main time phase; the orbit number of the hard-mosaic product of the main time phase is automatically read through the Python code, and the data information on the same track in the scored total table is searched. The products that meet the cloud amount and quality are selected to be mosaicked with the hard-mosaic product of the main time phase to obtain the hard-mosaic product of the reference benchmark product, and the vector range is synchronously merged to obtain the vector coverage range corresponding to the hard-mosaic product of the reference benchmark product.

[0096] In blocks 130 to 150, the above-mentioned sequentially inlaying of each product includes:

[0097] Based on the preset mosaic parameters, the product with the highest total score is mosaicked with the reference benchmark product. The reference benchmark product is always placed on the top layer during mosaicking. The mosaicked new product is used as the new reference benchmark product to mosaic the next product until the mosaicking of all products in the product set to be processed is completed.

[0098] In some embodiments, the preset mosaic parameters may include setting mosaic parameters such as histogram matching, feathering distance, and edge line.

[0099] In some embodiments, the products in set A are sorted from high to low according to their total scores, and the products with the highest scores are inlaid with the reference benchmark product S through inlay parameters. The reference benchmark product S is always placed on the top layer during inlaying, and the new product after inlaying is used as the reference benchmark product S. Then, the inlaying of the next product is started in the same way, and so on, until the inlaying of all products in set A is completed to obtain a new reference benchmark product S1. Based on the new reference benchmark product S1, all the products to be processed that overlap with its space are searched and included in the second set of products to be processed A1. Automatic scoring is performed using a preset evaluation index model, and the inlaying is completed by repeating the above steps. And so on, until the inlaying of all products to be processed is completed.

[0100] It should be noted that, in combination with the above-mentioned hard mosaicking, a vector coverage range buffer corresponding to the hard mosaicking result of the reference benchmark product can also be created. For example, if the distance is 20km, the xml file within the range is automatically searched, and the corresponding product information within the range is read, that is, the first set of products to be processed A, the scoring index information of the products within the range is extracted from the total scored table, the products within the range are sorted from large to small according to the scores, and the above-mentioned method is used to mosaic with the hard mosaicking result of the reference benchmark product to obtain the mosaicking result, and the vector range corresponding to the mosaicking result is obtained synchronously; the above process is repeated until no products to be processed are found in the buffer using the Python code, and the mosaicking is completed to obtain the mosaicking result of the final quantitative inversion product.

[0101] like Figure 2 and Figure 3 As shown in the figure, by comparing the spaceborne hyperspectral remote sensing quantitative inversion products before and after mosaicking, the spaceborne hyperspectral remote sensing quantitative inversion products obtained without the above mosaicking method have obvious color difference and obvious differentiation in the overlapping area, while the spaceborne hyperspectral remote sensing quantitative inversion products obtained by the above mosaicking method have uniform coloring and a more natural transition in the overlapping area.

[0102] It should be noted that in the above mosaicking method, the first set of products to be processed can be automatically mosaicked based on the main phase, main load and main track data based on the preset evaluation index model. It is also possible to automatically mosaic the main phase, main load and main track data based on the preset evaluation index model.

[0103] According to the embodiments of the present disclosure, the following technical effects are achieved:

[0104] The invention can obtain each to-be-processed satellite-borne hyperspectral remote sensing quantitative inversion product and a reference benchmark product, and use the to-be-processed satellite-borne hyperspectral remote sensing quantitative inversion product that has a spatial overlap with the reference benchmark product as a first to-be-processed product set; then input each product in the first to-be-processed product set into a preset evaluation index model, and output the total score corresponding to each product; then sort the products in the first to-be-processed product set from high to low according to the total score of each product, and perform mosaic processing on each product in turn until the mosaic of each product in the first to-be-processed product set is completed to obtain a new reference benchmark product, and use the to-be-processed satellite-borne hyperspectral remote sensing quantitative inversion product that has a spatial overlap with the new reference benchmark product as a second to-be-processed product set; then input each product in the second to-be-processed product set into a preset evaluation index model, and output the total score corresponding to each product; then sort the products in the second to-be-processed product set from high to low according to the total score of each product, and perform mosaic processing on each product in turn until the mosaic of each product in the second to-be-processed product set is completed to obtain a new reference benchmark product. Until all the spaceborne hyperspectral remote sensing quantitative inversion products to be processed are mosaicked; based on this, a method for automatic selection and mosaicking of spaceborne hyperspectral remote sensing quantitative inversion products can be proposed according to the characteristics of spaceborne hyperspectral remote sensing quantitative inversion products. After the data scoring indicators are determined in advance, the construction of the evaluation indicator model is completed, and the mosaicking order is automatically selected through the comprehensive evaluation scores. The specific implementation process and method of scoring rule formulation, automatic selection of mosaicking product rules, and overall mosaicking order are given, which fully considers the data volume and data quality issues. It has the characteristics of simple operation, high degree of automation, and is suitable for large-scale and batch processing of spaceborne hyperspectral data. It solves the problem of consuming a lot of time in manual mosaicking, can perform automatic mosaicking, and select inversion products with better quality according to automatic scoring for mosaicking to achieve the best mosaicking effect, thereby realizing the batch, large-scale, and automatic selection and mosaicking of spaceborne hyperspectral remote sensing quantitative inversion products, and providing important technical support for large-scale natural resource surveys and environmental monitoring of spaceborne hyperspectral data.

[0105] It should be noted that, for the aforementioned method embodiments, for the sake of simplicity, they are all described as a series of action combinations, but those skilled in the art should be aware that the present disclosure is not limited by the order of the actions described, because according to the present disclosure, certain steps can be performed in other orders or simultaneously. Secondly, those skilled in the art should also be aware that the embodiments described in the specification are all optional embodiments, and the actions and modules involved are not necessarily required by the present disclosure.

[0106] The above is an introduction to the method embodiment. The following is a further explanation of the scheme disclosed in the present invention through an apparatus embodiment.

[0107] Figure 4 FIG. 4 is a block diagram of a mosaic device 400 for spaceborne hyperspectral remote sensing quantitative inversion products according to an embodiment of the present disclosure. Figure 4 As shown, the device 400 includes:

[0108] The acquisition module 410 is used to acquire each spaceborne hyperspectral remote sensing quantitative inversion product to be processed and the reference benchmark product, and take the spaceborne hyperspectral remote sensing quantitative inversion product to be processed that has a spatial overlap with the reference benchmark product as the first set of products to be processed;

[0109] Processing module 420, used to input each product in the first set of products to be processed into a preset evaluation index model, and output a total score corresponding to each product;

[0110] The processing module 420 is further used to sort the products in the first set of products to be processed from high to low according to the total score, and perform mosaic processing on each product in turn until the mosaic of each product in the first set of products to be processed is completed to obtain a new reference benchmark product, and the space-borne hyperspectral remote sensing quantitative inversion products to be processed that have a spatial overlap with the new reference benchmark product are used as the second set of products to be processed;

[0111] The processing module 420 is further used to input each product in the second set of products to be processed into a preset evaluation index model and output a total score corresponding to each product;

[0112] The processing module 420 is also used to sort the total scores of the products in the second set of products to be processed from high to low, and to mosaic each product in turn until the mosaic of each product in the second set of products to be processed is completed to obtain a new reference benchmark product, until all the space-borne hyperspectral remote sensing quantitative inversion products to be processed are mosaicked.

[0113] Those skilled in the art can clearly understand that, for the convenience and brevity of description, the specific working process of the described module can refer to the corresponding process in the aforementioned method embodiment, and will not be repeated here.

[0114] In the technical solution disclosed herein, the acquisition, storage and application of user personal information involved are in compliance with the provisions of relevant laws and regulations and do not violate public order and good morals.

