Elevation defuzzification method of array SAR point cloud in mountainous area based on external DEM

Abstract

The present invention provides an array SAR mountain elevation defuzzification method based on an external DEM, which can accurately identify fuzzy point clouds and solve the number of fuzzy cycles, solve the problem of elevation fuzziness in mountain point clouds, and realize high-precision three-dimensional reconstruction in mountain areas. The present invention provides an array SAR mountain elevation defuzzification method based on an external DEM, which can accurately and efficiently identify fuzzy point clouds and determine the fuzzy cycle, and move them to the real elevation position, realize the fuzzy solution of mountain elevation, and further provide a certain reference and guarantee for the research of high-precision three-dimensional reconstruction methods in mountain areas.

Description

Technical Field

[0001] The invention relates to the technical field of synthetic aperture radar signal processing, and in particular to an array SAR mountain point cloud elevation defuzzification method based on an external DEM. Background Art

[0002] As an all-day, all-weather, high-resolution, penetrable active microwave imaging radar, the Synthetic Aperture Radar (SAR) imaging system has important application value in military reconnaissance, disaster monitoring, topographic mapping, marine research, resource exploration and other fields. The traditional SAR can only obtain two-dimensional images projected on the azimuth-slant range plane, which not only loses the height information of the observed target, but also has geometric distortion phenomena such as overlap, which is far from meeting the needs of production, life and national strategic management in other fields for three-dimensional information. Array SAR can obtain multi-angle SAR images of the same observation scene by adding multiple antennas in turn across the heading to form a virtual synthetic aperture. It has high-dimensional geometric resolution capability, while improving the coherence between multiple channels to ensure the accuracy of coherent measurement, thereby achieving high-precision three-dimensional reconstruction of the target.

[0003] Mountainous areas are typical natural targets, accounting for about two-thirds of my country's land area. High-precision three-dimensional imaging of them is crucial. The surface of mountainous areas is covered with a large number of distributed targets that are very sensitive to temporal decorrelation, which makes it difficult for traditional single-channel multi-flight tomographic SAR to obtain high-coherence mountainous three-dimensional point cloud data. Array SAR can effectively solve the temporal decorrelation problem caused by the revisit cycle with only one flight, making high-precision mountainous area three-dimensional reconstruction based on array SAR possible. Due to the limited discrete antenna elements of the array SAR system, the array SAR system cannot meet the full sampling requirements. When the elevation span of the observed scene is greater than the unambiguous elevation of the array SAR system, the point cloud will have elevation ambiguity. Due to the large undulations in the mountainous terrain, it is difficult to know the elevation span a priori. In addition, the unambiguous elevation and elevation resolution of the system restrict each other. Under the premise of a fixed number of channels, blindly improving the unambiguous elevation of the system by encrypting the baseline will sacrifice the elevation resolution. Therefore, limited by the scene characteristics and system design, the elevation ambiguity problem of array SAR mountainous point clouds is extremely prominent. However, there are few reports on the research of array SAR three-dimensional reconstruction in mountainous areas at home and abroad. The problem of solving the elevation ambiguity of mountainous point clouds needs to be solved urgently. Summary of the invention

[0004] In view of this, the present invention provides an array SAR mountain elevation defuzzification method based on external DEM, which can accurately identify fuzzy point clouds and solve the fuzzy cycle number, solve the problem of elevation fuzziness in mountain point clouds, and realize high-precision three-dimensional reconstruction in mountainous areas.

[0005] To achieve the above object, the technical solution of the present invention is:

[0006] A method for defuzzifying elevation of an array SAR mountain point cloud based on an external DEM comprises the following steps:

[0007] Step 1, obtaining three-dimensional point cloud data of mountainous areas through array SAR;

[0008] Step 2, calculating the unambiguous elevation of the target area;

[0009] Step 3, introduce the external DEM corresponding to the target area;

[0010] Step 4, calculate the height difference h1 between the target area and the external DEM;

[0011] Step 5, lower the target area to an unambiguous elevation and calculate the height difference h2 between it and the external source DEM;

[0012] Step 6: Raise the target area to an unambiguous elevation and calculate the height difference h3 between it and the external source DEM;

[0013] Step 7, determining the point cloud of a blurred period according to the height relationship among h1, h2 and h3;

[0014] Among them, if h2 of a certain point

[0015] Step 8, repeating steps 4 to 7 to find a point cloud with multiple blurred periods; for blurred points with multiple blurred periods, repeating step 8 until there are no blurred points;

[0016] Step 9: Move the blurred point cloud to the real location to achieve elevation deblurring of the mountain point cloud.

[0017] Wherein, in step 1, the array SAR uses an elevation super-resolution algorithm to generate a three-dimensional point cloud of the target area.

