Remote sensing image screening method and device

Abstract

The invention relates to a remote sensing image screening method and device, and belongs to the technical field of remote sensing image screening, and the method comprises the steps: obtaining a target region and a plurality of first remote sensing images corresponding to the target region; wherein the plurality of first remote sensing images are respectively overlapped with the target area, and an area formed by the plurality of first remote sensing images covers the target area; segmenting the plurality of first remote sensing images according to the overlapping condition of the plurality of first remote sensing images to obtain a plurality of segmented regions; wherein the segmentation areas are not overlapped with each other; determining a second remote sensing image from the plurality of first remote sensing images; wherein the second remote sensing image singly comprises at least one segmented region or at most segmented regions. The first remote sensing images are screened based on the segmented regions, redundant remote sensing images in the first remote sensing images can be removed, meanwhile, it is guaranteed that the reserved first remote sensing images can completely cover the target region, and the accuracy of the screening process is guaranteed.

Description

Technical Field

[0001] The present invention relates to the technical field of remote sensing image screening, and particularly to a remote sensing image screening method and device. Background Art

[0002] With the rapid development of space remote sensing technology, the collected remote sensing image data gradually presents characteristics such as multi-source, multi-scale, multi-temporal, global coverage, and high resolution, making the remote sensing image data show a trend of big data and massive amounts. Remote sensing technology is widely used in fields such as land resource survey, mineral resource exploration, military reconnaissance, crop yield estimation, disaster monitoring, weather forecasting, etc. These applications are all based on image data sets that have been screened to meet the requirements. However, due to the large amount of remote sensing image data, it is difficult to obtain images that meet the application scenarios from the massive data. Especially when the area of the region of interest is large, the image screening is extremely time-consuming and laborious. Therefore, in remote sensing applications in different scenarios, how to efficiently and accurately screen images that meet specific remote sensing monitoring tasks from massive remote sensing images is crucial. Summary of the Invention

[0003] In view of this, it is necessary to provide a remote sensing image screening method and device to solve the problem that the screening is not accurate enough when screening images that meet specific remote sensing monitoring tasks from massive remote sensing images in the prior art.

[0004] To solve the above problems, in a first aspect, the present invention provides a remote sensing image screening method, including: Obtain a target area and a plurality of first remote sensing images corresponding to the target area; wherein, the plurality of first remote sensing images respectively overlap with the target area, and the area formed by the plurality of first remote sensing images completely covers the target area; According to the overlapping situation among the plurality of first remote sensing images, segment the plurality of first remote sensing images to obtain a plurality of segmented areas; wherein, each of the segmented areas does not overlap with each other; Determine a second remote sensing image from the plurality of first remote sensing images; wherein, the second remote sensing image solely contains at least one of the segmented areas or contains the most of the segmented areas.

[0005] Optionally, the determining a second remote sensing image from the plurality of first remote sensing images includes: Determine a two-dimensional array A; wherein, the number of rows of A is the number of first remote sensing images, the number of columns of A is the number of segmented areas, and for the elements in A corresponding to the first remote sensing images and segmented areas with an inclusion relationship, the value is assigned as 1, otherwise the value is assigned as 0; Determine the second remote sensing image through the following formula:

[0006] Wherein, m is the serial number of the first remote sensing image, and n is the number of segmented regions; Take the first remote sensing image that meets the condition as the second remote sensing image.

[0007] Optionally, there are multiple target regions, and the multiple target regions are obtained by splitting the region of interest.

[0008] Optionally, the obtaining of the multiple first remote sensing images corresponding to the target region includes: Obtain multiple initial remote sensing images associated with the region of interest, and the quality scores of each of the initial remote sensing images; Determine a target initial remote sensing image from each of the initial remote sensing images according to the quality scores of each of the initial remote sensing images; Determine whether the target initial remote sensing image intersects with the region of interest; If the target initial remote sensing image intersects with the region of interest, retain the target initial remote sensing image, determine the intersection region between the target initial remote sensing image and the region of interest, and remove the intersection region from the region of interest to update the region of interest; If the target initial remote sensing image does not intersect with the region of interest, delete the target initial remote sensing image; Return to the step of "determining a target initial remote sensing image from each of the initial remote sensing images according to the quality scores of each of the initial remote sensing images" until the intersection region is equal to the region of interest; Determine multiple first remote sensing images corresponding to the target region from the retained target initial remote sensing images.

