Unmanned aerial vehicle surveying and mapping method based on landform surveying and mapping precision

Abstract

The invention discloses an unmanned aerial vehicle surveying and mapping method based on landform surveying and mapping precision, relates to the technical field of landform surveying and mapping, and solves the problem that surveying and mapping route planning of an unmanned aerial vehicle is unreasonable due to the fact that a specified landform area is not known. Radiation deviation caused by atmospheric absorption, scattering and sensor characteristics is greatly eliminated; in a pixel feature extraction link, point location pixel values and information of eight pixel points around the point location pixel values are creatively and comprehensively considered, and it is ensured that point locations with the same feature can be accurately positioned in multiple sets of images; a series of subsequent fine operations based on feature point locations, including mean processing of connecting line lengths and image plane stretching and zooming, enable geometric differences among different images to be corrected properly, finally realize high-precision fusion of multiple groups of images, and enable the generated fused image to reflect the actual situation of a geomorphic area to be surveyed more truly and comprehensively.

Description

Technical Field

[0001] The present invention relates to the technical field of topographic surveying and mapping, and in particular to an unmanned aerial vehicle surveying and mapping method based on topographic and geomorphic surveying and mapping accuracy. Background Art

[0002] Topographic surveying and mapping is a vital basic geographic information acquisition work. It is like drawing a detailed "blueprint" for the earth's surface, and plays an indispensable supporting role in the development of many fields. From a professional perspective, topographic surveying and mapping covers a wide variety of technical means and processes. The first is the data collection link, among which satellite remote sensing technology is unique. Various high-resolution satellites are like "eyes in space", continuously sending back massive remote sensing images to the ground. These images not only cover a wide range, allowing a panoramic view of large areas, but also use advanced sensors to capture information about land objects in different bands, just like covering the surface of the earth with a colorful "information coat".

[0003] For specific surveying of designated terrain areas, relevant operators usually control the UAVs to conduct related surveying of designated terrain areas and confirm the surveying images to improve the specific surveying accuracy of designated terrain areas. In the actual surveying process, due to the lack of understanding of the designated terrain areas, the surveying route planning of the UAVs will be unreasonable. In addition, when the UAVs are changing in height, if the height difference is large, it will cause great problems in the surveying accuracy of the UAVs during surveying. Summary of the invention

[0004] In view of the shortcomings of the prior art, the present invention provides a UAV surveying and mapping method based on topographic and geomorphic surveying and mapping accuracy, which solves the problem of unreasonable UAV surveying and mapping route planning caused by failure to understand the designated terrain area.

