A multi-frame remote sensing image data effective range automatic calculation method

Abstract

The application discloses a kind of multi-frame remote sensing image data effective range automatic calculation method, comprising: obtaining preset frame remote sensing image, determining metadata and circumscribed rectangle;Each circumscribed rectangle is merged to generate original total circumscribed rectangle and constructs spatial index to determine spatial adjacency graph;Based on pixel, metadata and preset determination rule, determine strong, weak invalid mask, synthesis total mask and morphological denoising;Total mask connected domain is simplified without vectorization and zero buffer repair, to determine strong, weak invalid area;Determine the real invalid area of each frame in combination with spatial adjacency graph;According to each circumscribed rectangle and each real invalid area, determine each local effective data polygon;And in the range of original total circumscribed rectangle, it is merged to generate global effective data area.The application can be widely applied to remote sensing image preprocessing, inlay cutting, unmanned aerial vehicle data cleaning and the like scene.

Description

Technical Field

[0001] This invention relates to the field of remote sensing data preprocessing and GIS spatial analysis technology, and in particular to an automatic calculation method for the effective range of multi-sheet remote sensing image data. Background Technology

[0002] In fields such as aerospace remote sensing, mapping, UAV aerial surveying, and online map base map production, remote sensing imagery is typically stored and processed in the form of massive datasets containing hundreds or thousands of map sheets (Tiles). During the collection, geometric correction, administrative division cropping, or lossy compression (such as JPEG compression), these images inevitably generate a large number of "no-data" regions at the edges or within the image. Common no-data regions include: external black / white borders resulting from irregular cropping or rotation; overlapping waste areas during UAV image sequence stitching; dark gradient noise (not pure black) at the edges caused by lossy compression; and internal holes caused by sensor imaging failures or cloud obstruction.

[0003] Before these image frames are imported into a database, seamlessly mosaicked, or published as a map service, extracting the true "effective data coverage" of each image frame is an indispensable preprocessing step. If invalid areas are not properly stripped, it will lead to serious data quality problems such as black borders obscuring underlying data and obvious gaps at the seams.

[0004] Currently, the main technical bottlenecks in methods for automating the calculation of effective data boundaries in the industry are as follows: The pixel-by-pixel diffusion method based on Flood-fill (region growing) relies on manually specifying initial "seed points". In multi-image stitching scenarios, invalid regions often span multiple images and have complex shapes. Seed points are prone to interruption of diffusion when encountering a single noisy pixel, resulting in missed detections or breaks. Furthermore, it is difficult to achieve automated joint calculation across image sheets.

[0005] Global binarization based on fixed thresholds: This method directly removes black borders by setting a fixed color tolerance. However, due to its lack of spatial topological considerations, when there are real dark features inside the image (such as deep lakes or shadows on the back side of mountains), the thresholding method is very likely to misclassify these real features as invalid areas, resulting in "false perforation of the image interior".

[0006] Machine learning-based land cover classification methods, while possessing strong semantic recognition capabilities, require a large number of training samples, resulting in extremely high computational costs. Furthermore, they often exhibit poor generalization stability when faced with various random edge compression noises, making them unsuitable for underlying basic preprocessing operations.

[0007] It is evident that current methods for automating the calculation of effective data boundaries in the industry have certain limitations, resulting in low efficiency and reliability in the production of remote sensing big data. Summary of the Invention

[0008] The main objective of this invention is to provide an automatic calculation method for the effective range of multi-frame remote sensing image data, aiming to solve the problem that current methods for automatically calculating the effective data boundary have certain limitations, resulting in low efficiency and reliability in the production of remote sensing big data.