[0009] Optionally, the obtaining of the multiple initial remote sensing images and the quality scores of each of the initial remote sensing images includes: Obtain multiple initial remote sensing images, and determine the quality scores of each of the initial remote sensing images according to the cloud amount, resolution and imaging time of each of the initial remote sensing images.

[0010] Optionally, the determining of the quality scores of each of the initial remote sensing images according to the cloud amount, resolution and imaging time of each of the initial remote sensing images includes: Calculate the quality scores of each of the initial remote sensing images through the following formula:

[0011] Wherein, is the quality score; They are respectively the cloud cover score, resolution score, and imaging time score of the initial remote sensing image; the cloud cover score is determined according to the cloud cover, the resolution score is determined according to the resolution, and the imaging time score is determined according to the imaging time; They are respectively the weights corresponding to the cloud cover score, resolution score, and imaging time score.

[0012] Optionally, the method further includes: Based on the second remote sensing images corresponding to each of the target regions, perform duplicate removal processing on the second remote sensing images to obtain the third remote sensing image corresponding to the region of interest.

[0013] Optionally, the method further includes: Determine a fourth remote sensing image from the third remote sensing image; wherein, the region formed by the fourth remote sensing images covers the region of interest, and the number of the fourth remote sensing images reaches the minimum value.

[0014] Optionally, the determining the fourth remote sensing image from the third remote sensing image includes: Determine a target third remote sensing image from the third remote sensing image; Judge whether the region formed by the remaining third remote sensing images after removing the target third remote sensing image can cover the region of interest; If the region formed by the remaining third remote sensing images covers the region of interest, then remove the target third remote sensing image from the third remote sensing image; If the region formed by the remaining third remote sensing images cannot cover the region of interest, then retain the target third remote sensing image; Return to the step of "determine a target third remote sensing image from the third remote sensing image" until the third remote sensing image is traversed; Use the remaining third remote sensing images as the fourth remote sensing image.

[0015] In a second aspect, the present invention further provides a remote sensing image screening device, including: A first remote sensing image determination module, configured to obtain a target region and a plurality of first remote sensing images corresponding to the target region; wherein, the plurality of first remote sensing images respectively overlap with the target region, and the region formed by the plurality of first remote sensing images completely covers the target region; A segmentation module, configured to segment the plurality of first remote sensing images according to the overlapping situation among the plurality of first remote sensing images to obtain a plurality of segmented regions; wherein, each of the segmented regions does not overlap with each other; The second remote sensing image determination module is configured to determine a second remote sensing image from the multiple first remote sensing images; wherein, the second remote sensing image solely includes at least one of the segmentation regions or includes the most of the segmentation regions.

[0016] The beneficial effects of the present invention are as follows: Based on the segmentation regions, the embodiments of the present invention screen the first remote sensing images, and only retain the first remote sensing images that include the most segmentation regions and solely include the segmentation regions, so as to eliminate redundant remote sensing images in the first remote sensing images, and at the same time ensure that the retained first remote sensing images can completely cover the target region, ensuring the accuracy of the screening process. Description of the Drawings

[0017] Figure 1 It is a schematic flowchart of an embodiment of the remote sensing image screening method provided by the present invention; Figure 2 It is a schematic flowchart of a segmentation region division method provided by the present invention; Figure 3 It is a schematic flowchart of a multi-threaded sub-region screening provided by the present invention; Figure 4 For the present invention Figure 1 It is a schematic flowchart of an embodiment of S101 in the present invention; Figure 5 It is a schematic flowchart of an initial remote sensing image screening method provided by the present invention; Figure 6 It is a set covering model diagram provided by the present invention; Figure 7 It is a schematic diagram of the optimization process of a third remote sensing image provided by the present invention; Figure 8 It is a schematic flowchart of a complete screening operation provided by the present invention; Figure 9 (a) It is a schematic diagram of comparing the screening efficiencies of multiple screening methods provided by the present invention; Figure 9 (b) It is a schematic diagram of comparing the screening redundancies of multiple screening methods provided by the present invention; Figure 10 It is a schematic structural diagram of an embodiment of the remote sensing image screening device provided by the present invention. Detailed Embodiments

[0018] Next, the technical solutions in the embodiments of the present invention will be clearly and completely described in conjunction with the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative efforts fall within the protection scope of the present invention.