[0005] To achieve the above objectives, the present invention is implemented through the following technical solutions: A UAV surveying and mapping method based on topographic and geomorphic surveying and mapping accuracy comprises the following steps: Step 1: Obtain remote sensing images of the landform area to be surveyed. There are multiple groups of remote sensing images. Confirm the pixel features associated with different points in the multiple groups of remote sensing images. Select points with the same features from the multiple groups of remote sensing images based on the confirmed pixel features. Then, based on the points with the same features determined in the multiple groups of remote sensing images, fuse the multiple groups of remote sensing images to confirm the fused image belonging to the landform area to be surveyed. The specific method is as follows: S11, from the multiple groups of remote sensing images associated with the landform area to be surveyed, confirm the pixel features of the corresponding points in a single group of remote sensing images: confirm the point pixel value of the corresponding point and mark it as X i, where i represents the corresponding point in the remote sensing image, and the pixel value of the surrounding pixels of this point is calibrated as H i-q , where q=1, 2, ..., 8; Based on the confirmed point pixel value X i And the pixel values ​​H of the surrounding pixels i-q , using: Cz i-q =(X i -H i-q ) Confirm the pixel value difference Cz between this pixel and the surrounding pixels i-q , and then the different pixel value differences Cz associated with different surrounding pixels i-q Perform ratio processing, confirm the ratio sequence, use this ratio sequence as the pixel feature of the pixel point, and confirm and record the different pixel features associated with different points in the remote sensing image in turn; S12, based on the pixel features associated with the corresponding points in the corresponding remote sensing images, identifying points with the same pixel features from multiple groups of remote sensing images, marking the identified points as feature points, confirming the pixel values ​​of the feature points in different remote sensing images, and performing mean processing on the confirmed groups of pixel values ​​about the feature points to determine the mean pixel, and using the determined mean pixel as the feature pixel of the feature point; S13, based on the several groups of feature points confirmed in the single group of remote sensing images, confirm the connecting lines between the corresponding adjacent feature points, and record the line length L of the connecting lines, and then perform mean processing on the line length L of the several groups of connecting lines confirmed from the multiple groups of remote sensing images, and confirm the feature mean value belonging to the adjacent feature points, and the feature points associated with the same connecting lines are all the same; Based on the feature mean values ​​determined between the corresponding adjacent feature points, multiple groups of remote sensing images are plane stretched or scaled, and the process stops when the length of the connecting line associated with the corresponding adjacent feature points in each remote sensing image is consistent with the determined feature mean value, and the remote sensing image after the plane adjustment is calibrated as the image to be fused; S14, fusing multiple groups of images to be fused, aligning the same feature points in each image to be fused based on the determined feature points, and fusing the images to confirm the fused image, and adjusting the pixel values ​​associated with the corresponding feature points after fusion to the determined feature pixels; Step 2: Based on the confirmed fused image, generate a mountain model about the fused image, and perform preliminary processing on the generated mountain model to decompose the mountain model into several grid areas. The specific sub-steps are: S21, generating a set of grid surfaces, wherein the grid surfaces contain a plurality of grids, and the area of ​​each grid is consistent, and the area of ​​the corresponding grid is a preset value, and then determining the bottom surface of the mountain model, wherein the bottom surface is parallel to the grid surface, and the grid surface is overlapped with the bottom surface of the mountain model; S22, based on the determined grid surface, the corresponding grid surface is moved upward, and during the movement, it is confirmed that the outer surface of the mountain model belongs to a partial area of ​​the grid surface, and the confirmed partial area is marked as a grid area, so that the outer surface of the mountain model is decomposed into a plurality of grid areas; Step 3: Based on several groups of grid areas confirmed on the upper surface of the mountain model, the grid features of the grid areas are determined, and then based on the determined grid features of different grid areas, several groups of grid areas are sorted in association to select the best grid sequence. The specific sub-steps are as follows: S31, based on the bottom surface determined by the mountain model, confirm the relevant distance J from different points in the corresponding grid area to the bottom surface k , where k represents different points in the grid area, and the corresponding groups of related distances J confirmed in the grid area k Perform mean processing to confirm the grid feature TZ belonging to this grid area; S32, randomly selecting a group of grid areas from the determined groups of edge grid areas as the starting area, and starting from the starting area, randomly selecting other grid areas as subsequent grid areas, and randomly combining them, until all the grid areas are combined to confirm the combination sequence, confirm the grid feature difference between adjacent grid areas in the combination sequence, and the grid feature difference = |grid feature of the previous group of grid areas - grid feature of the next group of grid areas|, and summing the grid feature differences of the groups associated with the combination sequence to confirm the combination feature corresponding to the combination sequence; Then confirm the combination features of other combination sequences one by one; Then randomly select other grid areas as the starting areas, and confirm the subsequent combination sequences, and simultaneously confirm the combination features associated with the corresponding groups of combination sequences; S33, selecting a minimum value from the combination features associated with the plurality of groups of combination sequences, and taking the combination sequence associated with the minimum value as the optimal grid sequence; Step 4: Based on the selected optimal grid sequence, confirm the feature points of the corresponding grid area in the optimal grid sequence, then connect the confirmed groups of feature points, confirm the surveying and mapping route associated with the optimal grid sequence, and control the drone to conduct field surveying with the determined optimal grid sequence and surveying and mapping route; the specific sub-steps for determining the surveying and mapping route are: S41, starting from the first grid area of ​​the best grid sequence and gradually performing correlation processing: based on the grid features confirmed in the corresponding grid area and the different correlation distances J associated with different points in the grid area k , confirm the center point of the corresponding grid area, mark it as the grid midpoint, take the grid midpoint as the center, construct a set of radiating circles with a radius of R1, where R1 is a preset value, mark the points inside this radiating circle as the points to be selected, and perform difference processing on the relevant distance of the points to be selected and the grid features of the corresponding grid area, confirm the relevant difference, and its relevant difference = |J k -Grid feature|, select the minimum value from the confirmed groups of related differences, and mark the point to be selected associated with the minimum value as the surveying and mapping point; S42, confirming the surveying and mapping points for each corresponding grid area, connecting several surveying and mapping points confirmed in sequence, confirming the surveying and mapping route, and controlling the UAV based on the confirmed surveying and mapping route to enable the UAV to perform field surveying and mapping.