[0009] To achieve the above objectives, this invention proposes an automatic calculation method for the effective range of multi-frame remote sensing image data, comprising: Acquire preset segmented remote sensing images, and determine the metadata and bounding rectangle of each preset segmented remote sensing image based on each preset segmented remote sensing image; The bounding rectangles are merged to generate the original total bounding rectangle, and a spatial index is constructed based on the bounding rectangles. The spatial adjacency graph between the preset segmented remote sensing images is determined based on the spatial index. Read the pixels of each preset segmented remote sensing image, and determine strong invalid masks and weak invalid masks according to the corresponding pixels, the metadata and preset judgment rules; An initial total mask is synthesized based on the strong invalid mask and the weak invalid mask, and morphological opening operation is performed on the initial total mask to denoise it, thereby determining the final total mask; Vectorization is performed on each connected region in the total mask. In the vectorization operation, the original cell mesh edges are preserved and no geometric simplification is performed. Zero-buffered topology repair is performed on the generated vector polygons. Then, the strong invalid mask and the weak invalid mask are combined to determine the strong invalid region and the weak invalid region respectively. Based on the strong invalid regions, the weak invalid regions, and the spatial adjacency map, the true invalid regions of each of the preset map-sliced ​​remote sensing images are determined; Based on the bounding rectangle of each preset segmented remote sensing image and each real invalid region, determine the local valid data polygon of each preset segmented remote sensing image; merge each local valid data polygon within the original total bounding rectangle to generate a global valid data area.

[0010] Preferably, the step of determining strong invalid masks and weak invalid masks based on the corresponding pixels, the metadata, and preset judgment rules includes: The preset judgment rules include a first condition, a second condition, a third condition, and a fourth condition; Based on the pixels and metadata of each preset frame remote sensing image, determine whether the pixel value of each pixel satisfies any one of the first condition, the second condition, the third condition, and the fourth condition; When the pixel value of any of the pixels satisfies any one of the first condition, the second condition, and the third condition, the pixel with that pixel value is a strongly invalid pixel, and the strongly invalid mask is constructed based on each of the strongly invalid pixels; When the pixel value of any of the pixels satisfies the fourth condition, the pixel with that pixel value is a weak invalid pixel, and the weak invalid mask is constructed based on each of the weak invalid pixels.

[0011] Preferably, the first condition is that the preset segmented remote sensing image where the pixel is located has an Alpha transparency channel, and the Alpha of the pixel is 0; the second condition is that the pixel is equal to a preset NODATA value; the third condition is that the pixel contains non-numeric values; and the fourth condition is that the maximum gray value of the pixel in all bands is less than or equal to a preset color tolerance.

[0012] Preferably, the step of combining the strong invalidation mask and the weak invalidation mask to determine the strong invalidation region and the weak invalidation region respectively includes: Obtain the fifth and sixth conditions, and determine whether the image framing area covered by the vector polygon meets the fifth or sixth condition; When the image framing region covered by the vector polygon meets the fifth condition, the image framing region covered by the vector polygon is a strongly invalid region. When the image framing region covered by the vector polygon meets the sixth condition, the image framing region covered by the vector polygon is a weakly invalid region.

[0013] Preferably, the fifth condition is that there is any strong invalid pixel in the image segmentation area covered by the vector polygon; the sixth condition is that all pixels in the image segmentation area covered by the vector polygon are weak invalid pixels.

[0014] Preferably, the step of determining the true invalid regions of each of the preset map remote sensing images based on the strong invalid regions, the weak invalid regions, and the spatial adjacency map includes: The strongly invalid region is activated as the first invalid region; Obtain the seventh, eighth, and ninth conditions, and determine whether the weakly invalid region satisfies any one of the seventh, eighth, and ninth conditions; When the weak invalid region satisfies any one of the seventh, eighth, and ninth conditions, the weak invalid region that satisfies any one of the seventh, eighth, and ninth conditions will be activated as a second invalid region. Obtain the tenth condition and determine whether the weakly invalid region satisfies the spatial adjacency graph and the tenth condition; When the weak invalid region satisfies the spatial adjacency graph and the tenth condition, the weak invalid region that satisfies the spatial adjacency graph and the tenth condition will be activated as the third invalid region. The first invalid region, the second invalid region, and the third invalid region are merged into the true invalid regions of each of the preset frame remote sensing images.