[0019] In the description of the embodiments of the present invention, unless otherwise specified, the meaning of "a plurality of" is two or more. The "first", "second", etc. involved in the embodiments of the present invention are used to distinguish similar objects, rather than to describe a specific order or sequence, nor to indicate or imply their relative importance or implicitly specify the quantity of the indicated technical features. It should be understood that such data can be interchanged under appropriate circumstances so that the embodiments of the present application can be implemented in an order other than those illustrated or described herein, and the objects distinguished by "first", "second", etc. are usually of the same type, and the number of objects is not limited. For example, the first object can be one or more.

[0020] Referring to "", a flowchart of an embodiment of the remote sensing image screening method provided by the present invention is shown. The method includes:

[0021] Refer to Figure 1 , a flowchart of an embodiment of the remote sensing image screening method provided by the present invention is shown. The method includes: S101, obtain a target area and a plurality of first remote sensing images corresponding to the target area; wherein, the plurality of first remote sensing images respectively overlap with the target area, and the area formed by the plurality of first remote sensing images completely covers the target area.

[0022] The target area can be an area of interest or a part of the area of interest. When the area of interest is large, the area of interest can be divided to obtain a plurality of target areas, and then S101-S103 are respectively executed on the plurality of target areas, so that the second remote sensing images corresponding to each target area can be screened from the first remote sensing images corresponding to each target area. For ease of understanding, the following takes one target area as an example to elaborate on the specific processes of S102-S103.

[0023] S102, according to the overlapping situation among the plurality of first remote sensing images, segment the plurality of first remote sensing images to obtain a plurality of segmented areas; wherein, the segmented areas do not overlap with each other.

[0024] The first remote sensing image is an image that overlaps with the target area, and the areas formed by each first remote sensing image can completely cover the corresponding target area. Due to the large amount of remote sensing images, there may be problems of high image overlap and high redundancy among the first remote sensing images corresponding to the target area. Therefore, the first remote sensing images corresponding to the target area can be screened.

[0025] Specifically, according to the overlapping situation among multiple first remote sensing images, the multiple first remote sensing images can be segmented to obtain multiple segmented areas, and then the first remote sensing images are screened based on the segmented areas; among them, the multiple segmented areas can be the smallest segmented areas, and the smallest segmented area is the area that cannot be further segmented according to the overlapping situation among multiple first remote sensing images, that is, the non-segmentable area, and the segmented areas do not overlap with each other.

[0026] Refer to Figure 2 , which shows a schematic flowchart of a method for dividing segmented areas provided by the present invention. There are a total of 5 scenes of first remote sensing images. When the second scene of the first remote sensing image is added, the second scene of the first remote sensing image overlaps with the first scene of the first remote sensing image. At this time, according to the overlapping situation between the two, 3 segmented areas can be obtained, namely the overlapping area, the non-overlapping area in the first scene of the first remote sensing image, and the non-overlapping area in the second scene of the first remote sensing image. And so on. When the fifth scene of the first remote sensing image is added, the first to fifth scenes of the first remote sensing images are divided into a total of 19 segmented areas, and at this time, the first scene of the first remote sensing image is completely covered by the area formed by the second to fifth scenes of the first remote sensing images, indicating that the first scene of the first remote sensing image is redundant data and can be removed.

[0027] In one example, the above division process can also be expressed by a formula. Let the set of segmented areas after dividing the first scenes of the first remote sensing image be , and the th scene of the image be . First, obtain the area in the set that intersects with the image . Let , will be divided into two parts, , . Update the set , process and update each area in . The part that is not included in is added as a new area to to obtain a new area division result . .

