CN113327259A

CN113327259A - Remote sensing data screening method and system for area coverage

Info

Publication number: CN113327259A
Application number: CN202110888797.3A
Authority: CN
Inventors: 刘巍; 刘士彬; 闫雪静
Original assignee: Aerospace Information Research Institute of CAS
Current assignee: Aerospace Information Research Institute of CAS
Priority date: 2021-08-04
Filing date: 2021-08-04
Publication date: 2021-08-31
Anticipated expiration: 2041-08-04
Also published as: CN113327259B

Abstract

The invention provides a method and a system for screening remote sensing data facing to area coverage, wherein the method comprises the following steps: primarily screening a candidate remote sensing data set from the remote sensing data set to be screened; cutting space geometric objects of each remote sensing image into pieces according to the range of a target area, cutting the pieces mutually and removing the weight to obtain a fragmented geometric element set covering the target area; calculating the fragments of each cut remote sensing image containing the fragmented geometric element set through spatial connection; and preferably selecting the target remote sensing image according to the number of fragments, imaging time and cloud cover contained in each cut remote sensing image. According to the method, the space calculation is directly utilized to obtain the fragmented geometric element set accurately covering the target area, remote sensing data screening is carried out through set relation operation and a multi-condition weight model, data redundancy is effectively reduced, and the comprehensive optimization of the number, time, cloud amount and space coverage of screened remote sensing images is guaranteed; support is provided for producing a long time sequence of large-area image mosaic products.

Description

Remote sensing data screening method and system for area coverage

Technical Field

The invention relates to the technical field of data processing, in particular to a method and a system for screening remote sensing data facing to regional coverage.

Background

The earth is a common home for human survival, and with the continuous progress of human civilization, the continuous discovery and understanding of the unknown world by using technical means becomes a powerful driving force for the progress of the human civilization. Due to the vast and vast surface of the earth, although mankind has evolved over the earth for tens of millions of years, awareness of the habitation of mankind from local to global is very limited. Until the middle of the 20 th century, with the advent of satellite remote sensing technology, the human beings do not really draw the relatively continuous cognition curtain of the whole earth by acquiring image data on the earth surface through the "sky eye". Particularly, in the early twenty-first century, with the rapid progress of the technology and the rapid development of the remote sensing technology, the number of remote sensing satellites is continuously increased, observation data on the earth surface obtained by human beings is continuously and rapidly accumulated, the data scale reaches EB level at present, and then the large data technology generated by due operation provides technical support for processing of mass data and information mining.

At present, although the remote sensing technology is greatly developed, the types and the fineness of remote sensing data are more and more abundant, the originally acquired remote sensing image has the characteristics of flakiness, space-time fragmentation and non-continuity relative to the whole earth due to the immense and wide earth surface and the spherical characteristic. Although the blank of the space-time data is filled to the maximum extent by utilizing the quantity advantage through the satellite constellation technology and the multi-satellite networking technology, and the space-time continuity degree of the whole image after the post-splicing processing is improved, the challenges in the aspects of data screening and data processing caused by the rapid increase of the data scale are brought. Aiming at the rapid generation requirement of large-scale regional images of time series and even global image data, firstly, a group of original data sets with the optimal combination is rapidly screened out from the large-scale image data sets under the space-time constraint condition so as to support the requirement of subsequent rapid splicing processing of the image data sets. This is crucial to the fast execution and result acquisition of the whole task, and if a manual screening mode is adopted, the efficiency is low, the cost consumption is high, and the manual screening mode cannot be effectively integrated into the existing IT (Information Technology) facility system, so as to realize a flexible and self-service on-demand image service scene. Therefore, it is one of the technical problems to be solved urgently how to quickly screen out a group of optimal data subsets from a mass of original remote sensing image data set after collecting the space-time constraint conditions of the user by a set of technical devices so as to match the generation of the subsequent large-scale area images.

Most of traditional remote sensing data screening methods search results meeting conditions from mass data based on a query engine provided by a spatial database. However, the traditional remote sensing data screening method has the following problems that on one hand, the setting of the query condition is relatively fixed, for example, for optical remote sensing satellite data, the remote sensing data can be screened only according to the space range, the time range and the threshold value of the highest cloud cover of a set specific target area, and the query condition setting facing to the specific application requirement is difficult to provide; on the other hand, a large amount of redundancy exists in the screened remote sensing data, the optimal remote sensing data needs to be obtained by means of manual selection, time and labor are wasted, and efficiency is low. Therefore, how to quickly and automatically screen out the optimal remote sensing image data and reduce the workload and time of manual selection is very important.

In order to solve the above problems, in the prior art, after a remote sensing data set is preliminarily screened according to a set specific spatial range, a set specific time range and a set specific threshold value of the highest cloud cover, a target area is subjected to grid division, and then the remote sensing data set is further screened by calculating a coverage relation between a remote sensing image and a grid. The size of the grid is very important for the screening result of the remote sensing data. On one hand, if the grid setting is too large, the area boundary covered by the final grid and the boundary of the target area have a large difference, so that the final optimal result cannot effectively cover the boundary of the target area. Moreover, if the size of the grid is set to exceed the size of the single scene of the remote sensing image, the problem that the grid cannot be covered when whether the grid is covered by the single scene image or not can be caused, so that the remote sensing image is selected in a missing mode, the screening result is inaccurate, and the coverage degree is reduced. On the other hand, if the mesh size is set too small, the amount of calculation increases exponentially, and the screening efficiency is low. .

Disclosure of Invention

The invention provides a method and a system for screening remote sensing data facing to area coverage, which are used for solving the defects of inaccurate screening result and low efficiency caused by secondary screening of the remote sensing data by adopting a grid in the prior art and realizing rapid and accurate screening of the optimal remote sensing data.