[0006] The present invention provides a UAV mapping method based on topographic mapping accuracy. Compared with the prior art, it has the following beneficial effects: The present invention processes multiple groups (at least three groups) of remote sensing images in detail, firstly performs radiation correction, thereby greatly eliminating the radiation deviation caused by atmospheric absorption, scattering and sensor characteristics, restoring the real radiation situation of the ground objects, and laying the foundation for subsequent accurate analysis; in the pixel feature extraction link, innovatively and comprehensively considers the information of the point pixel value and the eight surrounding pixels, accurately defines the pixel features through the difference ratio sequence, effectively overcomes the interference of different lighting conditions on image interpretation, and ensures that the same feature points can be accurately located in multiple groups of images; a series of subsequent fine operations based on the feature points, including the mean value processing of the connecting line length and the stretching and scaling of the image plane, enable the geometric differences between different images to be properly corrected, and finally achieve high-precision fusion of multiple groups of images, and the generated fused image can more realistically and comprehensively reflect the actual situation of the landform area to be surveyed; Based on the grid area, through a unique grid feature calculation method, the distance between the points in the grid and the bottom surface is comprehensively considered and the average is taken as the key feature, which deeply excavates the intrinsic structural information of the mountain model; in the process of selecting the best grid sequence, the random combination and feature difference sum analysis are innovatively adopted to accurately select the best sequence with a gentle change in height difference from many possible combination sequences; based on the surveying and mapping route determined by this sequence, the concept of grid midpoint and radiation circle is fully utilized to cleverly select surveying and mapping points, so that the drone does not need to frequently and drastically adjust the flight altitude during actual surveying and mapping, but only needs to make small and precise adjustments. This not only greatly reduces the error caused by altitude adjustment and ensures the stability of drone surveying and mapping accuracy, but also significantly improves surveying and mapping efficiency, ensuring that the terrain data of the landform area to be surveyed can be obtained quickly and accurately. BRIEF DESCRIPTION OF THE DRAWINGS

[0007] Figure 1 It is a schematic diagram of the process of the present invention. DETAILED DESCRIPTION