[0028] S103. Determine a second remote sensing image from multiple first remote sensing images, where the second remote sensing image solely contains at least one segmented area or contains the most segmented areas not yet covered.

[0029] When it is said that the second remote sensing image solely contains a segmented area, it means that a certain segmented area is completely covered by the second remote sensing image and only by the second remote sensing image. For example, taking the segmented area F as an example, if a first remote sensing image A completely covers F and F is only completely covered by A, then A can be used as the second remote sensing image.

[0030] The second remote sensing image can also be the first remote sensing image that contains the most segmented areas. Containing a segmented area means completely covering the segmented area. For example, the segmented areas include F1, F2, F3, and the first remote sensing images include A1, A2, A3. Among them, A1 completely covers F1, F2, F3, A2 completely covers F1, F2, and A3 completely covers F2, F3. Then A1 contains the most segmented areas, and A1 can be selected as the second remote sensing image.

[0031] In the above example, the first remote sensing image of the first scene does not meet the condition of solely containing at least one segmented area.

[0032] In the embodiments of the present invention, by dividing the region of interest into several target regions and performing screening by region, the screening efficiency can be improved. At the same time, screening multiple target regions can divide a long and complex thread into multiple threads, and each thread runs independently, which can effectively improve the utilization rate of the CPU of the computer system and enhance the efficiency of image screening.

[0033] In addition, the embodiments of the present invention screen the first remote sensing images based on the segmented areas, only retaining the first remote sensing images that contain the most segmented areas and solely contain segmented areas, so as to eliminate redundant remote sensing images in the first remote sensing images, and at the same time ensure that the retained first remote sensing images can completely cover the target region, ensuring the accuracy of the screening process.

[0034] Refer to Figure 3 , which shows a multi-threaded sub-region screening flowchart provided by the present invention. The region of interest is divided into 4 target regions, and at the same time, the target initial remote sensing images corresponding to the region of interest are divided to obtain the first remote sensing images corresponding to each target region. Then, using multi-threading, S102 - S103 are respectively executed on the 4 target regions to obtain the second remote sensing images corresponding to the 4 target regions. There may be duplicates among the second remote sensing images of different target regions. Therefore, finally, duplicate removal is performed on all the second remote sensing images of each target region to obtain the third remote sensing image corresponding to the region of interest.

[0035] In one embodiment, S103 may specifically include: determining a two-dimensional array A; wherein the number of rows of A is the number of first remote sensing images, the number of columns of A is the number of segmented regions, and the elements corresponding to the first remote sensing images and the segmented regions in A that have a containment relationship are assigned a value of 1, otherwise they are assigned a value of 0; and determining the second remote sensing image by the following formula:

[0036] Among them, m is the serial number of the first remote sensing image, and n is the number of segmented regions; Will satisfy the condition The first remote sensing image is used as the second remote sensing image.

[0037] This embodiment mainly uses the greedy algorithm to search the first remote sensing image to find the second remote sensing image that meets the requirements from the first remote sensing image. The greedy algorithm needs to define the evaluation function and the heuristic function Two functions. Evaluation function Describes how to select subsequent nodes based on the evaluation function starting from the current node. This evaluation function determines how to select nodes. Heuristic function It describes the minimum cost of the path from the computing node to the target node.

[0038] In this embodiment, the evaluation function It is used to determine whether a first remote sensing image satisfies a single included segmentation area or contains the most unincluded segmentation areas. The heuristic function It is used to determine the state of the first remote sensing image, that is, the number of unincluded segmented areas contained in the first remote sensing image and the number of single included segmented areas, and record the conditions that meet the conditions. The first remote sensing image, and then update the two-dimensional array , update all elements of a row corresponding to the first remote sensing image that meets the conditions to 0. When judging subsequent nodes, use the updated two-dimensional array When a two-dimensional array When all elements in are 0, the screening operation of the second remote sensing image is completed.