The invention provides a remote sensing data screening method facing area coverage, which comprises the following steps:

selecting a remote sensing image which has a spatial intersection relation with a target area and meets a preset non-spatial condition from a remote sensing data set to be screened to generate a candidate remote sensing data set;

performing spatial cutting on the spatial geometric object of each remote sensing image in the candidate remote sensing data set and the target area, and mutually cutting each cut remote sensing image to split each cut remote sensing image into a plurality of fragments;

carrying out duplicate removal processing on repeated fragments in a cutting result, generating a fragmented geometric element set from all the fragments subjected to duplicate removal processing, and carrying out spatial connection on each remote sensing image subjected to cutting and the fragments in the fragmented geometric element set to obtain the fragments contained in each remote sensing image subjected to cutting;

according to the number of fragments, imaging time and cloud cover contained in each cut remote sensing image, screening out a target remote sensing image from the candidate remote sensing data set; and the target area is completely covered by a union result of fragments contained in all target remote sensing images.

According to the method for screening the remote sensing data facing the area coverage, provided by the invention, the mutual cutting of the cut remote sensing images is carried out, and the method comprises the following steps:

combining the cut remote sensing images pairwise to form a combined remote sensing data set, and mutually cutting the two cut remote sensing images in each combined remote sensing data set;

combining the cutting results of the combined remote sensing data sets in pairs to form new combined remote sensing data sets, continuously cutting the two cutting results in each new combined remote sensing data set mutually again, and iteratively executing the processes of combining in pairs and cutting mutually until each remote sensing image after being cut is mutually cut with other remote sensing images after being cut; and the other cut remote sensing images are remote sensing images except the cut remote sensing images.

According to the method for screening the remote sensing data facing the area coverage, provided by the invention, the target remote sensing image is screened out from the candidate remote sensing data set according to the number of fragments, imaging time and cloud cover contained in each cut remote sensing image, and the method comprises the following steps:

if any fragment in the fragmented geometric element set belongs to the unique remote sensing image, taking the remote sensing image to which the fragment belongs as a first target remote sensing image of the target remote sensing image;

and screening out a second target remote sensing image of the target remote sensing image from all other images except the first target remote sensing image in the candidate remote sensing data set according to the number of fragments, imaging time and cloud cover contained in each other image.

According to the method for screening remote sensing data for area coverage provided by the present invention, before screening out the second target remote sensing image of the target remote sensing image from all other images except the first target remote sensing image in the candidate remote sensing data set according to the number of fragments, imaging time and cloud amount contained in each other image, the method further comprises:

deleting all fragments contained in the first target remote sensing image from the fragmented geometric element set;

and taking the number of fragments obtained by taking the intersection of the fragments contained in each other image and the deleted fragmented geometric element set as the number of fragments of each other image.

According to the method for screening the remote sensing data facing the area coverage, provided by the invention, the second target remote sensing image of the target remote sensing image is screened out from all other images except the first target remote sensing image in the candidate remote sensing data set according to the number of fragments, imaging time and cloud cover contained in each other image, and the method comprises the following steps:

calculating the score of each other image according to the number of fragments of each other image, the imaging time and the cloud cover;

selecting other images with the highest scores to be added into the second target remote sensing image, and deleting all fragments contained in the other images with the highest scores from the deleted fragmented geometric element set to obtain a new deleted fragmented geometric element set;

continuing to combine the fragments contained in the other unselected images with the new deleted fragmented geometric element set to obtain the new fragment number of the other unselected images;

iteratively performing the process of calculating and selecting until the new set of deleted fragmented geometric elements is empty;

and taking the set of all other selected images with the highest scores as the second target remote sensing image.

According to the method for screening the remote sensing data facing the area coverage, the score of each other image is calculated according to the number of fragments of each other image, the imaging time and the cloud cover, and the method comprises the following steps:

for each other image, calculating a time difference value between the imaging time of each other image and a preset time;

weighting and adding the time difference value and the cloud cover;

the number of patches is divided by the weighted addition result, and the division result is taken as the score of each other image.

According to the method for screening the remote sensing data facing the area coverage, the preset non-space conditions comprise that the imaging time is within a first preset range, the cloud cover is within a second preset range and the type of a sensor for obtaining the remote sensing image is the same.

The invention also provides a remote sensing data screening system facing the area coverage, which comprises the following components:

the selection module is used for selecting a remote sensing image which has a spatial intersection relation with a target area and meets a preset non-spatial condition from the remote sensing data set to be screened to generate a candidate remote sensing data set;

the cutting module is used for performing spatial cutting on the space geometric object and the target area of each remote sensing image in the candidate remote sensing data set, and cutting each cut remote sensing image into a plurality of fragments;

the acquisition module is used for carrying out duplicate removal processing on repeated fragments in a cutting result, generating a fragmented geometric element set from all the fragments subjected to duplicate removal processing, carrying out spatial connection on each cut remote sensing image and the fragments in the fragmented geometric element set, and acquiring the fragments contained in each cut remote sensing image;

the screening module is used for screening out target remote sensing images from the candidate remote sensing data set according to the number of fragments, imaging time and cloud cover contained in each cut remote sensing image; and the target area is completely covered by a union result of fragments contained in all target remote sensing images.

The invention also provides an electronic device, which comprises a memory, a processor and a computer program stored on the memory and capable of running on the processor, wherein the processor executes the program to realize the steps of any one of the above remote sensing data screening methods facing the area coverage.

The present invention also provides a non-transitory computer readable storage medium having stored thereon a computer program which, when being executed by a processor, implements the steps of the method for remote sensing data screening for area coverage as described in any of the above.

