KR20120009186A - method for manufacturing a digital elevation model using a SAR data - Google Patents

Abstract

PURPOSE: A digital elevation model manufacturing method using SAR(Synthetic Aperture Radar) data is provided to simplify a complex processing progress by selecting a target area for using the SAR data. CONSTITUTION: Main SAR data and sub SAR data are matched(S102). Sub images are rearranged by using a matching result. A phase difference between the main SAR data and the sub SAR data is calculated. Interference is filtered by using the calculated interference(S103). DEM(Digital Elevation Model) is manufactured in order to extract a displacement index through a phase unwrapping method which changes phase information into an elevation value(S104, S105).

Description

Method for manufacturing a digital elevation model using a SAR data

The present invention relates to a method for manufacturing a numerical value model using SAR data, and in particular, by using energy generated by the sensor itself to receive energy reflected from the ground surface and using SAR data to obtain observation data of the ground surface. It is about how to make a model.

In general, numerical values representing the elevation of the earth's surface can be obtained by various methods, but there are advantages and disadvantages for each method. The most common method is to use the contour line in the digital topographic map, and there are other methods such as total station, GPS field survey, and aerial survey.

The method of using a digital topographic map has a disadvantage in that it cannot include information on the changes that have occurred since the digital topographic map was created due to the time difference between the time of the digital map production and the present time. It is not suitable for producing high level models. Aerial photographic surveys that acquire optical images have disadvantages such as a small imaging area, complicated processing, professional manpower, cost, and time. In addition, optical images may be deteriorated depending on the timing of shooting and the weather, and are subject to time constraints.

SAR interferometry, which is a radar interferometry method, can extract elevation data from radar images, which can overcome the shortcomings of optical images, and accuracy is also increasing due to the development of technology. It is expected to increase.

In addition, satellite remote sensing has the advantage of acquiring information of the target area from a long distance, and this advantage is currently used in various fields such as mapping, disaster area monitoring, terrain change detection, and military purpose. Sensors used in remote sensing are data acquisition equipment mounted on satellites or aircrafts and are classified into passive sensors and active sensors.

Passive sensor senses the light energy reflected from the earth surface using external energy such as the sun, and means an optical image sensor that mainly acquires visible and infrared images. An active sensor receives energy reflected from the earth's surface using energy generated by the sensor itself, and mainly acquires observation data on the earth's surface using microwave waves.

The SAR (Synthetic Aperture Radar) data acquired by the active sensor is also called image radar or synthetic aperture radar data, and the earth observation data can be acquired in an active manner by using the radar principle. SAR has the advantage of being able to acquire a wide range of images in harsh environments such as day and night, backlight, clouds, and bad weather, and can acquire images in case of disaster such as flood, typhoon, tsunami, oil spill, earthquake, volcanic eruption, and military As it can monitor the deployment of weapons and movements of the enemy nations, it is expected that the technological developments in the field of use of earth observation data will be innovative.

However, the SAR data has a problem that the processing is complicated because the resolution of the image is low, visual readability is difficult, and the geometric configuration is complicated, compared with the data obtained by the conventional optical sensor.

The present invention has been invented to solve the above problems, using a Single Look Complex (SLC) that has undergone initial signal processing in raw data, selects a processing area in the main and sub-images, and compensates for the baseline and cross-coordinates. Through registration, the coherence calculation is performed, the coupling calculation is performed, the interference is calculated, the interference pattern is generated, and the phase information is converted to the elevation through phase unwrapping. The purpose of this paper is to provide a method for producing a numerical model using SAR data that increases the resolution of images and simplifies the process.