[0015] Preferably, the seventh condition is that the geometric boundary of the weakly invalid region intersects with the circumscribed rectangle of the preset segmented remote sensing image covered by the weakly invalid region; the eighth condition is that the geometric boundary of the weakly invalid region intersects with the original total circumscribed rectangle or the distance is less than the minimum tolerance; and the ninth condition is that an external instruction parameter for forcibly clearing internal island regions is received.

[0016] Preferably, the tenth condition is that, based on the spatial adjacency graph, the geometric boundary of the weakly invalid region intersects with the geometric boundary of any of the first invalid regions, or the geometric boundary of the weakly invalid region intersects with the geometric boundary of any of the second invalid regions.

[0017] Preferably, the step of determining the local valid data polygons of each preset remote sensing image based on the bounding rectangle of each preset image segment and each real invalid region, and merging the local valid data polygons within the original total bounding rectangle to generate a global valid data area includes: Each real invalid region within each preset segmented remote sensing image is locally geometrically fused to determine the total local invalid region corresponding to each preset segmented remote sensing image. Subtract the corresponding local total invalid region from the bounding rectangle of each preset segmented remote sensing image to determine the local valid data polygon of each preset segmented remote sensing image; The local valid data polygons are merged within the original total bounding rectangle to generate a global valid data area.

[0018] Compared with the prior art, the present invention has at least the following beneficial effects: Based on vector polygon analysis and a strong / weak invalid pixel hierarchical identification model, this method first constructs a strong invalid mask reflecting absolutely invalid data by analyzing pre-defined remote sensing images, and then constructs a weak invalid mask reflecting suspected invalid data by combining color tolerance. Next, morphological filtering is applied to the synthesized total mask to denoise it, and strong and weak invalid regions are selected as candidate invalid regions. These are then extracted using vector polygon analysis, with pixel-level boundaries strictly preserved during vectorization to ensure topological rigor. Subsequently, an intelligent hybrid activation mechanism based on map edge contact, global outer contour contact, and spatial adjacency graph propagation is used to determine and activate true invalid regions. Finally, the activated invalid regions are uniformly merged, and the final valid data boundary (i.e., the global valid data area) with strict topological closure is extracted by performing a difference operation with the original total bounding rectangle.

[0019] It can automatically and accurately remove black borders, compressed noise, and internal NODATA areas in multi-map stitching, while perfectly preserving real dark features such as lakes and shadows. It can be widely used in remote sensing image preprocessing, mosaicking and cropping, base map production, and UAV data cleaning. Attached Figure Description

[0020] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on the structures shown in these drawings without creative effort.

[0021] Figure 1 This is a flowchart illustrating an embodiment of a method for automatically calculating the effective range of multi-frame remote sensing image data.

[0022] The realization of the objective, functional features and advantages of the present invention will be further explained in conjunction with the embodiments and with reference to the accompanying drawings. Detailed Implementation

[0023] Embodiments of the present invention are described in detail below, examples of which are illustrated in the accompanying drawings. The embodiments described below with reference to the accompanying drawings are exemplary and intended to explain the present invention, and should not be construed as limiting the present invention.

[0024] The following describes an embodiment of the present invention with reference to the accompanying drawings, an automatic calculation method for the effective range of multi-frame remote sensing image data.

[0025] Figure 1 This is a flowchart illustrating an embodiment of a method for automatically calculating the effective range of multi-frame remote sensing image data.

[0026] Please see Figure 1To achieve the above objectives, the first embodiment of the present invention provides an automatic calculation method for the effective range of multi-frame remote sensing image data, comprising: Step S10: Obtain preset segmented remote sensing images, and determine the metadata and bounding rectangle of each preset segmented remote sensing image based on each preset segmented remote sensing image. Step S20: Merge the bounding rectangles to generate the original total bounding rectangle, construct a spatial index based on the bounding rectangles, and determine the spatial adjacency map between the preset sheet remote sensing images based on the spatial index. Step S30: Read the pixels of each preset frame remote sensing image, and determine strong invalid masks and weak invalid masks according to the corresponding pixels, metadata and preset judgment rules; Step S40: Synthesize an initial total mask based on the strong invalid mask and the weak invalid mask, and perform morphological opening operation on the initial total mask to denoise it, and determine the final total mask; Step S50: Perform vectorization operation on each connected region in the total mask. In the vectorization operation, the original cell mesh edges are preserved and geometric simplification is not performed. Perform zero-buffered topology repair operation on the generated vector polygons. Then, combine the strong invalid mask and the weak invalid mask to determine the strong invalid region and the weak invalid region respectively. Step S60: Determine the real invalid regions of each preset map remote sensing image based on the strong invalid regions, weak invalid regions, and spatial adjacency map; Step S70: Based on the bounding rectangle of each preset segmented remote sensing image and each real invalid region, determine the local valid data polygon of each preset segmented remote sensing image; merge each local valid data polygon within the original total bounding rectangle to generate a global valid data area.