According to the method and the system for screening the remote sensing data facing the area coverage, on one hand, the target area is directly adopted to cut the space geometric object of each remote sensing image in the candidate remote sensing data set, the part, overlapped with the target area, of each remote sensing image in the candidate remote sensing data set is obtained, the cut remote sensing images are further cut mutually, all fragments covering the target area are obtained, the problem that the size of a grid is inaccurate when the target area is divided by adopting the grid is avoided, and the precision of screening the space coverage is effectively improved; on the other hand, the number of fragments, imaging time and cloud cover contained in each remote sensing image in the candidate remote sensing data set are comprehensively considered to screen the candidate remote sensing data set, so that the screened target remote sensing image has the characteristic of comprehensive optimal number, time, cloud cover and space coverage.

Drawings

In order to more clearly illustrate the technical solutions of the present invention or the prior art, the drawings needed to be used in the description of the embodiments or the prior art will be briefly described below, and it is obvious that the drawings in the following description are some embodiments of the present invention, and it is obvious for those skilled in the art to obtain other drawings based on these drawings without creative efforts.

FIG. 1 is a schematic flow chart of a method for screening remote sensing data facing to area coverage provided by the invention;

FIG. 2 is a schematic structural diagram of distribution of remote sensing images in a candidate remote sensing data set in the remote sensing data screening method for area coverage provided by the invention;

FIG. 3 is a schematic structural diagram of distribution of overlapping portions of each remote sensing image and a target area in the method for screening remote sensing data for area coverage provided by the invention;

FIG. 4 is a schematic structural diagram of mutual cutting among remote sensing images in the remote sensing data screening method facing area coverage provided by the invention;

FIG. 5 is a schematic structural diagram of a process of iteratively selecting the highest remote sensing image as a target remote sensing image in each cycle in the method for screening remote sensing data for area coverage provided by the present invention;

FIG. 6 is a schematic structural diagram of distribution of a target remote sensing image finally selected in the method for screening remote sensing data for area coverage provided by the present invention;

FIG. 7 is a second schematic flowchart of the method for screening remote sensing data oriented to area coverage according to the present invention;

FIG. 8 is a schematic structural diagram of a remote sensing data screening system for area coverage provided by the present invention;

fig. 9 is a schematic structural diagram of an electronic device provided by the present invention.

Detailed Description

In order to make the objects, technical solutions and advantages of the present invention clearer, the technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings, and it is obvious that the described embodiments are some, but not all embodiments of the present invention. All other embodiments, which can be derived by a person skilled in the art from the embodiments given herein without making any creative effort, shall fall within the protection scope of the present invention.

With the increasing number of satellites, the revisit period is shorter and shorter, the research on regional and global changes is wider and wider, the demand on regional time-sequence data is more and more urgent, and how to quickly screen the required remote sensing images from the massive overlapped remote sensing images is a crucial problem at present. The traditional method is that after the remote sensing image is pre-screened, the remote sensing image is selected manually or in a grid-based mode, so that the efficiency is low and the optimal selection result is difficult to achieve. In the embodiment, the computer is used for cutting the space geometric object of the remote sensing image by using the target area to obtain the part overlapped with the target area, the cut remote sensing images are cut mutually and then are subjected to de-duplication to obtain the fragment set for representing the target area, and then according to the number of the fragments, the imaging time and the cloud cover contained in each cut remote sensing image, the optimal remote sensing image is automatically and accurately screened out from the candidate remote sensing image data set, so that the manual selection is avoided, the screening speed is greatly increased, the working efficiency is improved, and the time cost is saved.

The method for screening remote sensing data facing to area coverage in the invention is described below with reference to fig. 1, and comprises the following steps: step 101, selecting a remote sensing image which has a spatial intersection relation with a target area and meets a preset non-spatial condition from a remote sensing data set to be screened to generate a candidate remote sensing data set.

The method for screening remote sensing data for area coverage in the embodiment may be executed by a computer.

The number of remote sensing images contained in the remote sensing data set to be screened is plural, and this embodiment does not limit this.

In the screening process, the spatial geometric objects of the remote sensing images are mainly screened.

Because the remote sensing data set to be screened contains a plurality of remote sensing images with different attribute characteristics, the remote sensing images in the remote sensing data set to be screened can be preliminarily screened to delete the remote sensing images irrelevant to the screened target.

Firstly, judging whether each remote sensing image in a remote sensing data set to be screened is spatially intersected with a target area or not and meets a preset non-spatial condition; and then, taking all remote sensing images which are intersected with the target area space and meet the preset non-space condition as a candidate remote sensing data set so as to screen the candidate remote sensing data set in the subsequent screening process. As shown in fig. 2, the structure diagram of the distribution of each remote sensing image in the candidate remote sensing data set is shown.

The preset non-spatial condition includes whether the imaging time satisfies a condition, whether the cloud amount satisfies a condition, whether the sensor type satisfies a condition, and the like, which is not specifically limited in this embodiment.

The target area is an area needing to be screened, namely a screening target, which is input by a user.

And 102, carrying out spatial cutting on the spatial geometric object of each remote sensing image in the candidate remote sensing data set and the target area, and mutually cutting each cut remote sensing image so as to split each cut remote sensing image into a plurality of fragments.

Specifically, after the remote sensing data set to be screened is subjected to preliminary screening to obtain a candidate remote sensing data set meeting conditions, the target area is used for cutting space geometric objects of all remote sensing images in the candidate remote sensing data set respectively, parts which do not overlap with the target area are cut off, and only overlapping parts of all remote sensing images and the target area are reserved to obtain the overlapping parts of all remote sensing images and the target area in the candidate remote sensing data set. As shown in fig. 3, a schematic diagram of the distribution of the overlapping portion between each remote sensing image and the target area is shown.

