A method for evaluating elevation accuracy based on optical remote sensing images of the lunar south pole

Abstract

The application discloses a kind of elevation accuracy evaluation methods based on lunar south pole optical remote sensing image, belong to deep space exploration technical field.The method according to the region position of the lunar south pole digital elevation model DEM to be evaluated corresponding, obtains the optical remote sensing image of the region position.The lunar south pole digital elevation model DEM data to be evaluated is rendered as the illumination rendering image of the moment of photographing.Optical remote sensing image and illumination rendering image are respectively first normalized, then binarization processing is carried out;The binarization result of optical remote sensing image and illumination rendering image is compared, mismatch chart is drawn, and the elevation accuracy evaluation data of the lunar south pole digital elevation model DEM to be evaluated is obtained.The present application significantly improves the data source reliability of landing area site selection, improves the accuracy of lunar south pole exploration task landing area site selection.

Description

Technical Field

[0001] This invention relates to a method for evaluating the elevation accuracy based on optical remote sensing images of the lunar south pole, belonging to the field of deep space exploration technology. Background Technology

[0002] In lunar soft landing and exploration missions, the Digital Elevation Model (DEM) serves as core data describing the lunar surface topography. The overall design of the probe utilizes the DEM to render lunar illumination and shading conditions. The accuracy of the rendering results is a fundamental prerequisite for safe landing and correct path planning, and it depends on the elevation accuracy of the DEM used. Inaccurate illumination rendering can lead to misjudgments of illuminated and shadowed areas, causing the probe to mistakenly enter shadowed regions and affecting the mission's success or failure. Furthermore, the lower the solar altitude angle, the more pronounced the shadow changes, meaning that illumination conditions are more sensitive to the accuracy of terrain elevation. Therefore, ensuring the elevation accuracy and reliability of the DEM data source has become a critical issue that urgently needs to be addressed in the field of lunar exploration, especially for lunar south pole exploration missions.

[0003] Currently, successful lunar soft landing missions, both domestically and internationally, are mainly concentrated in the mid-to-low latitude regions of the Moon. These regions have high solar altitude angles and exhibit periodic variations in illumination, meaning that terrain elevation accuracy has little impact on lunar illumination conditions. No methods for analyzing terrain elevation accuracy at low solar altitude angles have yet been discovered. However, in future lunar exploration missions, the lunar south pole, due to its abundant and diverse resources and the presence of water ice, is a hot research area for major spacefaring nations both domestically and internationally. Therefore, it is necessary to develop methods for assessing lunar surface terrain elevation accuracy at low solar altitude angles specifically for the lunar south pole. Summary of the Invention

[0004] The technical problem solved by this invention is to overcome the shortcomings of the prior art and propose an elevation accuracy evaluation method based on lunar south pole optical remote sensing images. This method evaluates lunar DEM data based on existing real-world optical remote sensing data, significantly improves the reliability of the data source for landing area selection, and enhances the accuracy of landing area selection for lunar south pole exploration missions.

[0005] The technical solution of this invention is:

[0006] A method for evaluating elevation accuracy based on lunar south pole optical remote sensing images includes:

[0007] Based on the location of the region corresponding to the digital elevation model (DEM) of the lunar south pole to be evaluated, acquire optical remote sensing images of the region and record the time when the optical remote sensing images were taken.

[0008] The digital elevation model (DEM) data of the lunar south pole to be evaluated is rendered as a light-rendered image at the time of the image capture.

[0009] The optical remote sensing image and the illumination rendering image are first normalized and then binarized. The binarization results of the optical remote sensing image and the illumination rendering image are compared to draw a mismatch map, obtain the length of each mismatch line segment, count the number of mismatch line segments, invert the terrain elevation error, and form the elevation accuracy evaluation data of the lunar south pole digital elevation model (DEM) to be evaluated.

[0010] Furthermore, the digital elevation model (DEM) data of the lunar south pole to be evaluated is rendered as an illumination rendering image at the time of the image capture. The specific method is as follows:

[0011] All pixels on the lunar south pole digital elevation model (DEM) are taken as the analysis points one by one. Rays are drawn outward from the analysis points as the center to calculate the terrain occlusion angle of the analysis points in various directions.

[0012] After obtaining the terrain occlusion angles of each analyzed point in various directions, the analysis point is then determined to be occluded at the time of image capture based on the solar ephemeris, thus obtaining the lighting rendering image.