[0115] According to an embodiment of the present disclosure, the present disclosure also provides an electronic device, a readable storage medium and a computer program product.

[0116] Figure 5 A block diagram of an exemplary electronic device 500 capable of implementing an embodiment of the present disclosure is shown. The electronic device is intended to represent various forms of digital computers, such as laptop computers, desktop computers, workstations, personal digital assistants, servers, blade servers, mainframe computers, and other suitable computers. The electronic device may also represent various forms of mobile devices, such as personal digital processing, cellular phones, smart phones, wearable devices, and other similar computing devices. The components shown herein, their connections and relationships, and their functions are merely examples and are not intended to limit the implementation of the present disclosure described and / or claimed herein.

[0117] The electronic device 500 includes a computing unit 501, which can perform various appropriate actions and processes according to a computer program stored in a ROM 502 or a computer program loaded from a storage unit 508 into a RAM 503. In the RAM 503, various programs and data required for the operation of the electronic device 500 can also be stored. The computing unit 501, the ROM 502, and the RAM 503 are connected to each other via a bus 504. An I / O interface 505 is also connected to the bus 504.

[0118] Multiple components in the electronic device 500 are connected to the I / O interface 505, including: an input unit 506, such as a keyboard, a mouse, etc.; an output unit 507, such as various types of displays, speakers, etc.; a storage unit 508, such as a disk, an optical disk, etc.; and a communication unit 509, such as a network card, a modem, a wireless communication transceiver, etc. The communication unit 509 allows the electronic device 500 to exchange information / data with other devices through a computer network such as the Internet and / or various telecommunication networks.

[0119] The computing unit 501 may be a variety of general and / or special processing components with processing and computing capabilities. Some examples of the computing unit 501 include, but are not limited to, a central processing unit (CPU), a graphics processing unit (GPU), various dedicated artificial intelligence (AI) computing chips, various computing units running machine learning model algorithms, digital signal processors (DSPs), and any appropriate processors, controllers, microcontrollers, etc. The computing unit 501 performs the various methods and processes described above, such as method 100. For example, in some embodiments, the method 100 may be implemented as a computer software program, which is tangibly contained in a machine-readable medium, such as a storage unit 508.

[0120] In some embodiments, part or all of the computer program may be loaded and / or installed on the electronic device 500 via the ROM 502 and / or the communication unit 509. When the computer program is loaded into the RAM 503 and executed by the computing unit 501, one or more steps of the method 100 described above may be performed. Alternatively, in other embodiments, the computing unit 501 may be configured to execute the method 100 in any other appropriate manner (e.g., by means of firmware).

[0121] Various implementations of the systems and techniques described above herein can be implemented in digital electronic circuit systems, integrated circuit systems, field programmable gate arrays (FPGAs), application specific integrated circuits (ASICs), application specific standard products (ASSPs), systems on chips (SOCs), load programmable logic devices (CPLDs), computer hardware, firmware, software, and / or combinations thereof. These various implementations can include: being implemented in one or more computer programs that can be executed and / or interpreted on a programmable system including at least one programmable processor, which can be a special purpose or general purpose programmable processor that can receive data and instructions from a storage system, at least one input device, and at least one output device, and transmit data and instructions to the storage system, the at least one input device, and the at least one output device.

[0122] The program code for implementing the method of the present disclosure may be written in any combination of one or more programming languages. These program codes may be provided to a processor or controller of a general-purpose computer, a special-purpose computer, or other programmable data processing device, so that the program code, when executed by the processor or controller, enables the functions / operations specified in the flow chart and / or block diagram to be implemented. The program code may be executed entirely on the machine, partially on the machine, partially on the machine and partially on a remote machine as a stand-alone software package, or entirely on a remote machine or server.