Because the boundary of the target area is usually irregular, compared with the existing grid division mode, the method directly cuts the space geometric objects of the remote sensing images by adopting the target area in the example, and can effectively avoid the problem that the boundary of the target area is difficult to be completely and accurately divided due to the fact that the target area is divided by adopting the grid, so that the screening precision is low or the screening efficiency is low.

Then, each cut remote sensing image and all other cut remote sensing images are cut mutually, and each cut remote sensing image can be divided into a plurality of fragments.

As shown in fig. 4, assuming that the overlapping portions of the remote sensing image A, B, C and the target area are both themselves, the remote sensing image A, B, C is mutually divided, and the remote sensing image a is divided into three parts, i.e., a fragment a, a fragment B, and a fragment C, by the remote sensing images B and C.

In this embodiment, the target area is directly used to cut the space geometric object of each remote sensing image to obtain the intersection part of each remote sensing image and the target area space, and the cut remote sensing images are cut mutually, so that each cut remote sensing image is separated into a plurality of fragments.

And 103, carrying out duplicate removal processing on repeated fragments in the cutting result, generating a fragmented geometric element set covering the target area from all the fragments subjected to the duplicate removal processing, and carrying out spatial connection on each remote sensing image subjected to the cutting and the fragments in the fragmented geometric element set to obtain the fragments contained in each remote sensing image subjected to the cutting.

Specifically, because the completely overlapped fragments exist in the cutting result, in order to avoid the completely overlapped fragments from influencing the screening result, the completely overlapped fragments in the cutting result are subjected to a deduplication processing so as to remove one fragment of the two completely overlapped fragments.

And then, performing space connection calculation on each cut remote sensing image and the fragments in the fragmented geometric element set to calculate the fragments contained in each remote sensing image. The condition of spatial connection is that the remote sensing image contains fragments.

Wherein the spatial join computation may be performed in a parallel manner. The step of performing spatially connected computations in parallel includes, first, splitting a set of fragmented geometric elements into a plurality of subsets; then, performing spatial connection calculation on each cut remote sensing image and the fragments of each subset; and finally, performing aggregation operation on all the space connection results to obtain fragments contained in each remote sensing image.

104, screening a target remote sensing image from the candidate remote sensing data set according to the number of fragments, imaging time and cloud cover contained in each remote sensing image in the candidate remote sensing data set; and the target area is completely covered by a union result of fragments contained in all target remote sensing images.

Specifically, the union set of fragments contained in all target remote sensing images can completely cover the target area;

the number of fragments contained in each remote sensing image is the number of fragments obtained by taking the intersection of the fragments contained in each remote sensing image and the fragments of the uncovered target area. Therefore, the larger the number of patches, the larger the contribution of each remote sensing image to the spatial coverage of the target area.

The closer the imaging time of the remote sensing image is to the preset time, the newer the imaging time is.

The less the cloud cover of the remote sensing image, the better the quality of the representation of the remote sensing image.

Therefore, spatial coverage contribution, imaging time and cloud cover may have a certain impact on the screening target.

In the embodiment, the influence of the remote sensing images on the screened target is obtained by integrating the number of the fragments, the imaging time and the cloud cover contained in each remote sensing image, so that the target remote sensing images are accurately and quickly screened from the candidate remote sensing data set, the union set of the fragments contained in all the target remote sensing images can cover the whole target area, and the number of the target remote sensing images is minimum, the imaging time is latest and the cloud cover is minimum.

On one hand, the space geometric objects of the remote sensing images in the candidate remote sensing data set are cut by directly adopting the target area, the parts, overlapped with the target area, of the remote sensing images in the candidate remote sensing data set are obtained, the parts, overlapped with the target area, are further cut mutually, all fragments covering the target area are obtained, the problem that the size of a grid is inaccurate when the target area is divided by adopting the grid is avoided, and the screening precision and efficiency are effectively improved; on the other hand, the number of fragments, imaging time and cloud cover contained in each remote sensing image in the candidate remote sensing data set are comprehensively considered to screen the candidate remote sensing data set, so that the screened target remote sensing image has the characteristic of comprehensive optimal number, time, cloud cover and space coverage.

On the basis of the foregoing embodiment, the mutually cutting the remote sensing images after being cut in this embodiment includes: combining the cut remote sensing images pairwise to form a combined remote sensing data set, and mutually cutting the two cut remote sensing images in each combined remote sensing data set; combining the cutting results of the combined remote sensing data sets in pairs to form new combined remote sensing data sets, continuously cutting the two cutting results in each new combined remote sensing data set mutually again, and iteratively executing the processes of combining in pairs and cutting mutually until each remote sensing image after being cut is mutually cut with other remote sensing images after being cut; and the other cut remote sensing images are remote sensing images except the cut remote sensing images.

Because the mutual cutting process between every two cut remote sensing images is independent, a plurality of distributed computers can be adopted to concurrently execute the mutual cutting process, and the algorithm execution efficiency is improved.

On the basis of the foregoing embodiment, in this embodiment, the screening out a target remote sensing image from the candidate remote sensing data set according to the number of pieces, imaging time, and cloud amount included in each cut remote sensing image includes: if any fragment in the fragmented geometric element set belongs to the unique remote sensing image, taking the remote sensing image to which the fragment belongs as a first target remote sensing image of the target remote sensing image; and screening out a second target remote sensing image of the target remote sensing image from all other images except the first target remote sensing image in the candidate remote sensing data set according to the number of fragments, imaging time and cloud cover contained in each other image.

Specifically, in order to ensure complete coverage of a target area, whether any fragment in the fragmented geometric element set only belongs to one remote sensing image but not to other remote sensing images is judged; and if so, taking the remote sensing image to which the fragment belongs as a first target remote sensing image of the target remote sensing image, and preferably selecting the remote sensing image from the candidate remote sensing data set.

