Array SAR point cloud three-dimensional reconstruction method based on multiple scattering of buildings

Abstract

The application provides a kind of array SAR point cloud three-dimensional reconstruction method based on building multiple scattering, can realize the reuse of multiple scattering information, improve three-dimensional reconstruction capability, solve the problem that obvious multiple scattering interference exists in array SAR urban three-dimensional point cloud.The application adopts array SAR technology, by sequentially increasing multiple antennas in the intersection rail, forms array antenna structure, has height resolution capability, so as to achieve the three-dimensional imaging effect of target scene.The application detects scattering points in the shadow of high building, and then performs symmetric displacement about the side surface of high building, not only can avoid multiple scattering interference and effectively suppress multiple scattering interference, but also realizes the reuse of three scattering information, obtains more building scattering information, can make the three-dimensional imaging result of building more delicate.

Description

Technical Field

[0001] This invention relates to the field of synthetic aperture radar signal processing technology, and specifically to a method for three-dimensional reconstruction of array SAR point clouds based on multiple scattering from buildings. Background Technology

[0002] SAR imaging systems possess all-weather, all-day, and penetrating operational characteristics, making them invaluable in remote sensing, topographic mapping, and other fields. With the rapid development of geographic information big data, high-resolution 3D imaging systems are receiving increasing attention. Array SAR systems can acquire not only the elevation information of target scatterers but also their distribution along the elevation axis. They also retain 3D reconstruction capabilities even in overlapping areas formed by complex terrain, enabling the 3D reconstruction of complex urban targets and gaining widespread attention both domestically and internationally.

[0003] However, due to the complex scattering mechanism of array SAR, there are a large number of interference points in array SAR point clouds, which limits the application of array SAR point clouds. Therefore, point cloud post-processing and target extraction have become research hotspots in array SAR signal processing.

[0004] Currently, array tomographic SAR signal processing, especially in complex urban areas, faces significant challenges. Complex urban environments, composed of buildings of different types and shapes, green spaces of varying sizes, infrastructure (roads, bridges, etc.), and complex suburban structures, have long been a difficult problem for remote sensing measurement and analysis.

[0005] Most domestic and international research focuses on height-based imaging algorithms for array SAR three-dimensional imaging technology, mainly including Fourier transform algorithms, spectral estimation algorithms, and compressed sensing algorithms. These methods are three-dimensional imaging algorithms and cannot solve the multiple scattering problem. Current research on multiple scattering primarily focuses on secondary scattering, neglecting tertiary scattering, and is limited to scattering suppression without fully utilizing scattering information. Summary of the Invention

[0006] In view of this, the present invention provides a method for three-dimensional reconstruction of array SAR point clouds based on multiple scattering of buildings, which can realize the reuse of multiple scattering information, improve the three-dimensional reconstruction capability, and solve the problem of obvious multiple scattering interference in the three-dimensional point clouds of urban areas in array SAR.

[0007] To achieve the above objectives, the technical solution of the present invention is as follows:

[0008] This invention discloses a 3D reconstruction method for array SAR point clouds based on multiple scattering of buildings. The method involves acquiring 3D point cloud data of buildings using array SAR; performing planar fitting on the side of the building to locate its position; estimating the building's height and width in the array SAR point cloud; calculating the building's shadow area based on the radar observation viewpoint; detecting multiple scattering points within the calculated shadow area; and then performing symmetrical displacement about the building's side to obtain the building's structural and scattering information; finally, achieving 3D imaging based on the building's structural and scattering information.

[0009] The specific formula for calculating the building shadow area based on the radar observation angle is l=w+h*tanθ, where l is the building shadow area, h and w are the building height and width respectively, and θ is the downward angle.

[0010] In this process, multiple scattering points are detected in the calculated building shadow area, and then symmetrical displacement is performed about the side of the building.

[0011] The process involves modeling the building based on its structural and scattering information, then applying the scattering information to the model to obtain a 3D imaging result.

[0012] Among them, the least squares method is used to perform planar fitting on the side of the building.