[0027] It can automatically and accurately remove black borders, compressed noise, and internal NODATA areas in multi-map stitching, while perfectly preserving real dark features such as lakes and shadows. It can be widely used in remote sensing image preprocessing, mosaicking and cropping, base map production, and UAV data cleaning.

[0028] Steps S10 and S20 extract the bounding rectangles of all preset sheet remote sensing images and construct a spatial index. The spatial indexing mechanism maps the two-dimensional physical envelopes of each preset sheet remote sensing image into a tree-like hierarchical data structure, enabling rapid spatial location retrieval. Subsequently, a spatial adjacency graph (G=(V, E)) containing all preset sheet remote sensing images is built based on this spatial index. Simultaneously, the computer quickly extracts peripheral nodes based on the spatial index and geometrically merges the bounding rectangles of each preset sheet remote sensing image to generate the original total bounding rectangle of the dataset.

[0029] In the second embodiment of the present invention, based on the first embodiment, the step of determining strong invalid masks and weak invalid masks according to the corresponding pixels, metadata, and preset judgment rules includes: Step S31, the preset judgment rules include the first condition, the second condition, the third condition and the fourth condition; Step S32: Based on the pixel and metadata of each preset sheet remote sensing image, determine whether the pixel value of each pixel satisfies any one of the first condition, the second condition, the third condition, and the fourth condition. Step S33: When the pixel value of any pixel satisfies any one of the first condition, the second condition, and the third condition, the pixel with that pixel value is a strongly invalid pixel, and a strongly invalid mask is constructed based on each strongly invalid pixel. Step S34: When the pixel value of any pixel satisfies the fourth condition, the pixel with that pixel value is a weak invalid pixel, and a weak invalid mask is constructed based on each weak invalid pixel.

[0030] Step S30, in conjunction with step S60, forms a strong / weak invalidity classification and activation strategy, which significantly preserves internal real features (lakes, shadows, dark features). Weak invalid cells are not automatically deleted; they are only activated when they touch the outer boundary. Therefore, internal dark features such as "black lakes" or "shadow areas" will not be mistakenly deleted. This is a core problem that traditional flood-fill or fixed threshold methods cannot solve.

[0031] To accurately distinguish between "real image edges / holes" and "dark features within the image," a hierarchical mask is constructed for each preset frame remote sensing image (i.e., strong invalid masks and weak invalid masks are constructed). Pixels included in the strong invalid mask are considered invalid pixels.

[0032] In the third embodiment of the present invention, based on the second embodiment, the first condition is that the preset map remote sensing image where each pixel is located in the metadata has an Alpha transparent channel and the Alpha of the pixel is 0; the second condition is that the pixel is equal to the preset NODATA value; the third condition is that the pixel contains non-numeric values; and the fourth condition is that the maximum gray value of the pixel in all bands is less than or equal to the preset color tolerance.

[0033] The NODATA value is the absolute matching threshold constructed using the "strong invalid mask".

[0034] To address the issue of background pixel fluctuations caused by lossy compression or resampling, a comprehensive verification of the multi-band characteristics of pixels is performed based on a preset color tolerance T. The expression for weak and invalid pixels is max(Band_1, Band_2, Band_3) ≤ T, where Band represents the frequency band.