And then, continuously screening other images except the first target remote sensing image in the candidate remote sensing data set.

In the embodiment, the influence of the number of fragments of each other image, the imaging time and the cloud cover on the screened target is comprehensively considered, so that the second target remote sensing image of the target remote sensing image is continuously and accurately and quickly screened from all other images.

By the screening method, the screened remote sensing images can completely cover the target area after being collected, and the screened remote sensing images have the comprehensive optimal characteristics of quantity, time, cloud cover and space coverage, so that the accuracy of the screened remote sensing images is higher.

On the basis of the foregoing embodiment, in this embodiment, before screening out the second target remote sensing image of the target remote sensing image from all other images in the candidate remote sensing dataset except the first target remote sensing image according to the number of fragments, imaging time, and cloud amount included in each other image, the method further includes: deleting all fragments contained in the first target remote sensing image from the fragmented geometric element set; and taking the number of fragments obtained by taking the intersection of the fragments contained in each other image and the deleted fragmented geometric element set as the number of fragments of each other image.

Specifically, after the first target remote sensing image is screened from the candidate remote sensing data set, in order to avoid interference of fragments contained in the screened first target remote sensing image on subsequent screening, all the fragments contained in the first target remote sensing image are deleted from the fragmented geometric element set.

Then, the number of fragments in the intersection of the fragments contained in each other image and the deleted fragmented geometric element set is counted, and the number is used as the number of fragments contained in each other image. The statistical number of fragments of each other image can be used to characterize its contribution rate to the deleted fragmented geometric element set.

On the basis of the foregoing embodiment, in this embodiment, screening out the second target remote sensing image of the target remote sensing image from all other images, except the first target remote sensing image, in the candidate remote sensing dataset according to the number of fragments, imaging time, and cloud amount included in each other image, includes: calculating the score of each other image according to the number of fragments of each other image, the imaging time and the cloud cover; selecting other images with the highest scores to be added into the second target remote sensing image, and deleting all fragments contained in the other images with the highest scores from the deleted fragmented geometric element set to obtain a new deleted fragmented geometric element set; continuing to combine the fragments contained in the other unselected images with the new deleted fragmented geometric element set to obtain the new fragment number of the other unselected images; iteratively performing the process of calculating and selecting until the new set of deleted fragmented geometric elements is empty; and taking the set of all other selected images with the highest scores as the second target remote sensing image.

Specifically, when the second target remote sensing image is screened from all other images, the second target remote sensing image needs to be screened from all other images according to the score of each other image.

The number of fragments contained in the image, the imaging time and the cloud cover can be fused to obtain the score of each other image.

The fusion mode may be subtracting the preset time from the imaging time, adding the subtraction result to the number of fragments, and dividing the number of fragments by the addition result; or, the subtraction result and the number of fragments are added in a weighted manner, and the number of fragments is divided by the weighted addition result, which does not specifically limit the fusion manner in this embodiment.

The step of screening out a second target remote sensing image from all other images comprises the following steps:

step (1), counting the number of fragments in the intersection of the fragments of each other image and the deleted fragmented geometric element set, and taking the number of fragments as the number of fragments contained in each other image;

step (2), calculating the score of each other image according to the number of fragments contained in each other image, the imaging time and the cloud cover;

step (3), screening the other images with the highest scores from all other images, and deleting all the fragments contained in the other images with the highest scores from the deleted fragmented geometric element set;

step (4), taking the re-deleted fragmented geometric element set as a new fragmented geometric element set, and judging whether the new fragmented geometric element set is empty or not; if not, turning to the step (1), and continuously screening all other images except the other image with the highest score; if so, the loop is ended.

And finally, taking all other images with the highest scores selected in the circulation process as second target remote sensing images. And all the screened first target remote sensing images and second target remote sensing images are the final screening result.

As shown in fig. 5, a schematic structural diagram of a process of selecting the highest remote sensing image as a target remote sensing image for each loop iteration; fig. 6 is a schematic structural diagram of coverage of the target area by all the finally screened target remote sensing images. The rectangular portions in fig. 5 and 6 represent the first target remote sensing image or the second target remote sensing image, and the region formed by the curves is the target region.

The dense areas of multiple overlapping in fig. 6 are for making the spatial union of all the screened target remote sensing images completely cover the target area, and avoiding gaps between the images.

Through the screening method, the whole target area can be completely covered by the spatial union set of fragments contained in all the screened first target remote sensing images and the screened second target remote sensing images.

When the first target remote sensing image or the second target remote sensing image is screened from the candidate remote sensing data set each time, all fragments of the screened first target remote sensing image or the screened second target remote sensing image need to be deleted from the fragmented geometric element set, so that the screened target remote sensing image is effectively ensured to completely cover the target area. When the second target remote sensing image is screened, the second target remote sensing image screened each time is the image with the closest imaging time and the largest number of fragments in the deleted fragmented geometric element set in the rest of the un-screened other remote sensing images, that is, the contribution rate to space coverage is the largest, and the cloud amount is the smallest.

For the problem of remote sensing image screening in a large target area, data omission or multiple selection is easily caused by manual selection. In the embodiment, the first target remote sensing image is preferentially screened, then, in an iterative loop mode, the second target remote sensing image selected in each step is the image which has the closest imaging time and the largest contribution rate to the residual fragmented geometric element set, and when the target area is completely covered, the loop is terminated, so that the problems of selection omission or multiple selections are avoided.

In the embodiment, the number of fragments, the imaging time and the cloud cover are fused to calculate the score of each remote sensing image in the candidate remote sensing data set, so that the second target remote sensing image is circularly screened out from the candidate remote sensing data set according to the score, the condition of missing or multiple selections is avoided, and the screening result is more accurate; and the imaging time difference of the remote sensing images can be smaller, the ground feature difference between the remote sensing images is effectively reduced, and the accuracy is improved. And the quantity of the screened remote sensing images is less, and the complexity of the subsequent remote sensing image mosaic work is reduced.

