Research Article Using Open-Source Components to Process ...downloads.hindawi.com/journals/ijap/2013/275635.pdf · Using Open-Source Components to Process Interferometric TerraSAR-X

Hindawi Publishing CorporationInternational Journal of Antennas and PropagationVolume 2013 Article ID 275635 13 pageshttpdxdoiorg1011552013275635

Research ArticleUsing Open-Source Components to Process InterferometricTerraSAR-X Spotlight Data

Michael Jendryke12 Timo Balz1 Houjun Jiang1 Mingsheng Liao1 and Uwe Stilla2

1 State Key Laboratory of Information Engineering in Surveying Mapping and Remote Sensing Wuhan UniversityWuhan Hubei 430079 China

2 Photogrammetry and Remote Sensing Technische Universitaet Muenchen 80333 Munich Germany

Correspondence should be addressed to Michael Jendryke michaeljendrykerubde

Received 21 November 2012 Accepted 20 January 2013

Academic Editor Gui Gao

Copyright copy 2013 Michael Jendryke et al This is an open access article distributed under the Creative Commons AttributionLicense which permits unrestricted use distribution and reproduction in any medium provided the original work is properlycited

We address the processing of interferometric TerraSAR-X and TanDEM-X spotlight data Processing steps necessary to deriveinterferograms at high spatial resolution from bi- and monostatic satellite images will be explained The spotlight image mode is abeam steering technique focusing the antenna on a specific ground areaThis results in a linear Doppler shift frequency in azimuthdirection which has to be matched to the master image While shifting the interpolation kernel in azimuth during resamplingthe frequency spectrum of the slave image is aligned to the master image We show how to process bistatic TanDEM-X imagesand propose an integrated processing option for monostatic TerraSAR-X data in the Delft Object-oriented Radar InterferometricSoftware (DORIS) The paper focuses on the implementation of this algorithm for high-resolution spotlight InSAR in a publicdomain tool hence it becomes available to a larger research community The results are presented for three test areas Uluru inAustralia Las Vegas in the USA and Luneburg in Germany

1 Introduction

Radar remote sensing with SAR (Synthetic Aperture Radar)satellites is used to observe the Earth from space (sim500 kmaway) [1] It is a space-borne image acquisition techniquethat is able to penetrate cloud coverage and that is notdependent on sunlightTheEarthrsquos surface is illuminatedwithelectromagnetic waves sent by the satellite antenna [2 3]Thebackscattered signal is received by the same antenna (mono-static) or other antennas (bi- or multistatic) and is storedas a complex image with amplitude and phase information[4] SAR images have a broad range of applications such asurban surveillance [5 6] damage assessment [7ndash9] glaciermonitoring [10] and deformation mapping [11 12]

New generation satellites as TerraSAR-X or COSMO-SkyMed provide images in single polarization with a resolu-tion of approximately 1m (with 300MHz bandwidth) [13]The highest-resolution image mode is called spotlight modeand coversmdashcompared to other image modesmdasha relativelysmall ground area [13] TerraSAR-X and COSMO-SkyMed

operate in the X-band spectrum which is particularly usefulfor observing man-made structures High resolution is ofinterest when monitoring small parts of urban areas andsingle building structures [5] In order to achieve resolutionin the meter or submeter region the antenna beam is steeredbackwards while the satellite passes the area of observation[14] One limitation is that continuous acquisition along theground track is not possible A full spotlight scene with twosubsets is shown in Figure 1

A consequence of beam steering is the linear Dopplershift that causes missmatched image spectra which have tobe aligned in order to create spotlight interferograms [14]Theground area has to be sampled more often to achieve highspatial resolution which is only possible by increasing theillumination time Therefore the beam is directed forward(from the perspective of the satellite) and graduallymotionedbackwards as the antenna passes the area of interest [13]This means a positive Doppler frequency at the first imageline which is linearly declining line by line reaching a singleline with zero Doppler in the image center and negative

2 International Journal of Antennas and Propagation

Look direction

Figure 1 TerraSAR-X Spotlight image (full scene) of Luneburg Germany Left subset shows a railway station right subset shows buildingstructures and a pond copy DLR 2010

frequencies towards the last line Matching the two imagespectra is the challenge in interferometric processing ofspotlight data

The purpose of this paper is (i) to explain high-resolutionspotlight interferogram generation for bi- and monostaticacquisitions (ii) the implementation of a resampling algo-rithm described in detail in Eineder et al [14] and (iii) mak-ing it available to the research community by incorporatingthe source code into DORIS [15]This will enable researchersto generate coherent high-resolution interferograms with anopen software solutionThe key point of this paper lies in theinstruction to process bi- and monostatic acquisitions