[0039] Reference Figure 4 , showing that the present invention Figure 1 The flowchart of an embodiment of S101 is shown in FIG. 1 , and this embodiment includes the following steps: S401, obtaining a plurality of initial remote sensing images associated with the region of interest and a quality score of each initial remote sensing image.

[0040] The initial remote sensing image may be a remote sensing image associated with the region of interest. For example, the region corresponding to the initial remote sensing image is adjacent to or overlaps with the region of interest.

[0041] The quality scores of each initial remote sensing image can be determined according to the cloud amount, resolution, and imaging time of each initial remote sensing image.

[0042] Specifically, the cloud amount score, resolution score, and imaging time score of each initial remote sensing image can be determined first according to the cloud amount, resolution, and imaging time of each initial remote sensing image.

[0043] Among them, the cloud amount score can be determined by the following formula:

[0044] is the cloud amount score, is the proportion of cloud amount in the initial remote sensing image. The less the proportion of cloud amount in the initial remote sensing image, the larger.

[0045] The resolution score can be determined by the following formula:

[0046] is the resolution score, is the resolution of the initial remote sensing image. The higher the resolution, the larger.

[0047] The imaging time score can be determined by the following formula:

[0048] is the imaging time score, is the time for screening processing, is the imaging time of the initial remote sensing image. The newer the imaging time, the larger.

[0049] After obtaining the cloud amount score, resolution score, and imaging time score of each initial remote sensing image, these three scores can be normalized to map these three scores to the range. The following formula can be used to normalize each score:

[0050] Finally, the quality scores of each initial remote sensing image can be calculated by the following formula:

[0051] is the quality score; They are respectively the cloud cover score, resolution score, and imaging time score of the initial remote sensing image; the cloud cover score is determined according to the cloud cover, the resolution score is determined according to the resolution, and the imaging time score is determined according to the imaging time; They are respectively the weights corresponding to the cloud cover score, resolution score, and imaging time score.

[0052] S402. Determine a target initial remote sensing image from each of the initial remote sensing images according to the quality score of each initial remote sensing image.

[0053] An initial remote sensing image with a high quality score can be preferentially selected as the target initial remote sensing image.

[0054] S403. Determine whether the target initial remote sensing image intersects with the region of interest.

[0055] S404. If the target initial remote sensing image intersects with the region of interest, then retain the target initial remote sensing image, determine the intersection region between the target initial remote sensing image and the region of interest, and remove the intersection region from the region of interest to update the region of interest.

[0056] S405. If the target initial remote sensing image does not intersect with the region of interest, then delete the target initial remote sensing image.

[0057] S406. Return to S402 until the intersection region is equal to the region of interest.

[0058] After returning to S402, an initial remote sensing image that has not been traversed and has a high quality score can be preferentially selected as the target initial remote sensing image. Then determine whether the target initial remote sensing image intersects with the updated region of interest. If it intersects, update the region of interest again. Repeat S402 - S406 until the intersection region is equal to the region of interest.

[0059] When the intersection region is equal to the region of interest, the region formed by the retained target initial remote sensing images can completely cover the region of interest.

[0060] S407. Determine multiple first remote sensing images corresponding to the target region from the retained target initial remote sensing images.

[0061] Based on the quality score and the intersection relationship between the initial remote sensing image and the region of interest, this embodiment screens the initial remote sensing images, and can achieve the following beneficial effects: ① Efficiently screen out most of the redundant images in the initial remote sensing images; ② Make the quality score of the selected target initial remote sensing images high; ③ Make the region formed by the target initial remote sensing images can completely cover the region of interest, ensuring the accuracy of the screening.

[0062] Refer to Figure 5, showing a schematic flow diagram of an initial remote sensing image screening method provided by the present invention. Let the spatial geometric object of the region of interest be an arbitrary polygon , and the spatial geometric object of the initial remote sensing image is denoted as ( , is the total number of initial remote sensing images to be screened, sorted by quality score), the region represents and intersection, that is , the region represents and difference set, that is , during the screening process, after the operation, is used to replace the region of interest , if , then this initial remote sensing image is recorded into the target initial remote sensing image set . When , it means that the region of interest is completely covered by the target initial remote sensing image. At this time, is the full coverage result set of the region of interest.