The present invention for performing such an object,

A matching process between two SAR data for correcting the earth curvature phase difference between the primary SAR data and the secondary SAR data located at pixels of different coordinates by observing through different orbits;

An interference calculation step of rearranging the sub-images using the matching result to calculate an interference degree by calculating a phase difference between two SAR data;

Azimuth filtering to calibrate to have the effect observed with the Doppler center frequency to increase the signal-to-noise ratio (SNR) of the interference using the calculated interference, and to correct the difference of the center frequency in the spectrum of the alignment direction of the two SAR data An interference filtering process for performing post-processing filtering to remove a large amount of noise remaining in the interference after the alignment curvature and the earth curvature phase are corrected; And,

It includes the step of creating a high level model to extract the high level data (DEM) and the surface displacement through phase unwrapping, which converts the phase information into an elevation value.

The method for producing a numerical value model using the SAR image according to the present invention is to increase the resolution of the image and simplify the processing process by selecting a target region using SAR data and applying a system to produce a complex value model. Its effectiveness is proven to improve the process, manpower, costs and time-consuming shortcomings.

1 is a flow chart showing a method for producing a numerical value model using SAR data.
2 is a view for showing a general geometrical configuration of a data acquisition method of SAR interference technology.
3 is a view showing a main SAR image in the method for producing a numerical value model using SAR data according to the present invention.
4 is a view showing a secondary SAR image in the method for producing a numerical value model using SAR data according to the present invention.
5 is a diagram showing an interference image in a method for manufacturing a numerical value model using SAR data according to the present invention.
Figure 6 is a numerical model produced by performing a method for manufacturing a numerical value model using SAR data according to the present invention.
FIG. 7 is a three-dimensional visualization of the numerical model of FIG. 6.

The interferometry SAR of the present invention is a technique for measuring topographic elevation and displacement by using phase information through two or more SAR data obtained at the same point, and differential radar interferometry is wider than tens of kilometers. It is possible to measure the displacement of the terrain by measuring sensitivity of several centimeters with spatial resolution of several tens of meters.

As shown in FIG. 1, a series of data processing procedures for extracting high-level data (DEM) and surface displacement using a radar interference technique from a SAR image according to the present invention is performed. Through the process of interference filtering and phase unwrapping, DEM and surface displacement are extracted.

Satellite radar interferometry uses SAR data from different orbits, so that the same scatterers are located at different coordinates in the image.

In order to correct the difference between the data, a matching process between the primary SAR image and the secondary SAR image as shown in FIGS. 3 and 4 is required.

 In the SAR data acquisition method, interference phases occur when two antennas use data observed from different planes or tracks. The interference phase is related to the difference in distance from two or two SAR observations of the same scatterer. The relationship between the interference phase, topography, and the ellipsoid is interconnected through the look angle as follows. First, the interference phase is related to the baseline and the observation angle, which in turn is related to the geospatial and topographical altitude. The general geometrical configuration of the data acquisition method of the SAR interference technology is shown in FIG.

와

로 놓았을 때, 두 안테나의 위상차이는 식 1과 같다.Referring to Figure 2, the distance from the two antennas to the ground

Wow

When set to, the phase difference between the two antennas is given by Equation 1.

(식 1)

(Equation 1)

는

센서의 파장이다. 는 궤도의 기하학적 구성에 따라 달라질 수 있으며, 식 1은 코사인 제 2법칙에 의해서 식 2와 같이 표현될 수 있다.here

Is

The wavelength of the sensor. May vary according to the geometry of the trajectory, and Equation 1 may be expressed as Equation 2 by the second law of cosine.

(식 2)

(Equation 2)

는 관측각이고,

를 안테나의 수평인 직선상에서 기선과 이루는 각이다. 식 2는 다음과 같이 요약될 수 있다.In Equation 2, B is the length of the baseline,

Is the viewing angle,

Is the angle formed with the base line on the horizontal straight line of the antenna. Equation 2 can be summarized as follows.

(식 3)

(Equation 3)

이므로 무시될 수 있다. 또한 식 3에서

은 기선거리의 수평성분

이므로At this time

Can be ignored. Also in equation 3

Is the horizontal component of the baseline distance

Because of

(식 4)

(Equation 4)

It can be expressed as Equation 4. Substituting Equation 4 back into Equation 1, the phase of the interference can be expressed as follows.