[0013] Furthermore, by drawing rays outward from the point being analyzed, the terrain shading angles of the analyzed point in various directions are calculated. The specific method is as follows:

[0014] Draw a ray outward from the point being analyzed in a direction of 0–360°, and calculate the occlusion angle of each pixel on the ray in each orientation:

[0015]

[0016] In the formula, Let D be the point being analyzed. ij At azimuth angle θ azi The occlusion angle H corresponding to the k-th pixel covered by the ray on the surface. k H is the elevation value corresponding to the k-th pixel, and H0 is the elevation value of the point D being analyzed. ij Elevation value, L k Let D be the point being analyzed. ij Distance to the k-th pixel;

[0017] Based on the occlusion angle of each pixel on the ray, the maximum value is taken as the terrain occlusion angle of the analyzed point in this orientation.

[0018] Furthermore, for each analyzed point, a ray is drawn outward from the analyzed point, and the effective length of the ray from the analyzed point is defined as the data range for analyzing the terrain occlusion angle; the method for calculating the effective length is as follows:

[0019] The outer latitude is calculated and analyzed when analyzing the terrain shading angle:

[0020]

[0021] In the formula, θ' is the apparent radius of the sun, θ lati R represents the outer latitude, R represents the lunar radius, and h represents the highest altitude of the lunar south pole region.

[0022] The effective length is:

[0023]

[0024] In the formula, L is the effective length.

[0025] Furthermore, after obtaining the terrain occlusion angles at each location of the analyzed point, the system determines whether the analyzed point was occluded at the time of image capture based on the solar ephemeris. The specific method is as follows:

[0026]

[0027] In the formula, S i,j Let D be the point being analyzed. ij The occlusion value at the moment of taking the picture. Let D be the point being analyzed. ij At azimuth angle θ azi The terrain obstruction angle above, The azimuth angle θ corresponding to the moment the image was taken azi The solar altitude angle is θ', where θ' is the apparent radius of the sun.

[0028] Analysis point D ij The occlusion value is used as the pixel value to obtain the lighting rendering image using DEM.

[0029] Furthermore, the optical remote sensing image and the illumination rendering image are normalized to distribute the image pixel values ​​evenly within the pixel grayscale range of 0 to 255.

[0030] Furthermore, the normalized optical remote sensing image and the illumination rendering image are binarized separately. The binarization method is as follows:

[0031] If the overall grayscale value of the image is below the minimum value of the set normal range, and the threshold is set to 10, that is, if the grayscale value is less than 10, then the pixel is considered to be in the shadow area; otherwise, the pixel is considered to be in the illuminated area.

[0032] If the overall grayscale value of the image is above the maximum value of the set normal range, and the threshold is set to 50, that is, if the grayscale value is less than 50, then the pixel is considered to be in the shadow area; otherwise, the pixel is considered to be in the illuminated area.

[0033] If the overall grayscale value of the image is within the set normal range, the threshold is selected as 10% of the maximum value of the normalized image pixels. That is, if the grayscale value is less than the threshold, the pixel is considered to be in the shadow area; otherwise, the pixel is considered to be in the illuminated area.

[0034] Pixels in the shadow area are set to 0, and pixels in the illuminated area are set to 1.

[0035] Furthermore, the binarized results of the optical remote sensing image and the illumination rendering image are compared and an XOR operation is performed to obtain a mismatch map. On the mismatch map, a ray is drawn along the direction of sunlight incidence to cover the entire map. For each ray, the mismatched area on the ray is marked, that is, each ray is divided into several mismatched line segments. The length of each mismatched line segment is measured, and the number of mismatched line segments is counted. All of these are used as evaluation results of elevation accuracy.

[0036] Furthermore, based on the mismatch map, the terrain elevation error is inverted, and the terrain elevation error is:

[0037]

[0038] In the formula, L k Let K be the length of the k-th mismatched segment, and K be the number of mismatched segments. The azimuth angle θ of the ray containing the mismatched line segment at the moment of image capture. azi The solar altitude angle, δ H This represents the terrain elevation error.

[0039] Furthermore, based on the mismatch map, the overall mismatch rate and the bright area mismatch rate are calculated. The specific method is as follows:

[0040] The overall mismatch rate is:

[0041]

[0042] In the formula, α1 is the mismatch rate across the entire region, and N Mis N represents the total number of mismatch points in the binarized optical remote sensing image and the illuminated rendering image, where N represents the total number of points in the entire region.