In this paper we showTanDEM-X single pass interferom-etry (bistatic) and TerraSAR-X (TSX) repeat pass interferom-etry (monostatic) We abbreviate the TerraSAR-X satellite asTSX TanDEM-X as TDX and the bistatic product generatedby the satellite constellation as TDM In the first (bistatic)case one satellite serves as a sender to illuminate the groundBoth antennas on satellite 1 (TSX) and 2 (TDX) are usedfor receiving [16] In order to form images in this modethe spacecrafts have to orbit in close formation and must belinked together to acquire two images at almost the same timefrom slightly different positions [4] In the second case onesatellite serves as a sending and receiving antenna during oneimage acquisition Images are recorded every time the satellitepasses the area of interest

This article is divided into four parts starting fromsimple to more complex operations It will first discuss TDMspotlight products (TanDEM-X bistatic) and the additionalprocessing steps required to generate a bistatic interferogramThe second part illustrates how proper spotlight interfero-grams are derived by describing how spotlight images areacquired and subsequently treated during the resamplingprocess The last two parts show examples and discuss theresults which are achieved with the presented implementa-tion

2 TanDEM-X Spotlight Interferometry

TanDEM-X (TDX) is the twin satellite of TerraSAR-X (TSX)and the mission purpose of this constellation is to acquire aglobal homogeneous DEM [4] Using InSAR to generate aDEM (digital elevation model) is well known since NASArsquosSRTM (Shuttle Radar Topography Mission) However thereare certain aspects that have to be considered tomaximize thequality of the interferogram and therefore the resultingDEMIn order to minimize phase decorrelation (due to temporalchanges or surface deformation) both satellites have torecord the images within a very short time delay Anotheraspect in DEM generation is the height ambiguity whichis determined by choosing an appropriate baseline Bothsatellites orbit consistently in very close formation recordingimages at almost the same time with a defined baselineImages taken in this setup show very high coherence and aretherefore particularly suitable for interferogram generationas shown in Figure 15

21 Mission Design for Bistatic Image Acquisition The pri-mary mission goal of TanDEM-X is the generation of aDEM of unprecedented accuracy with a spatial resolution of12m and a relative vertical accuracy of 2m [16] The globalcoverage will be larger towards the poles compared to theSRTM [16]

Themission design is a constellation of TSX and TDX as atwin concept It is based on two almost identical SAR systemsorbiting in close formation When TSX was launched italready incorporated all necessary equipment and proceduresto operate with a second satellite Both satellites need tobe able to synchronize the phase therefore X-band hornantennas are installed on both satellites The dual frequencyGPS receiver estimates the orbit parameters precisely Thepulse repetition frequency (PRF) is synchronized by GPSmeasurements An additional cold gas propulsion system

International Journal of Antennas and Propagation 3

line = 119861perp

Equa

tor

Equa

tor

Equa

tor

Sout

h po

le

Nor

th p

ole

TanDEM-X orbit helix of TSX and TDX

Orbit tube

TDX orbitTSX orbit

Distance betweenred and green

Theoreticalground path

Figure 2 TerraSAR-X (red) and TanDEM-X (green) orbitconstellationmdashclose formation helix-like orbit X-axis is onerevolution around the globe

allows TDX to keep the right orbit attitude Both satellitescan operate separately in the imagemode spotlight (see imageformation in Section 31) stripmap or ScanSAR Operatingas a twin system brings up three different data acquisi-tion modes these include bistaticmdashwhich will be discussedbelowmdashpursuit monostatic and alternating bistatic [17]

In bistatic imaging one satellite is serving as transmitterand both satellites receive the backscattered signal Temporaldecorrelation is minimized and the atmospheric effects areequal in both images [4] Ground objects that move withinthis velocity range for example water or trees affected bywind are still incoherent Coherent results are achieved bysynchronizing the PRF between the two receiving antennasThe desired difference between both images is a slightly dif-ferent view angle That is expressed by the effective baselinewhich is half the perpendicular baseline [4] The modelingand determination of the orbital baseline for bistatic arediscussed in [18]

To ensure a constant baseline both satellites orbit in ahelix-like fashion with a short distance along the flight passOne satellite rotates around the orbit tube of the secondsatellite This is necessary to keep an effective baseline ofaround 200m [4] Schematic sketches are shown in Figures2 and 3 One can imagine two circular orbits with a crosstrack distance of sim360m at the Equator and sim plusmn400mradial distance at the poles [4] With two orbits as describedabove one satellite (eg TSX) is considered as the referenceThe second spacecraft (eg TDX) is then revolving aroundthe orbit of the reference One whole circle is performedduring one revolution around the Earth Having a controlledorbit gives a predictable perpendicular baseline and heightambiguity