[0123] In the context of the present disclosure, a machine-readable medium may be a tangible medium that may contain or store a program for use by or in conjunction with an instruction execution system, device, or equipment. A machine-readable medium may be a machine-readable signal medium or a machine-readable storage medium. A machine-readable medium may include, but is not limited to, an electronic, magnetic, optical, electromagnetic, infrared, or semiconductor system, device, or device, or any suitable combination of the foregoing. A more specific example of a machine-readable storage medium may include an electrical connection based on one or more lines, a portable computer disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the foregoing.

[0124] To provide interaction with a user, the systems and techniques described herein can be implemented on a computer having: a display device for displaying information to the user; and a keyboard and a pointing device (e.g., a mouse or trackball) through which the user can provide input to the computer. Other types of devices can also be used to provide interaction with the user; for example, the feedback provided to the user can be any form of sensory feedback (e.g., visual feedback, auditory feedback, or tactile feedback); and input from the user can be received in any form (including acoustic input, voice input, or tactile input).

[0125] The systems and techniques described herein may be implemented in a computing system that includes back-end components (e.g., as a data server), or a computing system that includes middleware components (e.g., an application server), or a computing system that includes front-end components (e.g., a user computer with a graphical user interface or a web browser through which a user can interact with implementations of the systems and techniques described herein), or a computing system that includes any combination of such back-end components, middleware components, or front-end components. The components of the system may be interconnected by any form or medium of digital data communication (e.g., a communication network). Examples of communication networks include: a local area network (LAN), a wide area network (WAN), and the Internet.

[0126] A computer system may include a client and a server. The client and the server are generally remote from each other and usually interact through a communication network. The relationship of client and server is generated by computer programs running on respective computers and having a client-server relationship with each other. The server may be a cloud server, a server of a distributed system, or a server combined with a blockchain.

[0127] It should be understood that the various forms of processes shown above can be used to reorder, add or delete steps. For example, the steps recorded in this disclosure can be executed in parallel, sequentially or in different orders, as long as the desired results of the technical solutions disclosed in this disclosure can be achieved, and this document does not limit this.

[0128] The above specific implementations do not constitute a limitation on the protection scope of the present disclosure. It should be understood by those skilled in the art that various modifications, combinations, sub-combinations and substitutions can be made according to design requirements and other factors. Any modification, equivalent substitution and improvement made within the spirit and principle of the present disclosure shall be included in the protection scope of the present disclosure.

Claims

1. A mosaic method for spaceborne hyperspectral remote sensing quantitative inversion products, characterized in that: include: Acquire each spaceborne hyperspectral remote sensing quantitative inversion product to be processed and a reference benchmark product, and use the spaceborne hyperspectral remote sensing quantitative inversion product to be processed that has a spatial overlap with the reference benchmark product as a first set of products to be processed; Input each product in the first set of products to be processed into a preset evaluation index model, and output a total score corresponding to each product; Sort the products in the first set of products to be processed from high to low according to their total scores, and perform mosaic processing on the products in turn until the mosaic of the products in the first set of products to be processed is completed to obtain a new reference benchmark product, and use the space-borne hyperspectral remote sensing quantitative inversion products to be processed that have a spatial overlap with the new reference benchmark product as the second set of products to be processed; Input each product in the second set of products to be processed into the preset evaluation index model, and output the total score corresponding to each product; The products in the second set of products to be processed are sorted from high to low according to their total scores, and each product is mosaicked in turn until all products in the second set of products to be processed are mosaicked to obtain new reference benchmark products until all the space-borne hyperspectral remote sensing quantitative inversion products to be processed are mosaicked.

2. The method according to claim 1, characterized in that The sequentially inlaying of the products comprises: Based on the preset mosaic parameters, the product with the highest total score is mosaicked with the reference benchmark product. The reference benchmark product is always placed on the top layer during mosaicking. The mosaicked new product is used as the new reference benchmark product to mosaic the next product until the mosaicking of all products in the product set to be processed is completed.