[0035] Weak invalid masks are mainly used to capture black border noise with numerical fluctuations caused by image cropping and lossy compression, but they may also mistakenly capture real dark features (such as lakes).

[0036] The first to third conditions, combined with step S60, form a strategy for weak invalid masking, morphological denoising, and boundary contact judgment, significantly improving the ability to identify black edges and compressed noise. The weak invalid mask combined with morphological filtering can effectively capture and purify black edge noise; through the edge activation mechanism, black edge removal is more stable, accurate, and has strong anti-interference ability.

[0037] In the fourth embodiment of the present invention, based on the second embodiment, and combining strong invalidation masks and weak invalidation masks, the step of determining strong invalidation regions and weak invalidation regions respectively includes: Step S51: Obtain the fifth and sixth conditions, and determine whether the image framing area covered by the vector polygon meets the fifth or sixth conditions. Step S52: When the image framing area covered by the vector polygon meets the fifth condition, the image framing area covered by the vector polygon is a strongly invalid area. Step S53: When the image framing area covered by the vector polygon meets the sixth condition, the image framing area covered by the vector polygon is a weakly invalid area.

[0038] The total mask is synthesized using a pixel-by-pixel Boolean OR operation; The initial master mask is calculated as either a strong invalid mask or a weak invalid mask. A morphological opening operation (erosion followed by dilation) is then performed on the master mask. This operation aims to filter out isolated "salt-and-pepper noise" generated by lossy compression in the image, preventing these tiny noise points from being incorrectly extracted as a large number of fragmented invalid polygons in subsequent steps, ultimately generating a cleaned master mask.

[0039] The image processing algorithm is invoked to extract all physically connected strong and weak invalid regions from the purified overall mask.

[0040] Steps S51 to S53 form the CCL full-image connected component extraction, and perform vectorization operation on each connected component. No seed points are needed, and the recognition result is completely independent of the image size and arrangement; eliminate the "geometric gaps" at the stitching of multiple images to achieve pixel-level seamless stitching.

[0041] By prohibiting geometric simplification during vectorization, the generated boundaries are ensured to strictly follow the pixel grid. This solves the problem of tiny triangular gaps (Slivers) appearing at the seams between adjacent images due to boundary smoothing in existing techniques.

[0042] CCL directly obtains all candidate regions, which is a deterministic global algorithm that avoids the risk of missed detection caused by the reliance on seed point positions and scanning order in traditional Flood-fill.

[0043] Vectorization is performed on each connected component. During this process, no geometric simplification algorithms are executed, and the jagged edges of the pixel grid are strictly preserved to ensure that the generated polygons perfectly match the original image pixels. Simultaneously, a "zero-buffering" operation is performed on the generated vector polygons to correct minor topological errors such as self-intersections that may occur during polygon generation.

[0044] In the fifth embodiment of the present invention, based on the fourth embodiment, the fifth condition is that there is any strong invalid pixel in the image segmentation area covered by the vector polygon; the sixth condition is that all pixels in the image segmentation area covered by the vector polygon are weak invalid pixels.

[0045] Specifically, all weakly invalid regions are grouped into a candidate invalid region set: {poly_i, is_strict_i}; Wherein, poly_i represents the geometric carrier of the object: it records the absolute coordinates and boundary contour of the i-th vector polygon in the real geographic space, which is used to perform all spatial topology and geometric intersection operations in subsequent steps; is_strict_i represents the judgment attribute of the object: it is a boolean value. When the value is True, it means that the area within the polygon is an absolutely invalid area. When the value is False, it means that the polygon belongs to suspected noise or dark features, and topological conditions need to be added for further judgment.