[0043] Extract pixels from both images that are illuminated, and designate them as corresponding pixels; count the number of corresponding pixels; calculate:

[0044]

[0045] In the formula, α2 is the mismatch rate in the bright area, and N Mis-Comp The number of pixels in the illuminated area of ​​the rendered image minus the number of pixels with the same name; N DEM The number of pixels in the illuminated area of ​​the image for lighting rendering.

[0046] The advantages of this invention compared to the prior art are:

[0047] (1) This invention uses existing optical remote sensing images of the moon to quantitatively evaluate the elevation accuracy of lunar DEM topographic data, filling the gap in the current lack of effective quantitative evaluation methods for lunar DEM elevation accuracy and improving the reliability of lunar topographic data sources.

[0048] (2) This invention proposes and defines an image mismatch rate index by comparing the binarized mismatch between optical remote sensing images and DEM terrain rendering images, and calculates the deviation between terrain rendering images and optical remote sensing images, thus solving the problem of quantitative evaluation of the accuracy of DEM terrain rendering results and realizing the quantitative evaluation of DEM terrain rendering images. Attached Figure Description

[0049] Various other advantages and benefits will become apparent to those skilled in the art upon reading the following detailed description of preferred embodiments. The accompanying drawings are for illustrative purposes only and are not intended to limit the invention. Furthermore, the same reference numerals denote the same parts throughout the drawings. In the drawings:

[0050] Figure 1 This is a flowchart of the elevation accuracy evaluation method according to an embodiment of the present invention;

[0051] Figure 2 This is a schematic diagram of a lunar surface optical remote sensing image according to an embodiment of the present invention;

[0052] Figure 3 This is a schematic diagram illustrating the calculation of terrain obstruction angle in an embodiment of the present invention;

[0053] Figure 4 This is a schematic diagram illustrating the determination of ray length for terrain shading angle calculation in an embodiment of the present invention;

[0054] Figure 5 This is a lunar surface illumination rendering based on lunar DEM data, as shown in an embodiment of the present invention.

[0055] Figure 6 This is a schematic diagram of the binarization result of a lunar surface optical remote sensing image according to an embodiment of the present invention;

[0056] Figure 7 This is a schematic diagram of the binarization result of the DEM illumination rendering image in an embodiment of the present invention;

[0057] Figure 8 This is a schematic diagram of the mismatch results between the illuminated area and the shaded area in an embodiment of the present invention;

[0058] Figure 9 This is a schematic diagram of the illumination area mismatch result in an embodiment of the present invention. Detailed Implementation

[0059] Exemplary embodiments of the present disclosure will now be described in more detail with reference to the accompanying drawings. While exemplary embodiments of the present disclosure are shown in the drawings, it should be understood that the present disclosure may be implemented in various forms and should not be limited to the embodiments set forth herein. Rather, these embodiments are provided so that this disclosure will be thorough and complete, and will fully convey the scope of the disclosure to those skilled in the art.

[0060] This invention proposes a method for evaluating the elevation accuracy based on optical remote sensing images of the lunar south pole, such as... Figure 1 As shown, it includes:

[0061] Step 1: Based on the location of the lunar south pole region in the Digital Elevation Model (DEM) of the Moon, acquire optical remote sensing images of the corresponding area. These images can be obtained from existing optical image data from lunar mapping satellites, such as images from the Narrow Angle Camera (NAC) of the Lunar Reconnaissance Orbiter (LRO). Figure 2 As shown.

[0062] Step 2: Record the actual time of taking the optical remote sensing image. In this embodiment, based on the optical remote sensing image information, the actual time of taking the image is obtained as 17:10:58 on December 13, 2009.

[0063] Step 3: Based on the solar ephemeris relationship, render the DEM data to be evaluated as an illumination rendering map at the same time as the actual optical remote sensing image.

[0064] First, consider the shading caused by terrain elevation variations, and calculate the terrain shading angle using DEM data. For example... Figure 3 As shown, each pixel on the DEM to be evaluated is taken as the analysis point, and rays are drawn outward from the analysis point to analyze the maximum terrain occlusion angle in each direction, and finally the occlusion angle data of each pixel and each direction are obtained.

[0065] When constructing the ray, the ray length is determined by considering the topographic relief of the lunar south pole. The lunar curvature ensures that distant mountains no longer obscure the analyzed point, thus eliminating the need for calculations over a larger area. The highest mountain in the lunar south pole region, h, is approximately 7 km high, and the apparent radius of the sun, θ', is approximately 0.27°. Considering the candidate point's elevation of 0 km (the actual candidate point is a relatively high location in a local area with an elevation greater than 0 km) and the maximum elevation of the obstructing object of 7 km, ... Figure 4 As shown, the outward latitude is used when calculating the shading angle:

[0066]

[0067] R is the radius of the Moon, 1737.4 km. The corresponding ray length L is:

[0068]

[0069] Considering a certain margin, the ray length L is taken as 164km.