[0008] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0009] See also Figure 1 , this application provides a UAV mapping method based on topographic mapping accuracy, including the following steps: Step 1: Acquire remote sensing images of the landform area to be surveyed. There are multiple groups of remote sensing images, which are input by the operator. Confirm the pixel features associated with different points in the multiple groups of remote sensing images, select the same feature points from the multiple groups of remote sensing images based on the confirmed pixel features, and then fuse the multiple groups of remote sensing images based on the same feature points determined in the multiple groups of remote sensing images to confirm the fused image belonging to the landform area to be surveyed. Specifically, since the remote sensing images are affected by atmospheric absorption, scattering and the characteristics of the sensor itself during the acquisition process, the radiation value of the image is biased; perform radiation correction on the multiple groups of remote sensing images respectively to eliminate these influences so that the brightness value of the image is closer to the actual radiation condition of the ground object. The multiple groups include at least three groups. The overall performance of the corresponding fused image is improved through the specific fusion of the multiple groups of images. Among them, the specific method of determining the fused image of the landform area to be surveyed is: S11, from the multiple groups of remote sensing images associated with the landform area to be surveyed, confirm the pixel features of the corresponding points in a single group of remote sensing images: confirm the point pixel value of the corresponding point and mark it as Xi , where i represents the corresponding point in the remote sensing image, and the pixel value of the surrounding pixels of this point is calibrated as H i-q , where q = 1, 2, ..., 8, there are eight groups of pixels around a group of pixels. In the two-dimensional plane representation of digital images, pixels are the basic units that constitute images, and they are arranged in a discrete grid form; usually we can regard the image as a matrix composed of rows and columns, and each element (i.e., pixel) has a specific position in this matrix. Therefore, in the associated remote sensing image, the number of pixels around a group of pixels is 8. The points at the edge of the image may not have eight points around them, so such points do not need to confirm the pixel features, because the surrounding points are insufficient, and in the remote sensing image, such points are located in the edge area, generally do not belong to the relevant features of the mountain, and have no specific surveying and mapping value, so such points need to be confirmed; Based on the confirmed point pixel value X i And the pixel values ​​H of the surrounding pixels i-q , using: Cz i-q =(X i -H i-q ) Confirm the pixel value difference Cz between this pixel and the surrounding pixels i-q , and then the different pixel value differences Cz associated with different surrounding pixels i-q Perform ratio processing, confirm the ratio sequence, and use this ratio sequence as the pixel feature of the pixel point (in different images, the same point may have different pixel values ​​under different lighting conditions, but the change ratios between corresponding pixels are the same. Since the corresponding points are located at fixed positions and their positions will not change, the change ratios associated with the corresponding points must be the same value). Confirm and record the different pixel features associated with different points in the remote sensing image in turn; S12, based on the pixel features associated with the corresponding points in the corresponding remote sensing images, points with the same pixel features are identified from the multiple groups of remote sensing images, and the identified points are marked as feature points (the pixel features associated with the feature points in the multiple groups of remote sensing images are the same), the pixel values ​​of the feature points in different remote sensing images are confirmed, and the confirmed groups of pixel values ​​about the feature points are averaged to determine the average pixel, and the determined average pixel is used as the feature pixel of the feature point (in the multiple groups of remote sensing images, there are the same feature points, and the pixel values ​​associated with the feature points are different, so the average processing method can better select the average pixel); S13, based on the several groups of feature points confirmed in the single group of remote sensing images, confirm the connecting lines between the corresponding adjacent feature points, and record the line length L of the connecting lines, and then perform mean processing on the line length L of the several groups of connecting lines confirmed from the multiple groups of remote sensing images, confirm the feature mean values ​​belonging to the adjacent feature points, and the feature points associated with the same connecting lines are the same (that is, in each remote sensing image, the points associated with the same connecting lines are the same feature points); Based on the feature mean values ​​determined between the corresponding adjacent feature points, multiple groups of remote sensing images are plane stretched or scaled, and the process stops when the length of the connecting line associated with the corresponding adjacent feature points in each remote sensing image is consistent with the determined feature mean value, and the remote sensing image after the plane adjustment is calibrated as the image to be fused; S14, fusing the multiple groups of images to be fused, aligning the same feature points in each image to be fused based on the determined feature points, and fusing the images to confirm the fused image, and adjusting the pixel values ​​associated with the corresponding feature points after fusion to the determined feature pixels. Since the method of fusing multiple groups of substantially the same images is relatively common in the prior art, it will not be described in detail here; Step 2: Based on the confirmed fused image, generate a mountain model for the fused image, and perform preliminary processing on the generated mountain model to decompose the mountain model into several grid areas. The method of generating a mountain model based on an image is relatively common in the prior art. Generally, a suitable edge detection algorithm is used to extract the edge information of the mountain from the fused image to outline the general outline. Common edge detection algorithms include the Canny edge detection algorithm and the Sobel operator. The Canny edge detection algorithm first performs a Gaussian filter on the fused image to remove noise from the image. This is because noise can cause erroneous results in edge detection. For example, in a mountain image, a noise point may be misjudged as an edge. Gaussian filtering effectively smoothes the image while retaining edge information by taking a weighted average of each pixel in the image and its surrounding pixels, and determining the weight according to the Gaussian distribution. The specific sub-steps for initial processing of the generated mountain model are as follows: S21, generating a set of grid surfaces, wherein a plurality of grids exist in the grid surface, and the area of ​​each grid is consistent, and the area of ​​the corresponding grid is a preset value, which is prepared by relevant operators based on experience, and then determining the bottom surface of the mountain model, whose bottom surface is parallel to the grid surface, so that the grid surface coincides with the bottom surface of the mountain model; S22, based on the determined grid surface, the corresponding grid surface is moved upward, and during the movement, it is confirmed that the outer surface of the mountain model belongs to a partial area of ​​the grid surface, and the confirmed partial area is marked as a grid area, so that the outer surface of the mountain model is decomposed into a plurality of grid areas; Specifically, the corresponding grid surface is parallel to the bottom surface of the corresponding mountain model. After the grid surface is generated, it can be moved upward to confirm a group of space bodies. When it moves upward, it is horizontally upward. Then, during the movement, the relevant area of ​​the mountain model covered by the corresponding grid surface can be determined, thereby determining the specific grid area. Each different mountain grid area has different associated related parameters. Step 3: Based on several groups of grid areas confirmed on the upper surface of the mountain model, the grid features of the grid areas are determined, and then based on the determined grid features of different grid areas, several groups of grid areas are sorted in association to select the best grid sequence. Specifically, different sorting methods have different feature presentations, so when sorting, there are several different sorting methods, each of which corresponds to a different grid sequence, so that the corresponding best grid sequence can be determined; The specific sub-steps for selecting the best grid sequence are: S31, based on the bottom surface determined by the mountain model, confirm the relevant distance J from different points in the corresponding grid area to the bottom surface k , where k represents different points in the grid area, and the corresponding groups of related distances J confirmed in the grid area k Perform mean processing to confirm the grid feature TZ belonging to this grid area; S32, randomly selecting a group of grid areas from the determined groups of edge grid areas as the starting area, and starting from the starting area, randomly selecting other grid areas as subsequent grid areas, and randomly combining them, until all the grid areas are combined to confirm the combination sequence, confirm the grid feature difference between adjacent grid areas in the combination sequence, and the grid feature difference = |grid feature of the previous group of grid areas - grid feature of the next group of grid areas|, and summing the grid feature differences of the groups associated with the combination sequence to confirm the combination feature corresponding to the combination sequence; Then confirm the combination features of other combination sequences one by one; Then randomly select other grid areas as the starting areas, and confirm the subsequent combination sequences, and simultaneously confirm the combination features associated with the corresponding groups of combination sequences; S33, selecting a minimum value from the combination features associated with the plurality of groups of combination sequences, and taking the combination sequence associated with the minimum value as the optimal grid sequence; Specifically, because the outer surface of the mountain model has been decomposed into several grid areas, there are different grid features between each grid area, that is, related parameters such as height difference, so this kind of height difference needs to be confirmed. From the associated several groups of combination sequences and the associated several groups of combination features, several groups of grid feature differences are confirmed and summed, and the combination features associated with the corresponding combination sequence are confirmed. The selected optimal combination sequence is relatively gentle in the overall change of height difference, which is convenient for the drone to carry out relevant surveying and mapping. It is not necessary to make specific adjustments to the up and down height differences all the time. Fine-tuning is sufficient, which can effectively guarantee the surveying and mapping accuracy of the drone during surveying and mapping, and ensure the overall surveying and mapping accuracy of the corresponding landform and terrain area. In actual surveying and mapping, the drone needs to adjust the actual flight altitude of the drone in real time based on the height difference between the drone and the ground. When the planned route is in the lowest state in terms of height difference, then relative to the drone, the difference in its adjusted parameters will also be lower, and there will be no errors due to large-scale adjustments. It can effectively guarantee the surveying and mapping accuracy during surveying and mapping, and improve the overall effect of its landform and terrain area.