[0063] In one embodiment, the remote sensing image screening method further includes: based on the second remote sensing images corresponding to each target region, performing duplicate removal processing on the second remote sensing images to obtain the third remote sensing images corresponding to the region of interest.

[0064] In one embodiment, the remote sensing image screening method further includes: determining the fourth remote sensing images from the third remote sensing images; wherein, the regions formed by the fourth remote sensing images cover the region of interest, and the number of the fourth remote sensing images reaches the minimum value.

[0065] To solve the local optimum situation existing in the greedy algorithm screening, this embodiment further optimizes the third remote sensing images. The optimization can be regarded as a typical set covering problem, and its goal is to find a series of sets , the set contains the third remote sensing images, and the regions formed by the third remote sensing images in the set completely cover the region of interest. The third remote sensing images in each set are not completely the same. Finally, it is necessary to find a target set containing the fewest third remote sensing images in a series of sets . The third remote sensing images in the target set are the fourth remote sensing images.

[0066] In one embodiment, the step of determining the fourth remote sensing image from the third remote sensing image may specifically include: determining a target third remote sensing image from the third remote sensing image; determining whether the area formed by the remaining third remote sensing images after removing the target third remote sensing image can cover the region of interest; if the area formed by the remaining third remote sensing images covers the region of interest, removing the target third remote sensing image from the third remote sensing image; if the area formed by the remaining third remote sensing images cannot cover the region of interest, retaining the target third remote sensing image; returning to the step of "determining a target third remote sensing image from the third remote sensing image" until all the third remote sensing images are traversed; and using the remaining third remote sensing images as the fourth remote sensing image.

[0067] Based on the third remote sensing image and the topological relationship between the third remote sensing image and the region of interest, this embodiment optimizes the third remote sensing image, which can improve the quality of the screening result.

[0068] Refer to Figure 6 , which shows a set coverage model diagram provided by the present invention. The set includes three third remote sensing images A, B, and C, and the area formed by the third remote sensing images in the set completely covers the region of interest .

[0069] Refer to Figure 7 , which shows a schematic diagram of the optimization process of the third remote sensing image provided by the present invention. Let the spatial object of the region of interest be an arbitrary polygon , the rd image in the set is , is the set of the remaining third remote sensing images after removing from the set , and the spatial object of is an arbitrary polygon . If , then is deleted from , and the set is updated, that is, . If , then is retained and the set is not updated. Traverse all the third remote sensing images in to complete the optimization operation. After the traversal is completed, the third remote sensing images in are the fourth remote sensing images.

[0070] Refer to Figure 8, which shows a schematic diagram of a complete screening operation process provided by the present invention. First, information such as cloud amount information, resolution information, imaging time information, and geometric information of the initial remote sensing image is obtained. Then, based on this information, the quality score of the initial remote sensing image is determined. Based on the quality score of the initial remote sensing image and the intersection relationship between the initial remote sensing image and the region of interest, the target remote sensing image is determined from the initial remote sensing image to complete the preliminary screening. Then, the region of interest is divided into multiple target regions, and the target remote sensing image is divided to obtain the first remote sensing image corresponding to each target region; the first remote sensing image corresponding to each target region is screened again respectively to obtain the second remote sensing image corresponding to each target region; the second remote sensing images are integrated to obtain the third remote sensing image corresponding to the region of interest to complete the detailed screening. Finally, the third remote sensing image is optimized to obtain the fourth remote sensing image to complete the entire screening process.

[0071] Refer to Figure 9 Figure (a), which shows a schematic diagram of the comparison of the screening efficiency of various screening methods provided by the present invention. The vertical coordinate is the screening time, and the horizontal coordinate is the region of interest, and there are 5 regions of interest. For each region of interest, the screening test is carried out respectively by the above 4 screening methods to obtain the screening time. For each region of interest, the bar columns from left to right correspond to the grid method, the spatial filtering method, the traditional greedy algorithm, and the screening time of the present invention. It can be found from Figure 9 Figure (a) that among the four methods, the screening time of the present invention is significantly lower than that of the traditional greedy algorithm and the grid method, that is, the screening efficiency is significantly higher than that of the traditional greedy algorithm and the grid method.