[0046] In the sixth embodiment of the present invention, based on the fourth embodiment, step S60 includes: Step S61: Activate the strongly invalid region as the first invalid region; Step S62: Obtain the seventh, eighth, and ninth conditions, and determine whether the weak invalid region satisfies any one of the seventh, eighth, and ninth conditions; Step S63: When a weak invalid region satisfies any one of the seventh, eighth, and ninth conditions, the weak invalid region that satisfies any one of the seventh, eighth, and ninth conditions is activated as a second invalid region. Step S64: Obtain the tenth condition and determine whether the weak invalid region satisfies the spatial adjacency graph and the tenth condition; Step S65: When a weak invalid region satisfies the spatial adjacency graph and the tenth condition, the weak invalid region that satisfies the spatial adjacency graph and the tenth condition is activated as the third invalid region. Step S66: Merge the first invalid region, the second invalid region, and the third invalid region into the real invalid regions of each preset frame remote sensing image.

[0047] To address the issue of accidental deletion of dark-colored features, the computer executes intelligent activation logic for polygonal regions in each strongly invalid region and each weakly invalid region.

[0048] If a polygonal region is classified as a strongly invalid region, it is immediately marked as the first "activated" invalid region. This mechanism ensures that genuine NODATA holes within the image caused by cloud cover or data acquisition failure are unconditionally considered invalid regions.

[0049] In the seventh embodiment of the present invention, based on the sixth embodiment, the seventh condition is that the geometric boundary of the weak invalid region intersects with the outer rectangle of the preset segmented remote sensing image covered by the weak invalid region; the eighth condition is that the geometric boundary of the weak invalid region intersects with the original total outer rectangle or the distance is less than the minimum tolerance; the ninth condition is that an external instruction parameter for forcibly clearing internal island regions is received.

[0050] For polygon regions with the attribute of weak invalid region, spatial topology determination is performed, and they are activated only if any one of the seventh, eighth and ninth conditions is met.

[0051] In the eighth embodiment of the present invention, based on the sixth embodiment, the tenth condition is that, based on the spatial adjacency graph, the geometric boundary of the weak invalid region intersects with the geometric boundary of any first invalid region, or the geometric boundary of the weak invalid region intersects with the geometric boundary of any second invalid region.

[0052] Weakly invalid regions at the edges often span multiple map sheets. Based on the spatial adjacency graph constructed in step S20, an iterative check is performed: for a currently inactive weakly invalid region (i.e., poly_i), it is determined whether it has a geometric intersection with any "activated" first or second invalid region in any adjacent map sheet. If there is an intersection, it means that the first or second invalid region is physically connected to the outside, and its state is then updated to "activated". This adjacency propagation iteration is continued until the states of all polygons in all weakly invalid regions converge, that is, no new weakly invalid regions are activated.

[0053] The tenth condition is the adjacency propagation mechanism. In the case of multi-map splicing, the continuity of invalid area identification is extremely strong. It automatically judges the cross-map topological contact and can stably and seamlessly identify the entire invalid area band. Even if the area does not touch the outer contour in a certain map segment (such as a "peninsula" distribution), it can still be correctly identified through the state propagation of neighbors.

[0054] In the ninth embodiment of the present invention, based on the first embodiment, step S70 includes: Step S71: Perform local geometric fusion on each real invalid region within each preset sheet remote sensing image to determine the local total invalid region corresponding to each preset sheet remote sensing image. Step S72: Subtract the corresponding local total invalid region from the bounding rectangle of each preset sheet remote sensing image to determine the local valid data polygon of each preset sheet remote sensing image. Step S73: Merge the polygons of each local valid data within the original total bounding rectangle to generate a global valid data area.

[0055] All "activated" first, second, and third invalid regions are subjected to polygon geometric fusion to obtain locally valid data polygons for the map sheet. Finally, a polygon merging operation is performed across the map sheet on all locally valid data polygons to generate the final global valid data area.

[0056] The final merged geometry (i.e., the entire effective data area) is output and saved. The output is a vector polygon describing the effective data spatial range. This geometry has strict topological closure and seamless stitching characteristics, and can be directly used for subsequent remote sensing image mosaicking, visualization, or precise cropping operations.

[0057] In the description of this specification, references are made to the terms "one embodiment", "another embodiment", and "other embodiments". Descriptions such as "example" or "first embodiment to Xth embodiment" refer to descriptions in conjunction with that embodiment or example. Specific features, structures, materials, or characteristics are included in at least one embodiment or example of the present invention.