[0010] Step 4: Based on the selected optimal grid sequence, confirm the feature points of the corresponding grid area in the optimal grid sequence, then connect the confirmed groups of feature points, confirm the surveying and mapping route associated with the optimal grid sequence, and control the drone to conduct field surveying with the determined optimal grid sequence and surveying and mapping route. The specific sub-steps for determining the surveying and mapping route are as follows: S41, starting from the first grid area of ​​the best grid sequence and gradually performing correlation processing: based on the grid features confirmed in the corresponding grid area and the different correlation distances J associated with different points in the grid area k , confirm the center point of the corresponding grid area (the center point of the grid area has been calibrated in advance and can be directly extracted and confirmed), calibrate it as the midpoint of the grid, take the midpoint of the grid as the center, and construct a set of radiation circles with a radius of R1, where R1 is a preset value, and its specific value is determined by the operator based on experience. The points inside this radiation circle are calibrated as points to be selected, and the relevant distances of the points to be selected are differenced with the grid features of the corresponding grid area to confirm the relevant differences, and the relevant differences = |J k -Grid feature|, select the minimum value from the confirmed groups of related differences, and mark the point to be selected associated with the minimum value as the surveying and mapping point; S42, confirming the surveying and mapping points for each corresponding grid area, connecting several surveying and mapping points confirmed in sequence, confirming the surveying and mapping route, and controlling the UAV based on the confirmed surveying and mapping route to enable the UAV to perform field surveying and mapping.