On the basis of the above embodiment, in this embodiment, calculating the score of each other image according to the number of fragments, the imaging time, and the cloud amount of each other image includes: for each other image, calculating a time difference value between the imaging time of each other image and a preset time; weighting and adding the time difference value and the cloud cover; the number of patches is divided by the weighted addition result, and the division result is taken as the score of each other image.

The formula for calculating the score of any other image is as follows:

Cost=N/(w1*|T-T0|+w2*C)；

wherein Cost is the score of the other images, w1 and w2 are weight coefficients, T, C and N are the imaging time, cloud cover and the number of fragments of the other images respectively; the number of the fragments is the number of the fragments in the fragmented geometric element set of the other images containing the deleted fragments; t0 is a preset time. The preset time is the optimal time of the remote sensing image to be acquired.

W1 and w2 can be set according to actual requirements, and can also be obtained by model calculation. For example, the imaging time and the preset time of each of the other images are input into the model, the similarity between the imaging time and the preset time is calculated, and the weight value is obtained.

Wherein the imaging time may be inclusive of year and day, such as 200 th day of 2020; it may include year, month and day, and this embodiment is not particularly limited thereto.

In the case where the imaging time includes year and day, and the preset time also includes year and day, the time difference is calculated by the formula:

|T-T0|=|Y-Y0|+|D-D0|；

wherein, Y and D are respectively the year and the day of the imaging time of the other images, and Y0 and D0 are respectively the year and the day of the preset time.

On the basis of the foregoing embodiments, in this embodiment, the preset non-spatial condition includes that the imaging time is within a first preset range, the cloud cover is within a second preset range, and the type of the sensor for acquiring the remote sensing image is the same.

Specifically, due to differences in imaging principles, parameter settings, technical requirements, and the like of the sensors, the remote sensing images acquired by the sensors have large differences. Therefore, when the remote sensing images in the remote sensing data set to be screened come from different sensors, the remote sensing images collected by the same type of sensors are selected as much as possible.

When the cloud amount of the remote sensing image is large, the quality of the remote sensing image is seriously influenced, and the usability of the remote sensing image is poor.

Therefore, the remote sensing data set to be screened can be preliminarily screened by integrating the overlapping part with the target area, the imaging time, the cloud cover and the type of the sensor, so that a better remote sensing image can be screened as much as possible.

As shown in fig. 7, the method for screening remote sensing data for area coverage provided by this embodiment includes the specific steps of:

the method comprises the following steps that (1) a remote sensing data set to be screened is preliminarily screened according to conditions such as a target area, imaging time, cloud cover and sensor type, and a candidate remote sensing data set is obtained;

step (2), cutting a space geometric object of each remote sensing image of the candidate remote sensing data set by using the target area to obtain an overlapped part of each remote sensing image and the target area;

mutually cutting each cut remote sensing image and other cut remote sensing images to split the overlapped part of each remote sensing image and the target area into a plurality of fragments, and generating a fragmented geometric element set after carrying out duplication removal treatment on the fragments in the mutual cutting treatment result;

step (4), performing spatial connection on each cut remote sensing image and fragments in the fragmented geometric element set to obtain fragments contained in each remote sensing image;

step (5), when the fragments in the fragmented geometric element set are only contained by one remote sensing image but not other remote sensing images, taking the remote sensing image as a first target remote sensing image, preferably selecting the first target remote sensing image from the candidate remote sensing data set, and deleting all the fragments contained in the first target remote sensing image from the fragmented geometric element set;

step (6), calculating the number of fragments in each screened remote sensing image, wherein each remote sensing image comprises the deleted fragmented geometric element set; calculating the difference value between the imaging time and the preset time, and calculating the score of each remote sensing image after screening according to the number of the fragments, the difference result and the cloud cover;

step (7), traversing all the screened remote sensing images, selecting the remote sensing image with the highest score as a second target remote sensing image, and deleting all fragments contained in the second target remote sensing image from the fragmented geometric element set again;

step (8), judging whether the new fragmented geometric element set is empty or not; and (6) if the number of the detected signals is not equal to the preset value, jumping to the step (6).

The remote sensing data screening system for area coverage provided by the invention is described below, and the remote sensing data screening system for area coverage described below and the remote sensing data screening method for area coverage described above can be referred to correspondingly.

As shown in fig. 8, the present embodiment provides an area coverage-oriented remote sensing data screening system, which includes a selection module 801, a cutting module 802, an obtaining module 803, and a screening module 804, where:

the selection module 801 is configured to select a remote sensing image from the remote sensing data set to be screened, where the remote sensing image has a spatial intersection relationship with the target area and meets a preset non-spatial condition, to generate a candidate remote sensing data set.

In the screening process, the spatial geometric objects of the remote sensing images are mainly screened. Because the remote sensing data set to be screened contains a plurality of remote sensing images with different attribute characteristics, the remote sensing images in the remote sensing data set to be screened can be preliminarily screened to delete the remote sensing images irrelevant to the screened target.

The target area is an area needing to be screened and input by a user, namely a screening target;

the cutting module 802 is configured to perform spatial cutting on the spatial geometric object of each remote sensing image in the candidate remote sensing data set and the target area, and perform mutual cutting on each cut remote sensing image, so as to split each cut remote sensing image into a plurality of fragments.

The obtaining module 803 is configured to perform deduplication processing on repeated fragments in the cutting result, generate a fragmented geometric element set from all fragments subjected to deduplication processing, perform spatial connection on each cut remote sensing image and fragments in the fragmented geometric element set, and obtain fragments included in each cut remote sensing image.