[0058] In this specification, the illustrative expressions of the terms used do not necessarily refer to the same embodiments or examples.

[0059] Furthermore, the specific features, structures, materials, method steps, or characteristics described may be combined in any suitable manner in one or more embodiments or examples.

[0060] It should be noted that, in this document, the terms "comprising," "including," or any other variations thereof are intended to... This encompasses non-exclusivity inclusion, thereby allowing a process, method, article, or device to include a range of elements. The setting includes not only those elements, but also other elements not explicitly listed, or may also include... Elements inherent to such a process, method, article, or apparatus. Without further limitation, an element defined by the phrase "comprising a..." does not exclude the presence of other identical elements in the process, method, article, or apparatus that includes that element.

[0061] The sequence numbers of the above embodiments of the present invention are merely for descriptive purposes and do not represent the superiority or inferiority of the embodiments. Through the above description of the embodiments, those skilled in the art can clearly understand that the methods of the above embodiments can be implemented by means of software plus necessary general-purpose hardware platforms. Of course, they can also be implemented by hardware, but in many cases, the former is a better implementation method. Based on this understanding, the technical solution of the present invention, in essence, or the part that contributes to the prior art, can be embodied in the form of a software product. This computer software product is stored in a storage medium (such as ROM / RAM, magnetic disk, optical disk), and includes several instructions to cause a terminal (which may be a mobile phone, computer, server, air conditioner, or network device, etc.) to execute the methods described in the various embodiments of the present invention.

[0062] The embodiments of the present invention have been described above with reference to the accompanying drawings, but the present invention is not limited thereto. The specific embodiments described above are merely illustrative and not restrictive. Those skilled in the art, guided by the teachings of this invention, will apply the principles and claims of this invention without departing from its spirit and scope. Within the scope of protection, many other forms can be made, all of which fall within the protection scope of this invention.

Claims

1. A method for automatically calculating the effective range of multi-frame remote sensing image data, characterized in that, include: Acquire preset segmented remote sensing images, and determine the metadata and bounding rectangle of each preset segmented remote sensing image based on each preset segmented remote sensing image; The bounding rectangles are merged to generate the original total bounding rectangle, and a spatial index is constructed based on the bounding rectangles. The spatial adjacency graph between the preset segmented remote sensing images is determined based on the spatial index. Read the pixels of each preset segmented remote sensing image, and determine strong invalid masks and weak invalid masks according to the corresponding pixels, the metadata and preset judgment rules; An initial total mask is synthesized based on the strong invalid mask and the weak invalid mask, and morphological opening operation is performed on the initial total mask to denoise it, thereby determining the final total mask; Vectorization is performed on each connected region in the total mask. In the vectorization operation, the original cell mesh edges are preserved and no geometric simplification is performed. Zero-buffered topology repair is performed on the generated vector polygons. Then, the strong invalid mask and the weak invalid mask are combined to determine the strong invalid region and the weak invalid region respectively. Based on the strong invalid regions, the weak invalid regions, and the spatial adjacency map, the true invalid regions of each of the preset map-sliced ​​remote sensing images are determined; Based on the bounding rectangle of each preset segmented remote sensing image and each real invalid region, determine the local valid data polygon of each preset segmented remote sensing image; merge each local valid data polygon within the original total bounding rectangle to generate a global valid data area.

2. The method for automatically calculating the effective range of multi-frame remote sensing image data as described in claim 1, characterized in that, The step of determining strong invalid masks and weak invalid masks based on the corresponding pixels, the metadata, and preset judgment rules includes: The preset judgment rules include a first condition, a second condition, a third condition, and a fourth condition; Based on the pixels and metadata of each preset frame remote sensing image, determine whether the pixel value of each pixel satisfies any one of the first condition, the second condition, the third condition, and the fourth condition; When the pixel value of any of the pixels satisfies any one of the first condition, the second condition, and the third condition, the pixel with that pixel value is a strongly invalid pixel, and the strongly invalid mask is constructed based on each of the strongly invalid pixels; When the pixel value of any of the pixels satisfies the fourth condition, the pixel with that pixel value is a weak invalid pixel, and the weak invalid mask is constructed based on each of the weak invalid pixels.