[0011] On the basis of the grid area, through a unique grid feature calculation method, the distance from the point in the grid to the bottom surface is comprehensively considered and the average is taken as the key feature, which deeply explores the intrinsic structural information of the mountain model. In the process of selecting the best grid sequence, the random combination and feature difference sum analysis are innovatively adopted to accurately select the best sequence with a smooth change in height difference from many possible combination sequences. Based on the surveying and mapping route determined by this sequence, the concepts of grid midpoints and radiation circles are fully utilized, and the surveying and mapping points are cleverly selected, so that the drone does not need to frequently and drastically adjust the flight altitude during actual surveying and mapping, but only needs to make small and precise adjustments. This not only greatly reduces the error caused by altitude adjustment and ensures the stability of drone surveying and mapping accuracy, but also significantly improves the surveying and mapping efficiency, ensuring that the terrain data of the landform area to be surveyed can be obtained quickly and accurately.

[0012] Some of the data in the above formulas are dimensionless and numerically calculated. Meanwhile, the contents not described in detail in this specification belong to the prior art known to those skilled in the art.

[0013] The above embodiments are only used to illustrate the technical method of the present invention rather than to limit it. Although the present invention has been described in detail with reference to the preferred embodiments, those skilled in the art should understand that the technical method of the present invention may be modified or replaced by equivalents without departing from the spirit and scope of the technical method of the present invention.

Claims

1. The UAV mapping method based on topographic mapping accuracy is characterized by: The following steps are involved: Step 1: Acquire remote sensing images of the landform area to be surveyed, where there are multiple groups of remote sensing images, confirm pixel features associated with different points in the multiple groups of remote sensing images, select points with the same features from the multiple groups of remote sensing images based on the confirmed pixel features, and then fuse the multiple groups of remote sensing images based on the points with the same features determined in the multiple groups of remote sensing images to confirm the fused image belonging to the landform area to be surveyed; Step 2: Based on the confirmed fused image, a mountain model of the fused image is generated, and the generated mountain model is preliminarily processed to decompose the mountain model into a number of grid areas; Step 3: Based on the several groups of grid areas confirmed on the upper surface of the mountain model, the grid features of the grid areas are determined, and then based on the determined grid features of different grid areas, the several groups of grid areas are correlated and sorted to select the best grid sequence; Step 4: Based on the selected optimal grid sequence, confirm the feature points of the corresponding grid area in this optimal grid sequence, then connect the confirmed groups of feature points, confirm the surveying and mapping route associated with this optimal grid sequence, and control the drone to conduct field surveying using the determined optimal grid sequence and surveying route.

2. The method for unmanned aerial vehicle mapping based on topographic mapping accuracy according to claim 1 is characterized in that: In the step 1, the specific method of determining the fused image of the landform area to be surveyed is: S11, from the multiple groups of remote sensing images associated with the landform area to be surveyed, confirm the pixel features of the corresponding points in a single group of remote sensing images: confirm the point pixel value of the corresponding point and mark it as X i , where i represents the corresponding point in the remote sensing image, and the pixel value of the surrounding pixels of this point is calibrated as H i-q , where q=1, 2, ..., 8; Based on the confirmed point pixel value X i And the pixel values ​​H of the surrounding pixels i-q , using: Cz i-q =(X i -H i-q ) Confirm the pixel value difference Cz between this pixel and the surrounding pixels i-q , and then the different pixel value differences Cz associated with different surrounding pixels i-q Perform ratio processing, confirm the ratio sequence, use this ratio sequence as the pixel feature of the pixel point, and confirm and record the different pixel features associated with different points in the remote sensing image in turn; S12, based on the pixel features associated with the corresponding points in the corresponding remote sensing images, identifying points with the same pixel features from multiple groups of remote sensing images, marking the identified points as feature points, confirming the pixel values ​​of the feature points in different remote sensing images, and performing mean processing on the confirmed groups of pixel values ​​about the feature points to determine the mean pixel, and using the determined mean pixel as the feature pixel of the feature point; S13, based on the several groups of feature points confirmed in the single group of remote sensing images, confirm the connecting lines between the corresponding adjacent feature points, and record the line length L of the connecting lines, and then perform mean processing on the line length L of the several groups of connecting lines confirmed from the multiple groups of remote sensing images, and confirm the feature mean value belonging to the adjacent feature points, and the feature points associated with the same connecting lines are all the same; Based on the feature mean determined between corresponding adjacent feature points, multiple groups of remote sensing images are stretched or scaled in the plane, and the process stops when the length of the connecting lines associated with the corresponding adjacent feature points in each remote sensing image is consistent with the determined feature mean, and the remote sensing images after the plane adjustment are calibrated as the images to be fused.