(식 5)

(Equation 5)

When radar interference processing is performed using SAR data, an interference degree is formed as shown in FIG. 5, and the interference degree produced at this time is mainly referred to as terrain interference since the phase difference caused by the terrain is the largest. The error that can occur in the formation of topographical interference is due to the distance between the satellite and the target (scatter), the distance between the satellite, the satellite altitude, and the phase difference. In radar interference, the phase assumes accurate satellite orbit geometry and ideal signal acquisition conditions, but the actual data do not meet the ideal conditions, which can lead to unexpected phase errors.

From radar geometries, the interference phase is influenced not only by the topography but also by the ellipsoid. The length and angle of the baseline (ie the orbit information) must be known accurately, and any residual phases due to satellite orbit errors should be removed by additional orbital correction. The phase due to the ellipsoid can be expressed by the following equation.

(식 6)

(Equation 6)

Since the interference phase from which the geospheroidal phase is removed is the difference between Equations 5 and 6, it can be briefly expressed as shown in Equation 8.

(식7)

(Eq. 7)

(식8)

(Eq. 8)

는 영상의 각 지점에서 지구타원체까지의 관측각이며,

이다.From the stomach

Is the angle of observation from each point in the image to the geospheroid,

to be.

를 구할 수 있다.In the present invention, the phase correction by the ellipsoid does not use the baseline length between satellites, but directly uses the orbital information of two satellites. Coordinates (x, y, z) of satellite orbits of master and slab corresponding to specific points (line, pixel) in the main image and coordinates (x, y, z) of points on the reference ellipsoid It is obtained using three equations (geospheroidal equation, range equation, Doppler equation). As shown in Fig. 2, the positions of the two satellites and the points on the ellipsoid are plotted on the ellipsoid using the following formula.

Can be obtained.

(식 9)

(Eq. 9)

를 구할 수 있다. 실제로 지구타원체에 의한 위상 변화는 매우 선형적이므로, 영상에 고루 분포하는 임의의 수백~수천 점을 이용하여 지구타원체 위상을 구한 후 2차원 다항식에 최적화(2D polynomial fitting)를 실시하여 영상 전체에서의 위상을 계산하는 방법을 구현할 수 있다.Substituting this into Equation 1, the phase by the earth curvature

Can be obtained. In fact, the phase change caused by the ellipsoid is very linear. Therefore, the geospheroidal phase is obtained by using arbitrary hundreds to thousands of points distributed evenly in the image, and then optimized by 2D polynomial fitting (2D polynomial fitting). Can be implemented to calculate

), 첫 번째 range 획득 시간(

)등 및 궤도정보가 요구되므로 리더파일 등으로부터 해당정보를 추출해야한다. 시뮬레이션을 위해 SAR 영상으로부터 선택된 지역의 좌표는 지리좌표계로 변환되어 입력된 DEM으로부터 해당영역을 추출한다.In order to simulate SAR images using the digital elevation model (DEM), satellite image acquisition coefficients (PRF, sensor wavelength, first azimuth acquisition time)

), First range acquisition time (

And orbital information is required, so the relevant information must be extracted from the reader file. The coordinates of the area selected from the SAR image for the simulation are converted to the geographic coordinate system to extract the area from the input DEM.

DHK 궤도에 놓여 있는 위성

사이의 방정식은 다음과 같다.Point on ellipsoid

Satellite in DHK orbit

The equation between

(식 10)

(Eq. 10)

)에 대한 미분은 다음과 같다.The exported numerical model (DEM) should be interpolated using bilinear and bicubic methods to fully correspond to the pixel spacing of the actual SAR image. The interpolated altitude values are converted to (x, y, z) in the geocentric Cartesian coordinate system using the coordinates of the DEM, and (x, y, z) values are converted to the radar coordinate system. To convert to a geographic coordinate system, only the Doppler equation can be obtained from the previous three equations. Azimuth time of the Doppler equation (

The derivative for) is

(식 11)

(Eq. 11)

는 위성의 가속도 벡터이다. 따라서 다음 Newton - Raphson 반복법을 이용하여 (

)을 구할 수 있다.here

Is the acceleration vector of the satellite. Therefore, using the Newton-Raphson iteration,

) Can be obtained.