3. The method for automatically calculating the effective range of multi-frame remote sensing image data as described in claim 2, characterized in that, The first condition is that the preset segmented remote sensing image where the pixel is located has an Alpha transparency channel, and the Alpha of the pixel is 0; the second condition is that the pixel value of the pixel is equal to the preset NODATA value; the third condition is that the pixel value of the pixel contains non-numeric values; the fourth condition is that the maximum gray value of the pixel value in all bands is less than or equal to the preset color tolerance.

4. The method for automatically calculating the effective range of multi-frame remote sensing image data as described in claim 2, characterized in that, The step of combining the strong invalid mask and the weak invalid mask to determine the strong invalid region and the weak invalid region respectively includes: Obtain the fifth and sixth conditions, and determine whether the image framing area covered by the vector polygon meets the fifth or sixth condition; When the image framing region covered by the vector polygon meets the fifth condition, the image framing region covered by the vector polygon is a strongly invalid region. When the image framing region covered by the vector polygon meets the sixth condition, the image framing region covered by the vector polygon is a weakly invalid region.

5. The method for automatically calculating the effective range of multi-frame remote sensing image data as described in claim 4, characterized in that, The fifth condition is that there is any strong invalid pixel in the image segmentation area covered by the vector polygon; the sixth condition is that all pixels in the image segmentation area covered by the vector polygon are weak invalid pixels.

6. The method for automatically calculating the effective range of multi-frame remote sensing image data as described in claim 4, characterized in that, The step of determining the true invalid regions of each preset map segmented remote sensing image based on the strong invalid regions, the weak invalid regions, and the spatial adjacency map includes: The strongly invalid region is activated as the first invalid region; Obtain the seventh, eighth, and ninth conditions, and determine whether the weakly invalid region satisfies any one of the seventh, eighth, and ninth conditions; When the weak invalid region satisfies any one of the seventh, eighth, and ninth conditions, the weak invalid region that satisfies any one of the seventh, eighth, and ninth conditions will be activated as a second invalid region. Obtain the tenth condition and determine whether the weakly invalid region satisfies the spatial adjacency graph and the tenth condition; When the weak invalid region satisfies the spatial adjacency graph and the tenth condition, the weak invalid region that satisfies the spatial adjacency graph and the tenth condition will be activated as the third invalid region. The first invalid region, the second invalid region, and the third invalid region are merged into the true invalid regions of each of the preset frame remote sensing images.

7. The method for automatically calculating the effective range of multi-frame remote sensing image data as described in claim 6, characterized in that, The seventh condition is that the geometric boundary of the weakly invalid region intersects with the outer rectangle of the preset segmented remote sensing image covered by the weakly invalid region; the eighth condition is that the geometric boundary of the weakly invalid region intersects with the original total outer rectangle or the distance is less than the minimum tolerance; the ninth condition is that an external instruction parameter for forcibly clearing internal island regions is received.

8. The method for automatically calculating the effective range of multi-frame remote sensing image data as described in claim 6, characterized in that, The tenth condition is that, based on the spatial adjacency graph, the geometric boundary of the weak invalid region intersects with the geometric boundary of any of the first invalid regions, or the geometric boundary of the weak invalid region intersects with the geometric boundary of any of the second invalid regions.

9. The method for automatically calculating the effective range of multi-frame remote sensing image data as described in claim 1, characterized in that, The step of determining the local valid data polygons of each preset remote sensing image based on the bounding rectangle of each preset image and each real invalid region, and merging the local valid data polygons within the original total bounding rectangle to generate a global valid data area includes: Each real invalid region within each preset segmented remote sensing image is locally geometrically fused to determine the total local invalid region corresponding to each preset segmented remote sensing image. Subtract the corresponding local total invalid region from the bounding rectangle of each preset segmented remote sensing image to determine the local valid data polygon of each preset segmented remote sensing image; The local valid data polygons are merged within the original total bounding rectangle to generate a global valid data area.

Images