[0013] Beneficial effects:

[0014] 1. This invention employs array SAR technology, forming an array antenna structure by sequentially adding multiple antennas along the cross-track direction, which possesses height resolution capability, thereby achieving a three-dimensional imaging effect of the target scene. This invention detects scattering points within the shadow of tall buildings and then performs symmetrical displacement about the side of the tall building. This not only avoids and effectively suppresses multiple scattering interference but also enables the reuse of three-dimensional scattering information, acquiring more scattering information from the building and resulting in more refined three-dimensional imaging results.

[0015] 2. This invention estimates the building height in array SAR point clouds and calculates the shadow range based on the observation geometry; then detects scattering points within the calculated shadow range to accurately obtain the building shadow area and effectively suppress multiple scattering interference.

[0016] 3. This invention obtains a three-dimensional imaging result by modeling the building and attaching the scattering information to the model. The three-dimensional imaging result clearly reflects the structural information of the building and the texture information of the side and roof, making it easier to apply and promote. Attached Figure Description

[0017] Figure 1 This is a schematic diagram of secondary scattering from a building.

[0018] Figure 2This is a schematic diagram of the three-dimensional scattering of a building.

[0019] Figure 3 This is a multiple scattering interference map of an array SAR point cloud.

[0020] Figure 4 This is a schematic diagram of the process of the present invention.

[0021] Figure 5 This is a schematic diagram of the building imaging geometry of the present invention.

[0022] Figure 6 This is a diagram showing the post-processing result of point clouds in this invention.

[0023] Figure 7 This is a three-dimensional imaging result diagram of the present invention. Detailed Implementation

[0024] The present invention will now be described in detail with reference to the accompanying drawings and embodiments.

[0025] Because the wavelength of the array SAR is close to the surface roughness of the building, the building has strong specular scattering of radar waves, resulting in significant scattering interference in the imaging. Figure 1 It is a secondary scattering (ABCD). Figure 2 This is known as triple scattering (ABCBA). Since the array SAR signal model is based on single scattering, multiple scattering from buildings will create interference points in the array SAR point cloud; secondary scattering will create strong angular reflection interference; and triple scattering will create mirror interference. For example... Figure 3 The image shows an actual point cloud from a multi-receiver SAR array. Because the radar observed from the upper left, the lower right area of ​​the building is shaded, therefore no observable scattering bodies exist in that area. However, due to... Figure 3 Two relatively dense point clouds can be observed on the lower right side of the building, one formed by secondary scattering and the other by tertiary scattering. Secondary scattering is mainly formed by the side of the tall building and the roof of the short building; the two clusters of secondary scattering point clouds are symmetrical about the dihedral angle. Tertiary scattering is mainly formed by the side of the tall building and the short building; the point cloud formed by tertiary scattering is a mirror image of the short building with respect to the side of the tall building. It should be noted that tertiary scattering observed the shaded side of the short building, information not observable in single-scattering point clouds. Secondary scattering point clouds exhibit symmetrical distribution characteristics and can be eliminated during signal processing. Tertiary scattering point clouds are mostly located in the shadow of the tall building, and their positions are symmetrical to the actual target with respect to the side of the tall building.

[0026] Based on the above analysis, this invention provides a method for 3D reconstruction of point clouds using array SAR based on multiple scattering from buildings. It detects scattering points within the shadow of tall buildings and then performs symmetrical displacement about the side of the tall building. This method not only avoids interference from multiple scattering but also enables the reuse of triple scattering information. Existing technologies are limited to removing scattering information and do not achieve the reuse of point cloud scattering information. The specific process of this invention is as follows: Figure 4 As shown, it includes the following steps:

[0027] Step 1: Acquire 3D point cloud data of buildings using array SAR;

[0028] Among them, array SAR obtains three-dimensional point cloud data of buildings by focusing the signal in the tomographic direction of the building. Array SAR technology is a further extension of traditional InSAR technology. By adding multiple antennas in the cross-track direction to form an array antenna structure, it has height resolution capability, thereby achieving three-dimensional imaging effect of the target scene.

