A method for extracting the height of vertical objects in SAR images

Abstract

The present invention discloses a method for extracting the height of an upright feature on a SAR image. After calculating a two-dimensional vector product based on the image coordinates of the upright feature and the position and velocity data of a SAR platform, the height value of the upright feature can be directly calculated. The present invention does not need to use the shadow or "overlapping" feature of the upright feature, nor does it need to construct an image pair using two images. The height calculation of the upright feature can be completed using only one SAR image, and the calculation process is direct and clear without any iteration or approximation process. Therefore, the present invention can be simply and conveniently applied to extracting the height information of various upright features on SAR images.

Description

Technical Field

[0001] The invention belongs to the technical field of SAR image processing, and in particular relates to a method for extracting the height of vertical objects on a SAR image. Background Art

[0002] Acquiring images based on synthetic aperture radar (SAR) has the advantages of all-day and all-weather, large ground coverage, and high resolution, which makes SAR images an important data source for remote sensing applications and has been widely used. In SAR image applications, it is often necessary to analyze and extract certain thematic information of interest, and vertical objects on SAR images are a special type of thematic vertical objects, such as various types of industrial chimneys, various tower-type structures, transmission poles, high-rise buildings, etc. These vertical objects usually stand vertically above the ground, and their height is much greater than their length or width. When using SAR images to analyze such vertical objects, it is often necessary to extract their height values.

[0003] At present, the methods for extracting the height of vertical objects from SAR images mainly include the extraction method based on a single SAR image and the extraction method based on two SAR images.

[0004] The method of extracting the height of vertical objects based on a single SAR image mainly utilizes the special information formed by vertical objects on SAR images, that is, "overlap" or shadow information. Vertical objects will show "overlap" and shadow structures in high-resolution SAR images. If the SAR incident wave encounters the dihedral structure formed by the vertical object and the ground, an obvious bright line structure will be formed on the SAR image, which is called the "overlap" structure. The height of the vertical object can be calculated using the extended distance of the "overlap" structure along the SAR incident direction.

[0005] The "overlapping" structure of vertical objects in high-resolution SAR images can be regarded as the forward mapping of the vertical objects, and its range width is related to the height of the vertical objects. The edge of the shadow structure is the projection of the top edge of the vertical object. The position of the shadow edge facing the SAR reflects the position of the bottom of the vertical object, and the length of the shadow along the range direction reflects the height of the vertical object. Therefore, the height of the vertical object can be calculated according to the projection formula. Therefore, the height of the vertical object can be solved by using the simple trigonometric function relationship between the "overlapping" or range width of the shadow structure of vertical objects in high-resolution SAR images and the height of the vertical object.

[0006] The methods based on two SAR images mainly include stereo SAR and interferometric SAR (InSAR). Among them, the stereo SAR method is to construct a stereo SAR image pair through two SAR images at different angles, and use the matching of the same-name points of the vertical objects and apply the distance equation and Doppler equation to solve the height of the vertical objects. Constructing stereo SAR generally requires a large viewing angle difference, and the sensitivity of SAR scattering characteristics to angles makes it difficult to break through the dense matching technology of multi-angle SAR images in complex scene areas such as urban areas. At present, commercial software such as PCI, ERDAS, and Vexcel all include SAR stereo measurement modules, and stereo SAR has been used in North America and South America to obtain hundreds of thousands of kilometers of elevation information. The interferometric SAR method uses two images with different observation angles and satisfying coherence to extract the height of vertical objects. The small viewing angle difference ensures the image registration performance. At the same time, the interference processing makes the pixel slant range difference accuracy reach an order of magnitude smaller than the wavelength, so that high-precision and continuous ground elevation information can be obtained. At present, interferometric SAR has been widely used in the field of terrain mapping. However, due to the noise and 360-degree entanglement of the interferometric phase, in urban areas with densely populated buildings, the phase changes suddenly and disorderly, making it difficult to solve the phase entanglement. Therefore, InSAR can usually only obtain the height of the surface, and cannot effectively realize the height extraction of vertical objects in complex scenes such as urban areas.

[0007] It can be seen that the method of extracting the height of vertical objects based on a single SAR image is essentially based on the shadow characteristics formed by the vertical objects blocking the sun due to their height above the ground or the "overlap" structure formed by the vertical objects and the ground. The height of the vertical objects is calculated through the geometric function relationship between the "overlap" or shadow structure formed by the vertical objects on the SAR image and the height of the vertical objects. Therefore, to obtain the height of the vertical objects, there must be "overlap" or shadow on the SAR image. Once there is no shadow or "overlap" on the SAR image, this method cannot extract the height of the vertical objects. In some cases, it is difficult to obtain a clear "overlap" or shadow of the vertical objects in the actual SAR images due to various reasons. At this time, the height of the vertical objects cannot be extracted based on the "overlap" or shadow. The method of extracting the height of upright objects based on SAR stereo and interferometric SAR requires extracting the height of the ground by obtaining two image pairs, that is, it is necessary to obtain two images at two locations, and calculate the height of the upright object points by extracting the same-name upright object points on the two images. However, when there is only one satellite CCD remote sensing image, the height information of the upright objects cannot be extracted by the stereo method.