[0070] Therefore, by drawing rays outward from the analyzed point in a 0–360° direction, the maximum terrain shading angle in each direction is obtained. This is used to calculate pixel D. ij At azimuth angle θ azi The following terrain obstruction angle For example:

[0071]

[0072] The azimuth angle θ azi The occlusion angle corresponding to the m pixels covered by the ray L is calculated as follows:

[0073]

[0074] H k Let H0 be the elevation value of the k-th pixel out of m pixels, and L be the elevation value of the point being analyzed. k The distance between the analyzed point and the k-th pixel is given.

[0075] After obtaining the terrain occlusion angle, calculate whether the analyzed point is occluded based on the solar ephemeris. If occluded, the point is set to 0; otherwise, the percentage is calculated based on the exposed area of ​​the sun. The solar ephemeris can be looked up using the DE405 ephemeris table or calculated based on the Earth-Moon analytical ephemeris.

[0076]

[0077] S i,j The occlusion status of the analyzed point (i, j) is represented by the pixel value of each point in the lighting rendering map. The calculation method is as follows. Due to subsequent normalization processing, this model does not consider the light intensity loss caused by lunar surface sunlight intensity and lunar surface reflection.

[0078]

[0079] The azimuth angle θ at that moment azi The solar altitude angle below.

[0080] Step 4: Normalize the DEM optical rendering and optical remote sensing images.

[0081] Optical image data is S light The image data rendered using the DEM is S DEM (like Figure 5As shown, the two images are normalized so that the pixel values ​​are evenly distributed within the pixel grayscale range of 0 to 255.

[0082]

[0083] Step 5: Binarize the normalized results of the DEM optical rendering and optical remote sensing images.

[0084] Convert the two floating-point images into binary images, denoted as S. light-01 and S DEM-01 ,like Figure 6 , Figure 7 As shown. The calculation method is detailed in the following formula:

[0085]

[0086] Where M is the grayscale threshold, in actual calculations, the threshold is selected in three cases for judging the shadow area of ​​optical images:

[0087] If the image is generally dark and the grayscale value is low (the overall grayscale value of the image is basically below 20, based on experience), select a threshold of 10, that is, a grayscale value less than 10 is a shadow area and a grayscale value greater than 10 is a light area.

[0088] If the image is generally bright and the grayscale value is high (the overall grayscale value of the image is generally above 40, based on experience), select a threshold of 50, that is, a grayscale value less than 50 is a shadow area and a grayscale value greater than 50 is a lit area.

[0089] Under normal circumstances, the image has a uniform distribution of light and dark areas. A threshold of 10% (grayscale value 10% × 255 = 25.5) is selected, meaning grayscale values ​​less than 25.5 are considered shadow areas, and values ​​greater than 25.5 are considered illuminated areas. The selection principle is to determine the optimal threshold by comparing the binarization results under different thresholds with human visual observation of the optical image.

[0090] Step 6: Compare the binarized results of the DEM optical rendering and the optical remote sensing image, count the mismatch points, and draw a mismatch map.

[0091] XORing the binarized DEM rendering and optical remote sensing images yields a mismatch map (showing all mismatched areas). In actual on-orbit exploration missions, the accuracy of lunar surface illumination times is paramount; therefore, in addition to calculating all mismatched areas, extra attention needs to be paid to mismatches in bright areas.

[0092] The mismatch rate is calculated, including the overall mismatch rate α1 and the bright area mismatch rate α2 (the number of mismatched bright area points in the DEM rendering / the total number of bright area points in the DEM rendering), as a global reference indicator.

[0093] The mismatch rate for the entire area (including shaded and lit areas) is:

[0094]

[0095] Where, N Mis For S light-01 and S DEM-01 The total number of mismatch points, where N represents the total number of points in the entire region.

[0096] Extract S light-01 and S DEM-01 The areas in both images that are illuminated (i.e., pixels with a value of 1, hereinafter referred to as "same-name pixels") are marked as S. Comp Pixels that are both 1 can be extracted by matrix multiplication; that is, the two images corresponding to pixels whose multiplication result is 1 are obtained by multiplying them. Figure 1 The values ​​are all set to 1.