[0018] Wherein, the elevation super-resolution algorithm includes compressed sensing and multiple signal classification.

[0019] Among them, the specific expression of the unambiguous elevation of the array SAR system in step 2 is: Among them, Δ h is the unambiguous elevation, λ is the wavelength of the transmitted signal, H is the height of the carrier aircraft, y is the scene reference ground height, θ is the downward viewing angle, Δb and β are the baseline interval and baseline horizontal inclination respectively.

[0020] Among them, in actual engineering applications, the indicator used is the array SAR unambiguous height, and its specific expression is ​

[0021] Among them, it is characterized in that in step 9, moving the blurred point to the real position is to lower the blurred point that is blurred upward for N cycles by N unambiguous heights, and to raise the blurred point that is blurred downward for N cycles by N unambiguous heights.

[0022] Beneficial effects:

[0023] 1. The array SAR mountain elevation defuzzification method based on external DEM in the present invention can accurately and efficiently identify the fuzzy point cloud and determine the fuzzy period, and move it to the real elevation position to achieve the fuzzy solution of mountain elevation, thereby providing a certain reference and guarantee for the research on high-precision three-dimensional reconstruction methods in mountainous areas.

[0024] 2. The present invention forms an array antenna structure by sequentially adding multiple antennas in the cross-track direction, which has a high degree of directional resolution, thereby achieving a three-dimensional imaging effect of the target scene.

[0025] 3. The SRTM DEM introduced in the present invention has a plane resolution of 30m×30m, a relative elevation accuracy of 6m, and an absolute elevation accuracy of 16m, which can achieve a simple and fast solution to the elevation ambiguity problem with accurate results. BRIEF DESCRIPTION OF THE DRAWINGS

[0026] Figure 1 is a flow chart of the method of the present invention.

[0027] Figure 2 Schematic diagram of the array SAR antenna array structure.

[0028] Figure 3 The array SAR system with incident angle and scene height as variables does not ambiguity height.

[0029] Figure 4 Array SAR point cloud image of the target area.

[0030] Figure 5 The target area introduced for the present invention corresponds to the exogenous DEM target area SRTM DEM.

[0031] Figure 6 This is a fuzzy point cloud image identified by the method of the present invention.

[0032] Figure 7 This is the elevation defuzzification result map using the method of the present invention.

[0033] Figure 8 This is the elevation defuzzification result map of the target area. DETAILED DESCRIPTION

[0034] The present invention is described in detail below with reference to the accompanying drawings and embodiments.

[0035] Since the limited discrete antenna array elements cannot meet the full sampling requirements, the array SAR point cloud will be aliased in the elevation direction. The array SAR system parameters are shown in Table 1.

[0036] Table 1 Array SAR system parameters

[0037]

[0038] The present invention proposes an array SAR mountain elevation defuzzification method based on an external DEM, which accurately identifies fuzzy point clouds and solves the fuzzy cycle number to solve the problem of elevation fuzziness in mountain point clouds and achieve high-precision three-dimensional reconstruction in mountainous areas. The implementation steps of the present invention are as follows:

[0039] Step 1, obtain three-dimensional point cloud data of mountain area through array SAR; wherein array SAR obtains three-dimensional point cloud data by focusing the signal in the tomographic direction of mountain area. Array SAR technology is a further development of traditional InSAR technology. By adding multiple antennas in the cross-track direction, an array antenna structure is formed, which has the ability of height resolution, thereby achieving the three-dimensional imaging effect of the target scene. In this embodiment, array SAR uses an elevation super-resolution algorithm to generate a three-dimensional point cloud of the target area. Specifically, the elevation super-resolution algorithm includes compressed sensing (CS) and multiple signal classification (MUSIC).

[0040] Step 2: Calculate the unambiguous elevation of the target area; the specific expression of the unambiguous elevation of the array SAR system is: Among them, Δ h is the unambiguous height, λ is the wavelength of the transmitted signal, H is the height of the carrier, y is the scene reference ground height, θ is the downward viewing angle, Δb and β are the baseline interval and baseline horizontal inclination respectively. In practical engineering applications, the commonly used indicator is the array SAR unambiguous height, and its specific expression is Array SAR antenna array structure diagram Figure 2 As shown. It can be seen from the above formula that when the signal wavelength, carrier height, baseline horizontal inclination and viewing angle of the observation scene are known, the baseline interval without elevation ambiguity can be reversely obtained according to the scene reference flat ground height. That is, for low targets such as buildings, the height of the scene can be known a priori, and the elevation ambiguity problem can be solved by appropriate system design. However, the mountainous terrain is undulating, and it is difficult to know the elevation span a priori. In addition, the unambiguous elevation and elevation resolution of the array tomography SAR system restrict each other. While pursuing the unambiguous elevation index, the elevation resolution will inevitably be sacrificed. Therefore, limited by system factors and scene characteristics, the elevation ambiguity problem of point clouds in highly undulating mountainous scenes is difficult to avoid.