[0008] In summary, the existing methods for extracting the height of vertical objects based on a single SAR image have the problem of relying on the shadow or "overlap" of the vertical objects, while the methods for extracting the height of vertical objects based on SAR stereo and interferometric SAR methods have the disadvantage of requiring two images. Summary of the invention

[0009] In view of this, the present invention provides a method for extracting the height of a vertical object on a SAR image, which can directly calculate the height value of the vertical object according to the image coordinates of the vertical object and the position and velocity data of the SAR platform.

[0010] The present invention provides a method for extracting the height of vertical objects on a SAR image, comprising the following steps:

[0011] Extracting the base image coordinates and the vertex image coordinates of the vertical object to be analyzed on the SAR image, and calculating the three-dimensional position coordinates of the base image and the vertex image of the vertical object to be analyzed corresponding to the SAR flight platform according to the base image coordinates and the vertex image coordinates, as well as the Doppler parameters corresponding to the base image and the vertex image respectively;

[0012] Calculating a two-dimensional vector product according to the three-dimensional position coordinates and the speed of the flight platform; calculating a Doppler difference index according to the base image coordinates and the vertex image coordinates, Doppler parameters and the properties of the flight platform;

[0013] The height of the vertical object to be analyzed is calculated based on the two-dimensional vector product, the Doppler difference index and the speed of the flying platform.

[0014] Furthermore, the method of extracting the bottom image coordinates and the top image coordinates of the vertical object to be analyzed on the SAR image is to analyze the SAR image using image analysis software to obtain the bottom image coordinates and the top image coordinates.

[0015] Furthermore, the method of calculating the three-dimensional position coordinates of the flight platform corresponding to the base image and the vertex image of the upright object to be analyzed based on the base image coordinates and the vertex image coordinates is: obtaining the three-dimensional position coordinates by interpolating the base image coordinates and the vertex image coordinates.

[0016] Furthermore, the two-dimensional vector product K is calculated using the following formula:

[0017] K=V X (X S2 -X S1 )+V Y (Y S2 -Y S1 )

[0018] Among them, X S1 and Y S1 are the coordinate values ​​of the X-axis and Y-axis of the flight platform corresponding to the bottom point image of the vertical object to be analyzed; S2 and Y S2V are the coordinate values ​​of the X-axis and Y-axis of the flight platform corresponding to the vertex image of the vertical object to be analyzed; X and V Y They are the speed values ​​of the flight platform in the X-axis and Y-axis directions respectively.

[0019] Furthermore, the Doppler difference index F is calculated using the following formula:

[0020]

[0021] Among them, M X is the slant range resolution of SAR, r0 is the perigee slant range of SAR, λ is the wavelength of SAR, and f D1 is the Doppler parameter corresponding to the bottom image of the vertical object to be analyzed, f D2 is the Doppler parameter corresponding to the vertex image of the vertical object to be analyzed, x1 is the column coordinate position of the base image of the vertical object to be analyzed, and x2 is the column coordinate position of the vertex image of the vertical object to be analyzed.

[0022] Furthermore, the height h of the vertical object to be analyzed is calculated using the following formula:

[0023]

[0024] Wherein, K is the two-dimensional vector product, F is the Doppler difference index, V Z It is the speed of the flight platform in the Z-axis direction.

[0025] Beneficial effects:

[0026] The present invention can directly calculate the height value of the upright feature after calculating the two-dimensional vector product according to the image coordinates of the upright feature and the position and velocity data of the SAR platform. The present invention does not need to use the shadow or "overlapping" feature of the upright feature, nor does it need two images to construct an image pair. The height calculation of the upright feature can be completed using only one SAR image, and the calculation process is direct and clear without any iteration and approximation process. Therefore, the present invention can be simply and conveniently applied to extracting height information of various upright features on SAR images. BRIEF DESCRIPTION OF THE DRAWINGS

[0027] Figure 1 The present invention provides a flow chart of a method for extracting the height of vertical objects on a SAR image. DETAILED DESCRIPTION

[0028] The present invention is described in detail below with reference to the accompanying drawings and embodiments.

[0029] The present invention provides a method for extracting the height of a vertical object on a SAR image. The core idea is to use a geometric relationship model of a SAR flight platform for rigorous imaging, and adopt the position information of the vertical object to be analyzed extracted from a SAR image containing the vertical object to be analyzed, the speed information of the flight platform, and the attribute information of the flight platform to directly calculate the height of the vertical object to be analyzed.

[0030] The present invention provides a method for extracting the height of vertical objects on a SAR image, the process is as follows: Figure 1 As shown, the specific steps include:

[0031] Step 1: Use image analysis software to extract the bottom image coordinates (x1, y1) and top image coordinates (x2, y2) of the vertical feature to be analyzed contained in the SAR image, and read the data in the flight platform auxiliary data file for acquiring the SAR image, and interpolate the data to obtain the three-dimensional position coordinates of the flight platform corresponding to the bottom image of the vertical feature to be analyzed (X1, y1). S1 ,Y S1 ,Z S1 ), the three-dimensional position coordinates of the flight platform corresponding to the vertex image are (X S2 ,Y S2 ,Z S2 ) and the three-dimensional velocity of the platform (V X ,V Y ,V Z ), and the Doppler parameter f corresponding to the bottom point image of the vertical object to be analyzed D1 And the Doppler parameter f corresponding to the vertex image D2 .