[0097] S Comp =S light-01 ·S DEM-01 (10)

[0098] First, calculate the matching rate, which is the percentage of corresponding points in the illuminated area of ​​the optical image. Then, calculate the mismatch rate. The calculation method is as follows:

[0099]

[0100] Where α2 represents the percentage of mismatch points in the illuminated area; N Mis-Comp Subtract S from the number of illuminated areas in the rendered image Comp The number of data values ​​of 1 (i.e., N) DEM -S Comp N represents the number of mismatch points. DEM For S DEM-01 The number of data values ​​of 1 indicates the illuminated area of ​​the rendered image. For example... Figure 8 , Figure 9 As shown.

[0101] The obtained mismatch rate α1 in the whole region and mismatch rate α2 in the bright area are used as global reference indicators and also as quantitative evaluation indicators of image mismatch.

[0102] Step 7: Analyze the mismatch length based on the mismatch diagram.

[0103] By performing an XOR operation on the binarized DEM rendering and optical remote sensing images, a mismatch map (displaying all mismatched areas) can be obtained. For the mismatch map, a ray is drawn along the direction of sunlight incidence to cover the entire map (① Note that repetition should be avoided; a ray can be drawn for every pixel along the top and right edges; ② Avoid ray lengths that are too short; only rays longer than 10 pixels are included in the statistics). For each ray, the mismatched areas on it are marked, meaning each ray is divided into several mismatched segments, each with a length.

[0104] Step 8: Statistical analysis of elevation errors across the entire DEM area to be evaluated; inversion of topographic elevation errors; and provision of the final topographic elevation accuracy evaluation results. The elevation error calculation method is as follows:

[0105]

[0106] L k Let K be the length of the mismatched line segment. There are K such segments.

[0107] The lengths of all mismatched line segments are statistically analyzed, and the elevation accuracy results are given. In this embodiment, the maximum elevation error is 0.2286m, and the elevation accuracy at 3σ is 0.2078m.

[0108] Table 1. Evaluation Results of Elevation Accuracy

[0109]

[0110] This invention addresses the lack of effective evaluation methods in current traditional technologies by transforming them into a quantitatively assessable algorithm through simulation modeling. It provides a quantitative and accurate evaluation method for assessing the elevation accuracy of DEM terrain data in landing areas, filling the gap in existing landing area selection methods for lunar south pole soft landing technology. This method has broad future application prospects. Future unmanned lunar exploration, manned lunar landings, Mars sample return missions, and manned Mars landings will increasingly focus on areas with higher scientific and application value, where terrain conditions and lighting variations will be more diverse and complex. This method can provide precise input data for landing area selection and path planning design for these missions, reducing risks during probe landing and mobile exploration. It guides the overall design, energy security analysis, work procedure formulation, and hibernation / wake-up design of deep space exploration missions, effectively improving the credibility of engineering implementation from multiple dimensions.

[0111] The embodiments described above are merely preferred embodiments of the present invention. Ordinary variations and substitutions made by those skilled in the art within the scope of the technical solution of the present invention should be included within the protection scope of the present invention.

Claims

1. A method for evaluating the elevation accuracy based on lunar south pole optical remote sensing images, characterized in that, include: Based on the location of the region corresponding to the digital elevation model (DEM) of the lunar south pole to be evaluated, acquire optical remote sensing images of the region and record the time when the optical remote sensing images were taken. The digital elevation model (DEM) data of the lunar south pole to be evaluated is rendered as a light-rendered image at the time of the image capture. The optical remote sensing image and the illumination rendering image are first normalized and then binarized. By comparing the binarization results of optical remote sensing images and illumination rendering images, a mismatch map is drawn, the length of each mismatched line segment is obtained, the number of mismatched line segments is counted, the terrain elevation error is inverted, and the elevation accuracy evaluation data of the lunar south pole digital elevation model (DEM) to be evaluated is formed. The digital elevation model (DEM) data of the lunar south pole to be evaluated is rendered as a lighting rendering image at the time of the image capture. The specific method is as follows: take all the pixels on the digital elevation model (DEM) of the lunar south pole as the points to be analyzed one by one, draw rays outward from the points to be analyzed as the center, and calculate the terrain occlusion angle of the points to be analyzed in each direction. After obtaining the terrain occlusion angles of each analyzed point in various directions, the analysis point is then determined to be occluded at the time of image capture based on the solar ephemeris, thereby obtaining the lighting rendering image. Draw rays outward from the point being analyzed, and calculate the terrain shading angle of the point being analyzed in various directions. The specific method is as follows: Draw a ray outward from the point being analyzed in a direction of 0~360°, and calculate the occlusion angle of each pixel on the ray in each direction: In the formula, For the point being analyzed In azimuth The first ray covered by the rays The occlusion angle corresponding to each pixel For the first The elevation value corresponding to each pixel. For the point being analyzed Elevation value, For the point being analyzed With the The distance of one pixel; Based on the occlusion angle of each pixel on the ray, the maximum value is taken as the terrain occlusion angle of the analyzed point in this orientation. After obtaining the terrain occlusion angles at each location of the analyzed point, the method is as follows: Based on the solar ephemeris, determine whether the analyzed point was occluded at the time of image capture. In the formula, For the point being analyzed The occlusion value at the moment of taking the picture. For the point being analyzed In azimuth The terrain obstruction angle above, The azimuth angle corresponding to the moment the picture was taken The solar altitude angle below, The apparent radius of the sun; Analysis point The occlusion value is used as the pixel value to obtain the lighting rendering image using DEM.