[0072] Figure 9 Figure (b), which shows a schematic diagram of the comparison of the screening redundancy of various screening methods provided by the present invention. Figure 9 In Figure (b), the vertical coordinate is the screening redundancy, and the horizontal coordinate is the same as that in Figure 9 Figure (a). In terms of indicators, the screening effect of the present invention is also significantly better than that of similar methods. In the region of interest with a small area, the redundancy rate is lower than 350%, and the utilization rate exceeds 5%. In the region of interest with a large area, the redundancy rate is lower than 200%, and the utilization rate exceeds 10%. For the screening result of Inner Mongolia, the redundancy rate is only 115%, and the utilization rate reaches 22%. It shows that the present invention can effectively screen the image data sets of different regions of interest, effectively proves the excellent performance of the method, can be well applied to regions of interest of different sizes, can efficiently obtain high-quality, small-number, and full-coverage image data sets according to the needs of users, can quickly screen a large amount of remote sensing image data within the region, and can effectively improve the efficiency of obtaining images within the region in multiple fields such as land cover type monitoring and land and resources survey, and has strong practicability.

[0073] Refer to Figure 10, showing a schematic structural diagram of an embodiment of the remote sensing image screening device provided by the present invention. The device 11 includes: A first remote sensing image determination module 111, configured to obtain a target area and a plurality of first remote sensing images corresponding to the target area; wherein, the plurality of first remote sensing images respectively overlap with the target area, and the area formed by the plurality of first remote sensing images completely covers the target area; A segmentation module 112, configured to segment the plurality of first remote sensing images according to the overlapping situation between the plurality of first remote sensing images to obtain a plurality of segmented areas; wherein, each of the segmented areas does not overlap with each other; A second remote sensing image determination module 113, configured to determine a second remote sensing image from the plurality of first remote sensing images; wherein, the second remote sensing image solely contains at least one of the segmented areas or contains the most of the segmented areas.

[0074] It should be noted that: the implementation principle or implementation process of the above-mentioned modules can refer to the embodiments of the foregoing remote sensing image screening method, and will not be elaborated here one by one.

[0075] Those skilled in the art can understand that all or part of the processes for implementing the methods of the above embodiments can be completed by instructing relevant hardware through a computer program, and the program can be stored in a computer-readable storage medium. Among them, the computer-readable storage medium is a magnetic disk, an optical disk, a read-only memory, or a random access memory, etc.

[0076] The above is only a preferred specific embodiment of the present invention, but the protection scope of the present invention is not limited thereto. Any changes or substitutions that can be easily thought of by those skilled in the art within the technical scope disclosed by the present invention should be covered by the protection scope of the present invention.

Claims

1. A remote sensing image screening method, characterized in that: include: Acquire a target area and a plurality of first remote sensing images corresponding to the target area; wherein the plurality of first remote sensing images overlap with the target area respectively, and an area formed by the plurality of first remote sensing images completely covers the target area; According to the overlap between the multiple first remote sensing images, the multiple first remote sensing images are segmented to obtain multiple segmented areas; wherein the segmented areas do not overlap with each other; A second remote sensing image is determined from the plurality of first remote sensing images; wherein the second remote sensing image contains only at least one of the segmented regions or contains the most of the segmented regions.

2. The remote sensing image screening method according to claim 1, characterized in that: The determining of the second remote sensing image from the plurality of first remote sensing images comprises: Determine a two-dimensional array A; wherein the number of rows of A is the number of first remote sensing images, the number of columns of A is the number of segmented regions, and the elements corresponding to the first remote sensing image and the segmented regions in A that have a containment relationship are assigned a value of 1, otherwise they are assigned a value of 0; The second remote sensing image is determined by the following formula: Among them, m is the serial number of the first remote sensing image, and n is the number of segmented regions; Will satisfy the condition The first remote sensing image is used as the second remote sensing image.