[0032] Among them, x1 and y1 are the column and horizontal coordinate positions of the bottom image of the vertical object to be analyzed respectively; x2 and y2 are the column and horizontal coordinate positions of the vertex image of the vertical object to be analyzed respectively; S1 , Y S1 and Z S1 are the coordinate values ​​of the X-axis, Y-axis and Z-axis of the flight platform corresponding to the bottom point image of the vertical object to be analyzed; S2 , Y S2 and Z S2 are the coordinate values ​​of the X-axis, Y-axis and Z-axis of the flight platform corresponding to the vertex image of the vertical object to be analyzed; V X 、V Y and V Z They are the speed values ​​of the flight platform in the X-axis, Y-axis and Z-axis directions respectively.

[0033] Step 2: Calculate the two-dimensional vector product K based on the three-dimensional position of the flight platform of the vertical feature to be analyzed and the speed of the platform obtained in step 1. The specific calculation formula is shown in formula (1):

[0034] K=V X (X S2 -X S1 )+V Y (Y S2 -Y S1 ) (1)

[0035] According to the position coordinates and Doppler parameters of the vertical object to be analyzed in the SAR image, the Doppler difference index F is calculated using formula (2):

[0036]

[0037] Among them, M X is the slant range resolution of the SAR image, r0 is the perigee slant range of the SAR image, and λ is the wavelength of the SAR. The above parameters are all known parameters of the SAR and can be directly read from the SAR auxiliary data.

[0038] Step 3: According to the two-dimensional vector product and Doppler difference index calculated in step 2 and the velocity value of the flight platform in the Z-axis direction obtained in step 1, the height h of the vertical object to be analyzed is calculated using formula (3). Formula (3) is as follows:

[0039]

[0040] It can be seen that the method for extracting the height of upright objects on a SAR image provided by the present invention strictly applies the strict imaging geometric relationship model of SAR images. The geometric relationship is strict, and there is no need to extract other features of the upright objects, nor to use two SAR images. There is no need to set initial values ​​and iterations in the calculation, and the height of the upright objects can be directly calculated.

[0041] 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 extracting the height of vertical objects on a SAR image, characterized in that: The following steps are involved: Extracting the base image coordinates and the vertex image coordinates of the vertical object to be analyzed on the SAR image, and calculating the three-dimensional position coordinates of the base image and the vertex image of the vertical object to be analyzed corresponding to the SAR flight platform according to the base image coordinates and the vertex image coordinates, as well as the Doppler parameters corresponding to the base image and the vertex image respectively; Calculating a two-dimensional vector product according to the three-dimensional position coordinates and the speed of the flight platform; calculating a Doppler difference index according to the base image coordinates and the vertex image coordinates, Doppler parameters and the properties of the flight platform; The height of the vertical object to be analyzed is calculated based on the two-dimensional vector product, the Doppler difference index and the speed of the flight platform; The two-dimensional vector product K is calculated using the following formula: K=V X (X S2 -X S1 )+V Y (Y S2 -Y S1 ) Among them, X S1 and Y S1 are the coordinate values ​​of the X-axis and Y-axis of the flight platform corresponding to the bottom point image of the vertical object to be analyzed; S2 and Y S2 V are the coordinate values ​​of the X-axis and Y-axis of the flight platform corresponding to the vertex image of the vertical object to be analyzed; X and V Y They are the speed values ​​of the flight platform in the X-axis and Y-axis directions respectively; The Doppler difference index F is calculated using the following formula: Among them, M X is the slant range resolution of SAR, r0 is the perigee slant range of SAR, λ is the wavelength of SAR, and f D1 is the Doppler parameter corresponding to the bottom image of the vertical object to be analyzed, f D2 is the Doppler parameter corresponding to the vertex image of the vertical object to be analyzed, x1 is the column coordinate position of the base image of the vertical object to be analyzed, and x2 is the column coordinate position of the vertex image of the vertical object to be analyzed; It is characterized in that the height h of the vertical object to be analyzed is calculated using the following formula: Wherein, K is the two-dimensional vector product, F is the Doppler difference index, V Z It is the speed of the flight platform in the Z-axis direction.

2. The extraction method according to claim 1, characterized in that The method of extracting the bottom image coordinates and the top image coordinates of the vertical object to be analyzed on the SAR image is to use image analysis software to analyze the SAR image to obtain the bottom image coordinates and the top image coordinates.

3. The extraction method according to claim 1, characterized in that The method of calculating the three-dimensional position coordinates of the flight platform corresponding to the base image and the vertex image of the upright object to be analyzed based on the base image coordinates and the vertex image coordinates is: obtaining the three-dimensional position coordinates by interpolating the base image coordinates and the vertex image coordinates.

Images