2. The elevation accuracy evaluation method based on lunar south pole optical remote sensing images according to claim 1, characterized in that, For each point being analyzed, a ray is drawn outward from the point being analyzed, and the effective length of the ray from the point being analyzed is defined as the data range for analyzing the terrain occlusion angle; the method for calculating the effective length is as follows: The outer latitude is calculated and analyzed when analyzing the terrain shading angle: In the formula, The apparent radius of the sun. To expand latitude, The radius of the moon, This is the highest altitude in the lunar south polar region; The effective length is: In the formula, This is the effective length.

3. The elevation accuracy evaluation method based on lunar south pole optical remote sensing images according to claim 1, characterized in that, The optical remote sensing images and the illumination rendering images are normalized to distribute the image pixel values ​​evenly within the pixel grayscale range of 0 to 255.

4. The elevation accuracy evaluation method based on lunar south pole optical remote sensing images according to claim 3, characterized in that, The normalized optical remote sensing image and the illumination rendering image are binarized separately. The binarization method is as follows: If the overall grayscale value of the image is below the minimum value of the set normal range, and the threshold is set to 10, that is, if the grayscale value is less than 10, then the pixel is considered to be in the shadow area; otherwise, the pixel is considered to be in the illuminated area. If the overall grayscale value of the image is above the maximum value of the set normal range, and the threshold is set to 50, that is, if the grayscale value is less than 50, then the pixel is considered to be in the shadow area; otherwise, the pixel is considered to be in the illuminated area. If the overall grayscale value of the image is within the set normal range, the threshold is selected as 10% of the maximum value of the normalized image pixels. That is, if the grayscale value is less than the threshold, the pixel is considered to be in the shadow area; otherwise, the pixel is considered to be in the illuminated area. Pixels in the shadow area are set to 0, and pixels in the illuminated area are set to 1.

5. The elevation accuracy evaluation method based on lunar south pole optical remote sensing images according to claim 1, characterized in that, By comparing the binarized results of the optical remote sensing image and the illumination rendering image, an XOR operation is performed to obtain a mismatch map. On the mismatch map, a ray is drawn along the direction of sunlight incidence to cover the entire map. For each ray, the mismatched area on the ray is marked, that is, each ray is divided into several mismatched line segments. The length of each mismatched line segment is measured, and the number of mismatched line segments is counted. All of these are used as evaluation results of elevation accuracy.

6. The elevation accuracy evaluation method based on lunar south pole optical remote sensing images according to claim 1, characterized in that, Based on the mismatch map, the terrain elevation error is retrieved and is as follows: In the formula, For the first The length of each mismatched segment The number of mismatched line segments. The azimuth angle of the ray containing the mismatched line segment at the moment of image capture. solar altitude angle, This represents the terrain elevation error.

7. The elevation accuracy evaluation method based on lunar south pole optical remote sensing images according to claim 1, characterized in that, Based on the mismatch map, the mismatch rate of the entire region and the mismatch rate of the bright area are calculated using the following method: The overall mismatch rate is: In the formula, The mismatch rate for the entire region. This represents the total number of mismatch points between the binarized optical remote sensing image and the illuminated rendered image. This represents the total number of points in the entire region. Extract pixels that are illuminated in both images and use them as corresponding pixels, then count the number of corresponding pixels. calculate: In the formula, The bright area mismatch rate, Subtract the number of pixels with the same name from the number of pixels in the illuminated area of ​​the light-rendered image; The number of pixels in the illuminated area of ​​the image for lighting rendering.

Images