3. The remote sensing image screening method according to claim 1, characterized in that: The target region includes multiple target regions, and the multiple target regions are obtained by segmenting the region of interest.

4. The remote sensing image screening method according to claim 3, characterized in that: The step of acquiring a plurality of first remote sensing images corresponding to the target area includes: Acquire a plurality of initial remote sensing images associated with the region of interest, and a quality score of each of the initial remote sensing images; Determining a target initial remote sensing image from each of the initial remote sensing images according to the quality scores of each of the initial remote sensing images; Determining whether the target initial remote sensing image intersects with the region of interest; If the target initial remote sensing image intersects with the region of interest, retaining the target initial remote sensing image, determining the intersection area between the target initial remote sensing image and the region of interest, and removing the intersection area in the region of interest to update the region of interest; If the target initial remote sensing image does not intersect with the region of interest, deleting the target initial remote sensing image; Return to the step of "determining a target initial remote sensing image from each of the initial remote sensing images according to the quality scores of each of the initial remote sensing images" until the intersection area is equal to the region of interest; A plurality of first remote sensing images corresponding to the target area are determined from the retained initial remote sensing images of the target.

5. The remote sensing image screening method according to claim 4, characterized in that: The obtaining of a plurality of initial remote sensing images and the quality score of each of the initial remote sensing images comprises: A plurality of initial remote sensing images are acquired, and a quality score of each of the initial remote sensing images is determined according to the cloud cover, resolution, and imaging time of each of the initial remote sensing images.

6. The remote sensing image screening method according to claim 5, characterized in that: Determining the quality score of each of the initial remote sensing images according to the cloud cover, resolution and imaging time of each of the initial remote sensing images includes: The quality score of each of the initial remote sensing images is calculated by the following formula: in, Rate the quality; They are the cloud amount score, resolution score and imaging time score of the initial remote sensing image respectively; the cloud amount score is determined according to the cloud amount, the resolution score is determined according to the resolution, and the imaging time score is determined according to the imaging time; They are the weights corresponding to the cloud cover score, resolution score and imaging time score respectively.

7. The remote sensing image screening method according to claim 3, characterized in that: The method further comprises: Based on the second remote sensing images corresponding to the target areas, the second remote sensing images are deduplicated to obtain third remote sensing images corresponding to the area of ​​interest.

8. The remote sensing image screening method according to claim 7, characterized in that: The method further comprises: A fourth remote sensing image is determined from the third remote sensing image; wherein the area formed by the fourth remote sensing image covers the area of ​​interest, and the number of the fourth remote sensing images reaches a minimum value.

9. The remote sensing image screening method according to claim 8, characterized in that: The step of determining a fourth remote sensing image from the third remote sensing image comprises: Determining a target third remote sensing image from the third remote sensing image; Determining whether the area formed by the remaining third remote sensing image after removing the target third remote sensing image can cover the area of ​​interest; If the area formed by the remaining third remote sensing images covers the area of ​​interest, removing the target third remote sensing image from the third remote sensing images; If the area formed by the remaining third remote sensing images cannot cover the area of ​​interest, retaining the target third remote sensing image; Return to the step of "determining a target third remote sensing image from the third remote sensing image" until the traversal of the third remote sensing image is completed; The remaining third remote sensing image is used as the fourth remote sensing image.

10. A remote sensing image screening device, characterized in that: include: A first remote sensing image determination module is used to obtain a target area and a plurality of first remote sensing images corresponding to the target area; wherein the plurality of first remote sensing images overlap with the target area respectively, and the area formed by the plurality of first remote sensing images completely covers the target area; A segmentation module, used for segmenting the plurality of first remote sensing images according to the overlap between the plurality of first remote sensing images to obtain a plurality of segmented areas; wherein the segmented areas do not overlap with each other; The second remote sensing image determination module is used to determine the second remote sensing image from the plurality of first remote sensing images; wherein the second remote sensing image contains only at least one of the segmented regions or contains the most of the segmented regions.

Images