(식 12)

(Eq. 12)

를 구하면 결국 위성의 위치가

가 결정되므로 다음 식을 이용하여 range 시간(

)을 구할 수 있다.

To get the position of the satellite

Is determined, so the range time (

) Can be obtained.

(식 13)

(Eq. 13)

와

은

, PRF, range sampling rate를 이용하여 레이더 좌표계에서의 line과 pixel로 변환된다. SAR영상을 시뮬레이션하기 위해서는

,

뿐만 아니라

지점에서 지표면의 수직 벡터를 구해야한다. 수직벡터

은 주어진 점의 수치고도모형(DEM)에서 상하좌우에 위치한 4점의

를 구한 후 다음의 벡터연산을 이용해 구한다.

Wow

silver

The signal is converted into lines and pixels in the radar coordinate system using the PRF and the range sampling rate. To simulate a SAR image

,

As well as

We need to find the vertical vector of the earth's surface at a point. Vertical Vector

Is the value of 4 points on top, bottom, left and right

After obtaining, use the following vector operation.

(식 14)

(Eq. 14)

(식 15)

(Eq. 15)

과

사이의 각

를 구한 후 다음 식을 이용해 구할 수 있다.Video brightness

and

Angle between

After we find, we can find it by

(식 16)

(Eq. 16)

(식 17)

(Eq. 17)

과

를 구해, 다음 식으로부터 지형과 지구타원체에 의한 위상을 포함한 간섭위상

값을 구한다.Two kinds of interference simulations are used, which have only phase information by terrain and both phase and earth curvature. Position vector of satellite in main and sub image

and

Find the interfering phase including the topography and the phase by the ellipsoid from the equation

Find the value.

(식 18)

(Eq. 18)

(식 19)

(Eq. 19)

(식 20)

(Eq. 20)

Removing the geospheroidal phase from Eq. 20 gives Eq.

(식 21)

(Eq. 21)

이므로

는 다음과 같이 정리된다.In Equation 21

Because of

Is summarized as follows.

(식 22)

(Eq. 22)

(식 23)

(Eq. 23)

는 주영상의 레이더좌표계 (i,j)지점의 고도가 0일 때의 지구타원체 상의 위치벡터이며,

는 식 22를 이용하여 구한 것으로 부영상 위성이

지점을 관측할 때의 위성 위치벡터이다.In Equation 23

Is the position vector on the ellipsoid when the altitude of the radar coordinate system (i, j) of the main image is 0.

Is obtained using Equation 22.

Satellite position vector when observing a point.

As shown in FIGS. 6 and 7, the numerical modeling method using SAR data performed by the above method has a higher resolution and easier visual reading than the data acquired by the conventional optical sensor. The process is not complicated.

Claims (1)

A matching process between the two SAR data for correcting the earth curvature phase difference between the primary SAR data and the secondary SAR data located at pixels of different coordinates by observing while passing through different orbits;
An interference calculation step of rearranging the sub-images using the matching result in the step to calculate an interference degree by calculating a phase difference between the two SAR data;
Azimuth filtering for correcting to have an effect observed at the Doppler center frequency in order to increase the signal-to-noise ratio (SNR) of the interference using the calculated interference, and the difference of the center frequency in the spectrum of the alignment direction of the two SAR data An interference filtering process for performing alignment filtering to correct the error and post-processing filtering to remove a large amount of noise remaining in the interference after the earth curvature phase is corrected; And,
By using SAR data including the step of producing a numerical value model for extracting the numerical value data (DEM) and the surface displacement through phase unwrapping that converts the phase information into an elevation value How to make a high school model.

Images