[0029] Step 2: Perform planar fitting on the side of the building to locate its position; estimate the building's height and width in the array SAR point cloud; calculate the building's shadow area based on the radar observation view, using the following formula:

[0030] l=w+h*tanθ

[0031] Where l represents the building's shadow area, h and w are the building's height and width, respectively, and θ is the downward viewing angle. The geometric schematic diagram of building imaging in this invention is shown below. Figure 5 As shown.

[0032] The least squares method is used to fit the plane of the building's side. Specifically, the least squares plane fitting mainly uses the coordinate values ​​x from the known point cloud data. i y i and z i To fit the plane, using x i z i The function values ​​obtained by fitting Compared with the actual value y i The square of the difference is minimized. Where x i y i , z i represents the point cloud coordinates in the azimuth, range, and altitude directions, respectively. Based on the data point (x... i y i , z i The equation of the plane is denoted as f(x, z) = p. T (x, z)A. Where p T A and B are the basis functions and coefficients of the plane equation, respectively, and can be taken as... From the definition of least squares, we can obtain:

[0033]

[0034] To obtain the result that minimizes the coefficients of the plane equation in J(A), equation (1) can be modified as follows:

[0035]

[0036] To find the solution to equation (2), define Where P is an n×3 matrix, and Y represents all known points y = ... i The coordinates of (2) are given, so (2) can be rewritten in matrix form:

[0037]

[0038] The coefficients A of the fitted plane equation can be obtained from equation (3).

[0039] Step 3: Detect scattering points in the calculated building shadow area;

[0040] Since the point clouds of the third scattering are mostly located in the shadow areas of buildings, the scattering points are detected in the shadow areas of buildings.

[0041] Step 4: Perform symmetrical displacement on the side of the building to obtain the building's structural information and scattering information;

[0042] Since the position of the triple-scattered point cloud is symmetrical to the actual target about the side of the building, a symmetrical displacement is performed about the side of the building. The post-processing results of the array SAR point cloud using the method of this invention are as follows: Figure 6 As shown, after point cloud post-processing, the shadow area on the lower right side of the building is free from multiple scattering interference, while the low building on the front side not only achieved front side reconstruction but also rear side reconstruction in the shaded area, obtaining more accurate building structure information and scattering information.

[0043] Step 5: Based on the building's structural information and scattering information, model the building and apply the scattering information to the model to obtain the three-dimensional imaging result.

[0044] The 3D imaging result obtained by modeling the building and applying scattering information to the model is as follows: Figure 7 As shown, the 3D imaging results clearly reflect the building's structural information and the texture information of its sides and rooftops, making it easier to apply and promote.

[0045] In summary, the above are merely preferred embodiments of the present invention and are not intended to limit the scope of protection of the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the scope of protection of the present invention.

Claims

1. A method for three-dimensional reconstruction of array SAR point clouds based on multiple scattering from buildings, characterized in that, Three-dimensional point cloud data of buildings are acquired using array SAR; Plane fitting is performed on the side of the building to locate its position; the height and width of the building are estimated in the array SAR point cloud, and the shadow area of ​​the building is calculated based on the radar observation view. Multiple scattering points are detected in the calculated building shadow area, and then symmetrical displacement is performed about the side of the building to obtain the building structure information and scattering information. Three-dimensional imaging is achieved based on building structural information and scattering information; The specific formula for calculating the building shadow area based on radar observation perspective is as follows: ,in The area shaded by the building and These are the building's height and width, respectively. This is a bottom-view perspective.

2. The method as described in claim 1, characterized in that, Multiple scattering points are detected in the calculated building shadow area, and then symmetrical displacement is performed about the side of the building.

3. The method as described in claim 2, characterized in that, Based on the building's structural and scattering information, the building is modeled, and the scattering information is applied to the model to obtain a three-dimensional imaging result.

4. The method as described in claim 1 or 3, characterized in that, The least squares method is used to perform planar fitting on the side of the building.

Images