Specifically, because the completely overlapped fragments exist in the cutting result, in order to avoid the completely overlapped fragments from affecting the screening result, the completely overlapped fragments in the cutting result are subjected to a duplicate removal treatment so as to remove one fragment of the two completely overlapped fragments;

The screening module 804 is used for screening out target remote sensing images from the candidate remote sensing data set according to the number of fragments, imaging time and cloud cover contained in each cut remote sensing image; and the target area is completely covered by a union result of fragments contained in all target remote sensing images.

The closer the imaging time of the remote sensing image is to the preset time, the newer the imaging time is. The less the cloud cover of the remote sensing image, the better the quality of the representation of the remote sensing image.

In addition, in the prior art, the remote sensing data is only screened according to the spatial overlapping rate of the remote sensing image and the target area, the maximum spatial overlapping rate of the screened remote sensing image and the target area can only be ensured, the optimal spatial coverage can not be ensured, the minimum number of the remote sensing images and the optimal cloud cover and time can not be ensured, and the optimal remote sensing image can not be screened. In the embodiment, the remote sensing image is screened by comprehensively considering the number of the fragments, the imaging time and the cloud cover, so that the screened target remote sensing image has the characteristic of comprehensively optimal image number, time, cloud cover and space coverage, and the screening requirement can be better met.

On the basis of the above embodiments, the cutting module in this example is specifically configured to: combining the cut remote sensing images pairwise to form a combined remote sensing data set, and mutually cutting the two cut remote sensing images in each combined remote sensing data set; combining the cutting results of the combined remote sensing data sets in pairs to form new combined remote sensing data sets, continuously cutting the two cutting results in each new combined remote sensing data set mutually again, and iteratively executing the processes of combining in pairs and cutting mutually until each remote sensing image after being cut is mutually cut with other remote sensing images after being cut; and the other cut remote sensing images are remote sensing images except the cut remote sensing images.

On the basis of the above embodiment, the screening module in this example is specifically configured to: if any fragment in the fragmented geometric element set belongs to the unique remote sensing image, taking the remote sensing image to which the fragment belongs as a first target remote sensing image of the target remote sensing image; and screening out a second target remote sensing image of the target remote sensing image from all other images except the first target remote sensing image in the candidate remote sensing data set according to the number of fragments, imaging time and cloud cover contained in each other image.

On the basis of the above embodiment, the present example further includes a deleting module for: deleting all fragments contained in the first target remote sensing image from the fragmented geometric element set; and taking the number of fragments obtained by taking the intersection of the fragments contained in each other image and the deleted fragmented geometric element set as the number of fragments of each other image.

On the basis of the above embodiment, the screening module in this example is further configured to: calculating the score of each other image according to the number of fragments, the imaging time and the cloud cover contained in each other image; selecting other images with the highest scores to be added into the second target remote sensing image, and deleting all fragments contained in the other images with the highest scores from the deleted fragmented geometric element set to obtain a new deleted fragmented geometric element set; continuing to combine the fragments contained in the other unselected images with the new deleted fragmented geometric element set to obtain the new fragment number of the other unselected images; iteratively performing the process of calculating and selecting until the new set of deleted fragmented geometric elements is empty; and taking the set of all other selected images with the highest scores as the second target remote sensing image.

On the basis of the above embodiment, the present example further includes a calculation module, configured to: for each other image, calculating a time difference value between the imaging time of each other image and a preset time; weighting and adding the time difference value and the cloud cover; the number of patches is divided by the weighted addition result, and the division result is taken as the score of each other image.

On the basis of the foregoing embodiments, in this example, the preset non-spatial conditions include that the imaging time is within a first preset range, the cloud cover is within a second preset range, and the types of the sensors for acquiring the remote sensing image are the same.

Fig. 9 illustrates a physical structure diagram of an electronic device, and as shown in fig. 9, the electronic device may include: a processor (processor)901, a communication Interface (Communications Interface)902, a memory (memory)903 and a communication bus 904, wherein the processor 901, the communication Interface 902 and the memory 903 are communicated with each other through the communication bus 904. Processor 901 may invoke logic instructions in memory 903 to perform a method for region coverage oriented remote sensing data screening, the method comprising: selecting a remote sensing image which has a spatial intersection relation with a target area and meets a preset non-spatial condition from a remote sensing data set to be screened to generate a candidate remote sensing data set; carrying out space cutting on each remote sensing image in the candidate remote sensing data set and the target area, and mutually cutting each cut remote sensing image so as to split each cut remote sensing image into a plurality of fragments; carrying out duplicate removal processing on repeated fragments in a cutting result, generating a fragmented geometric element set from all the fragments subjected to duplicate removal processing, and carrying out spatial connection on each remote sensing image subjected to cutting and the fragments in the fragmented geometric element set to obtain the fragments contained in each remote sensing image subjected to cutting; according to the number of fragments, imaging time and cloud cover contained in each cut remote sensing image, screening out a target remote sensing image from the candidate remote sensing data set; and the target area is completely covered by a union result of fragments contained in all target remote sensing images.

In addition, the logic instructions in the memory 903 may be implemented in a software functional unit and stored in a computer readable storage medium when the logic instructions are sold or used as a separate product. Based on such understanding, the technical solution of the present invention may be embodied in the form of a software product, which is stored in a storage medium and includes instructions for causing a computer device (which may be a personal computer, a server, or a network device) to execute all or part of the steps of the method according to the embodiments of the present invention. And the aforementioned storage medium includes: a U-disk, a removable hard disk, a Read-Only Memory (ROM), a Random Access Memory (RAM), a magnetic disk or an optical disk, and other various media capable of storing program codes.