3. The method according to claim 1, characterized in that The construction of the preset evaluation index model includes: Obtain the index information of each spaceborne hyperspectral remote sensing quantitative inversion product to be processed and establish an index set; Calculating the relative root mean square error between each product in the first set of products to be processed and the reference product in the spatial overlapping area; Based on a preset impact estimation model, determine the impact of each indicator in the indicator set on the relative root mean square error, and score the impact; The score of each indicator in the indicator set is divided by its maximum value, normalized, and converted into weights, sorted from high to low, and the indicators whose cumulative weights exceed the preset threshold are retained, and the ratio of the weight of each retained indicator to the total weight of all indicators is calculated as the final weight of each retained indicator; According to the preset indicator scoring rules, score each retained indicator, multiply the score of each indicator by its corresponding weight, sum the weighted scores of all indicators, and obtain the total score of each product in the first set of products to be processed; The effectiveness of the scoring rule is evaluated using the correlation index, and the preset threshold and the preset index scoring rule are adjusted until the preset evaluation index model is effective.

4. The method according to claim 3, characterized in that The preset indicator scoring rules include: For numerical indicators, linear scoring functions are used for scoring; for text indicators, scoring is performed based on the scoring criteria determined according to the distribution of indicator values.

5. The method according to claim 3, characterized in that: The reference benchmark product is a main phase and main payload satellite-borne hyperspectral remote sensing quantitative inversion product.

6. The method according to claim 5, characterized in that After obtaining the index information of each spaceborne hyperspectral remote sensing quantitative inversion product to be processed, the method further includes: Generate a summary table of product information to be processed based on each of the space-borne hyperspectral remote sensing quantitative inversion products to be processed, and pre-process the summary table of product information to be processed; Based on the preset monitoring requirements and according to the summary table of preprocessed product information to be processed, the main phase and main payload satellite-borne hyperspectral remote sensing quantitative inversion products are determined.

7. The method according to any one of claims 3 to 6, characterized in that: The index information includes: satellite name, payload type, imaging time, scene number, product number, spatial resolution, satellite zenith angle, satellite azimuth angle, solar zenith angle, solar azimuth angle, cloud cover, radiation correction parameters and longitude and latitude coordinates of the center point.

8. A mosaic device for spaceborne hyperspectral remote sensing quantitative inversion products, characterized in that: include: An acquisition module is used to acquire each spaceborne hyperspectral remote sensing quantitative inversion product to be processed and a reference benchmark product, and to use the spaceborne hyperspectral remote sensing quantitative inversion product to be processed that has a spatial overlap with the reference benchmark product as a first set of products to be processed; A processing module, used for inputting each product in the first set of products to be processed into a preset evaluation index model, and outputting a total score corresponding to each product; The processing module is further used to sort the products in the first set of products to be processed from high to low according to the total score of each product, and perform mosaic processing on each product in turn until the mosaic of each product in the first set of products to be processed is completed to obtain a new reference benchmark product, and the space-borne hyperspectral remote sensing quantitative inversion products to be processed that have a spatial overlap with the new reference benchmark product are used as the second set of products to be processed; The processing module is further used to input each product in the second set of products to be processed into the preset evaluation index model and output the total score corresponding to each product; The processing module is also used to sort the total scores of the products in the second set of products to be processed from high to low, and to mosaic each product in turn until the mosaic of each product in the second set of products to be processed is completed to obtain a new reference benchmark product until all the space-borne hyperspectral remote sensing quantitative inversion products to be processed are mosaicked.

9. An electronic device, characterized in that: include: at least one processor; as well as a memory communicatively coupled to the at least one processor; The memory stores instructions that can be executed by the at least one processor, and the instructions are executed by the at least one processor to enable the at least one processor to perform the method of any one of claims 1 to 7.

10. A non-transitory computer-readable storage medium storing computer instructions, characterized in that: The computer instructions are used to cause the computer to execute the method according to any one of claims 1-7.

Images