[0041] Figure 3 The unambiguous height of the array SAR with the incident angle and scene height as variables is shown. Because the mountainous terrain is undulating and changeable, the scene height here is the reference flat ground height. The system parameters are shown in Table 1. Figure 3 It can be seen that as the scene height increases, the unblurred height of the system decreases.

[0042] Step 3, introduce the external DEM corresponding to the target area; the external DEM corresponding to the target area introduced in this embodiment has a plane resolution of 30m×30m, a relative elevation accuracy of 6m, and an absolute elevation accuracy of 16m. Figure 4 , Figure 5 They are the target area array SAR point cloud result map and the corresponding external source DEM map.

[0043] Step 4, calculate the height difference h1 between the target area and the external DEM;

[0044] Step 5, lower the target area to an unambiguous elevation and calculate the height difference h2 between it and the external source DEM;

[0045] Step 6: Raise the target area to an unambiguous elevation and calculate the height difference h3 between it and the external source DEM;

[0046] Step 7, determining the point cloud of a blurred period according to the height relationship among h1, h2 and h3;

[0047] Among them, if h2 of a certain point Figure 6 As shown, the blue point cloud is the unblurred points, and the red point cloud is the blurred points that are blurred for one cycle.

[0048] Step 8, repeat steps 4 to 7 to find a point cloud with multiple blurred periods; for blurred points with multiple blurred periods, repeat step 8 until there are no blurred points.

[0049] Step 9: Move the blurred point cloud to the real location to achieve elevation deblurring of the mountain point cloud.

[0050] Among them, moving the fuzzy point to the real position means lowering the fuzzy point that has been blurred upward for N (N≥1) cycles by N unfuzzy heights, and raising the fuzzy point that has been blurred downward for N cycles by N unfuzzy heights. Figure 7 As shown, the blurred point that has been blurred upward for one cycle is lowered to an unblurred height. Figure 8 This is the result map of elevation defuzzification in mountainous area, where color represents height. The fuzzy points are moved to the real positions, thus solving the elevation fuzzy problem.

[0051] ​In summary, the above are only preferred embodiments of the present invention and are not intended to limit the protection scope of the present invention. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present invention should be included in the protection scope of the present invention.

Claims

1. A method for defuzzifying the elevation of point clouds in mountainous areas using array SAR based on external DEM. It is characterized in that The following steps are involved: Step 1, obtaining three-dimensional point cloud data of mountainous areas through array SAR; Step 2, calculating the unambiguous elevation of the target area; Step 3, introduce the external DEM corresponding to the target area; Step 4: Calculate the height difference h between the target area and the external DEM 1 ; Step 5: Lower the target area to an unambiguous elevation and calculate the height difference h between it and the external DEM. 2 ; Step 6: Raise the target area to an unambiguous elevation and calculate the height difference h between it and the external DEM. 3 ;, Step 7, according to h 1 、h 2 and h 3 The height relationship between them determines the point cloud that blurs a period; Among them, if h at a certain point 2 <h 1 , then the point is a fuzzy point and is blurred upward for one cycle. If the h of a point 3 <h 1 , then the point is a fuzzy point and is blurred downward by one cycle; Step 8, repeating steps 4 to 7 to find a point cloud with multiple blurred periods; for blurred points with multiple blurred periods, repeat step 8 until there are no blurred points; Step 9, move the blurred point cloud to the real location to achieve elevation deblurring of the mountain point cloud; In step 1, the array SAR uses an elevation super-resolution algorithm to generate a three-dimensional point cloud of the target area; The specific expression of the unambiguous elevation of the array SAR system in step 2 is: Among them, Δ h is the unambiguous elevation, λ is the wavelength of the transmitted signal, H is the height of the carrier aircraft, y is the scene reference ground height, θ is the downward viewing angle, Δb and β are the baseline interval and baseline horizontal inclination respectively.

2. According to the method of claim 1, the elevation super-resolution algorithm includes compressed sensing and multiple signal classification.

3. The method according to claim 1, It is characterized in that In practical engineering applications, the indicator used is the unambiguous height of the array SAR, and its specific expression is:

4. The method according to any one of claims 1 to 3, It is characterized in that In step 9, the blurred points are moved to the real positions by lowering the blurred points blurred upward for N cycles by N unambiguous heights, and raising the blurred points blurred downward for N cycles by N unambiguous heights.

Images