In another aspect, the present invention further provides a computer program product, the computer program product includes a computer program stored on a non-transitory computer-readable storage medium, the computer program includes program instructions, when the program instructions are executed by a computer, the computer can execute the method for filtering remote sensing data facing to area coverage provided by the above methods, the method includes: selecting a remote sensing image which has a spatial intersection relation with a target area and meets a preset non-spatial condition from a remote sensing data set to be screened to generate a candidate remote sensing data set; carrying out space cutting on each remote sensing image in the candidate remote sensing data set and the target area, and mutually cutting each cut remote sensing image so as to split each cut remote sensing image into a plurality of fragments; carrying out duplicate removal processing on repeated fragments in a cutting result, generating a fragmented geometric element set from all the fragments subjected to duplicate removal processing, and carrying out spatial connection on each remote sensing image subjected to cutting and the fragments in the fragmented geometric element set to obtain the fragments contained in each remote sensing image subjected to cutting; according to the number of fragments, imaging time and cloud cover contained in each cut remote sensing image, screening out a target remote sensing image from the candidate remote sensing data set; and the target area is completely covered by a union result of fragments contained in all target remote sensing images.

In yet another aspect, the present invention further provides a non-transitory computer-readable storage medium, on which a computer program is stored, the computer program being implemented by a processor to execute the above-mentioned method for filtering remote sensing data facing to regional coverage, the method comprising: selecting a remote sensing image which has a spatial intersection relation with a target area and meets a preset non-spatial condition from a remote sensing data set to be screened to generate a candidate remote sensing data set; carrying out space cutting on each remote sensing image in the candidate remote sensing data set and the target area, and mutually cutting each cut remote sensing image so as to split each cut remote sensing image into a plurality of fragments; carrying out duplicate removal processing on repeated fragments in a cutting result, generating a fragmented geometric element set from all the fragments subjected to duplicate removal processing, and carrying out spatial connection on each remote sensing image subjected to cutting and the fragments in the fragmented geometric element set to obtain the fragments contained in each remote sensing image subjected to cutting; according to the number of fragments, imaging time and cloud cover contained in each cut remote sensing image, screening out a target remote sensing image from the candidate remote sensing data set; and the target area is completely covered by a union result of fragments contained in all target remote sensing images.

The above-described system embodiments are merely illustrative, and the units described as separate parts may or may not be physically separate, and parts displayed as units may or may not be physical units, may be located in one place, or may be distributed on a plurality of network units. Some or all of the modules may be selected according to actual needs to achieve the purpose of the solution of the present embodiment. One of ordinary skill in the art can understand and implement it without inventive effort.

Through the above description of the embodiments, those skilled in the art will clearly understand that each embodiment can be implemented by software plus a necessary general hardware platform, and certainly can also be implemented by hardware. With this understanding in mind, the above-described technical solutions may be embodied in the form of a software product, which can be stored in a computer-readable storage medium such as ROM/RAM, magnetic disk, optical disk, etc., and includes instructions for causing a computer device (which may be a personal computer, a server, or a network device, etc.) to execute the methods described in the embodiments or some parts of the embodiments.

Finally, it should be noted that: the above examples are only intended to illustrate the technical solution of the present invention, but not to limit it; although the present invention has been described in detail with reference to the foregoing embodiments, it will be understood by those of ordinary skill in the art that: the technical solutions described in the foregoing embodiments may still be modified, or some technical features may be equivalently replaced; and such modifications or substitutions do not depart from the spirit and scope of the corresponding technical solutions of the embodiments of the present invention.

Claims

1. A remote sensing data screening method facing area coverage is characterized by comprising the following steps:

2. The method for screening remote sensing data for area coverage according to claim 1, wherein the mutually cutting the cut remote sensing images comprises:

combining the cutting results of the combined remote sensing data sets in pairs to form new combined remote sensing data sets, continuously cutting the two cutting results in each new combined remote sensing data set mutually again, and iteratively executing the processes of combining in pairs and cutting mutually until each remote sensing image after being cut is mutually cut with other remote sensing images after being cut; and the other cut remote sensing images are remote sensing images except the cut remote sensing images in all the cut remote sensing images.

3. The method for screening remote sensing data for area coverage according to claim 1, wherein screening out a target remote sensing image from the candidate remote sensing data set according to the number of fragments, imaging time and cloud cover contained in each cut remote sensing image comprises:

4. The method for screening remote sensing data for area coverage according to claim 3, wherein before screening out a second target remote sensing image of the target remote sensing image from all other images except the first target remote sensing image in the candidate remote sensing data set according to the number of fragments, imaging time and cloud amount contained in each other image, the method further comprises:

5. The method for screening remote sensing data for area coverage according to claim 4, wherein screening out a second target remote sensing image of the target remote sensing image from all other images except the first target remote sensing image in the candidate remote sensing data set according to the number of fragments, imaging time and cloud amount contained in each other image comprises:

6. The method for screening remote sensing data for area coverage according to claim 5, wherein the calculating the score of each other image according to the number of fragments, the imaging time and the cloud cover of each other image comprises:

weighting and adding the time difference value and the cloud cover;

7. The method for screening remote sensing data for area coverage according to any one of claims 1 to 6, wherein the predetermined non-spatial conditions include that an imaging time is within a first predetermined range, a cloud cover is within a second predetermined range, and a type of a sensor for acquiring the remote sensing image is the same type.

8. An area coverage oriented remote sensing data screening system, comprising:

9. An electronic device comprising a memory, a processor and a computer program stored on the memory and executable on the processor, wherein the processor implements the steps of the method for remote sensing data screening for area coverage according to any one of claims 1 to 7 when executing the program.

10. A non-transitory computer readable storage medium, on which a computer program is stored, wherein the computer program, when being executed by a processor, implements the steps of the method for remote sensing data filtering for area coverage according to any one of claims 1 to 7.