3. The method for unmanned aerial vehicle mapping based on topographic mapping accuracy according to claim 2 is characterized in that: The step one also includes: S14, fusing multiple groups of images to be fused, aligning the same feature points in each image to be fused based on the determined feature points, and performing image fusion to confirm the fused image, and adjusting the pixel values ​​associated with the corresponding feature points after fusion to the determined feature pixels.

4. The method for unmanned aerial vehicle mapping based on topographic mapping accuracy according to claim 1 is characterized in that: In the step 2, the specific sub-steps of performing initial processing on the generated mountain model are: S21, generating a set of grid surfaces, wherein the grid surfaces contain a plurality of grids, and the area of ​​each grid is consistent, and the area of ​​the corresponding grid is a preset value, and then determining the bottom surface of the mountain model, wherein the bottom surface is parallel to the grid surface, and the grid surface is overlapped with the bottom surface of the mountain model; S22. Based on the determined grid surface, the corresponding grid surface is moved upward, and during the movement, it is confirmed that the outer surface of the mountain model belongs to a partial area of ​​the grid surface, and the confirmed partial area is marked as a grid area, so the outer surface of the mountain model is decomposed into a plurality of grid areas.

5. The method for unmanned aerial vehicle mapping based on topographic mapping accuracy according to claim 4 is characterized in that: In step 3, the specific sub-steps for selecting the best grid sequence are: S31. Based on the bottom surface determined by the mountain model, determine the relevant distance J from different points in the corresponding grid area to the bottom surface. k , where k represents different points in the grid area, and the corresponding groups of related distances J confirmed in the grid area k Perform mean processing to confirm the grid feature TZ belonging to this grid area; S32, randomly selecting a group of grid areas from the determined groups of edge grid areas as the starting area, and starting from the starting area, randomly selecting other grid areas as subsequent grid areas, and randomly combining them, until all the grid areas are combined to confirm the combination sequence, confirm the grid feature difference between adjacent grid areas in the combination sequence, and the grid feature difference = |grid feature of the previous group of grid areas - grid feature of the next group of grid areas|, and summing the grid feature differences of the groups associated with the combination sequence to confirm the combination feature corresponding to the combination sequence; Then confirm the combination features of other combination sequences one by one; Then randomly select other grid areas as the starting areas, and confirm the subsequent combination sequences, and simultaneously confirm the combination features associated with the corresponding groups of combination sequences; S33. Select a minimum value from the combination features associated with the plurality of groups of combination sequences, and use the combination sequence associated with the minimum value as the optimal grid sequence.

6. The method for unmanned aerial vehicle mapping based on topographic mapping accuracy according to claim 5 is characterized in that: In step 4, the specific sub-steps for determining the surveying route are: S41, starting from the first grid area of ​​the best grid sequence and gradually performing correlation processing: based on the grid features confirmed in the corresponding grid area and the different correlation distances J associated with different points in the grid area k , confirm the center point of the corresponding grid area, mark it as the grid midpoint, take the grid midpoint as the center, construct a set of radiating circles with a radius of R1, where R1 is a preset value, mark the points inside this radiating circle as the points to be selected, and perform difference processing on the relevant distance of the points to be selected and the grid features of the corresponding grid area, confirm the relevant difference, and its relevant difference = |J k -Grid feature|, select the minimum value from the confirmed groups of related differences, and mark the point to be selected associated with the minimum value as the surveying and mapping point; S42, confirming the surveying and mapping points for each corresponding grid area, connecting several surveying and mapping points confirmed in sequence, confirming the surveying and mapping route, and controlling the UAV based on the confirmed surveying and mapping route to enable the UAV to perform field surveying and mapping.

Images