The differences between repeat pass and single passinterferometry are visualized in Figure 3 Repeat pass inter-ferometry refers to a single satellite system (M) acquiring oneimage at a time To observe the Earth two times in a row withsimilar image geometrymdasha prerequisite for interferometry[2]mdashthe satellite has to pass the same locationorbit againHowever this is only possible after a certain amount of timeexpressed by the temporal baseline for example sim11 daysfor TSX Single pass interferometry refers to an image pairgeneration within the same overpass The temporal baselinebetween both images is then nearly zero The advantages oftwo satellites (M and S) being just several hundred meterapart is that the temporal phase decorrelation is very lowA coherent image pair will produce an interferogram ofhigh accuracy and quality with minimized temporal decor-relation due to vegetation or varying weather conditionsAtmospheric effects on two different dates can be neglected[4]

22 Processing Considerations In this part we will discussthe actual processing steps for TDM images necessary togenerate an interferogram and subsequently flatten it Abistatic complex image pair is the recorded backscatter signalof two identical satellite antennas with a defined baselineThedelivered TDM spotlight product is an already coregisteredpair of standard SSC images and named CoSSC Obligatoryprocessing steps before interferogram formation are alreadydone This information is mentioned in the auxiliary XMLfile of a TDM CoSSC image pair During the processingwith DORIS three issues occurred (i) the changes in thedata format to 16-bit floating point (ii) extracting the rightbaseline and (iii) the different range (travel) times for thereference phase

TDM data are delivered in the CoSSC file formatwith complex half-precision float values (float16) Float16 isdefined by the IEEE 754 standard and explained in [19] SSCdata is delivered as complex integer with 16-bit precisionAs a result one complex number consists of two 16-bitfloating-point values A conversion from float16 to float32was performed with a library developed by [20] Correctvalues aremaintained and the data can be used in non-float16compatible software

The metadata states that both images are cropped to thesame extent corrected for bistatic mode spectral filtered inrange and azimuth coregistered and resampled Thereforeinterferogram generation can be started directly Dependingon the setup either one of the satellites is used as the master(TSX or TDX) By taking the complex phase information ofthe master and slave an interferogram can be derived with

119868 = 119872 sdot 119878

lowast

sdot 119877

lowast

(1)

M is the complex signal of the master 119878lowast is the complexconjugate of the slave119877lowast the conjugate of the reference phaseand I is the resulting interferogram [21]

DORIS extracts the orbital parameters slightly differentfor TDM By default it reads the data as if both images wereacquired independently in the monostatic mode Bistaticimage processing with TDM data is not considered [15] The


M M

Ground target

MonostaticPossible locationsfor satellite withinorbit tube space

ReceiveSend

MS

Bistatic

Possible locationsfor slave satellite

around master orbit

Figure 3 Monostatic (left side) image acquisition on two dates the satellite (M) sends a signal (dotted line) and receives the echo (dashedline) during each overpass Bistatic image acquisition generates two images in a single overpass with almost no time difference One satelliteserves as transmitter and both serve as receiving antennasThe baseline of the slave satellite is desired to be on a circle around the master seeFigure 2

Hei

ght o

fam

bigu

ity

Distance between tracks

Sat 1

Sat 2119861perp = (effective baseline) times 2

Figure 4 Relation of baseline annotations and orbit geometry forTanDEM-X

orbit information is retrieved based on the annotation filesof TSX and TDX separately Both files are treated as if theywere two separate acquisitions with a temporal baseline of0 days This is sufficient at the initial reading step Whenestimating the baseline in DORIS a mismatch between theperpendicular baseline and the effective baseline occurredAccording to Krieger and Moreira [4] the effective baselineis half the length of the perpendicular baseline which is dueto the bistatic acquisition mode (see Figure 4) The DORISresult file has to be modified accordingly

Due to the bistatic acquisition one has to take care ofthe range times by taking half the travel time instead of thefull range distance from and to the antenna The reference

phaseΦreference as mentioned and annotated with 119877 in (1) fora mono- or bistatic image pair is calculated by

Φreference = ΦMaster const sdot t rangeMaster

minus ΦSlaveconst sdot trangeSlave (2)

ΦMaster const is a constant value based on the wavelength ofthe master image and derived with either option in

Monostatic ΦMaster const =minus40 sdot 120587 sdot 119888

120582Master

Bistatic ΦMaster const =minus20 sdot 120587 sdot 119888

120582Master

(3)

where 119888 is the speed of light and 120582Master is the wavelengthof the master image For a signal traveling to the groundand back to the satellite (monostatic) the factor would beminus40 for two full phase circles instead of minus20 This comesfrom the radar range equation and has to be changed for thebistatic mode t rangeMaster is the range time based on thepixel location as stated as follows

t rangeMaster = t range0+

(pixel minus 10)RSR

(4)

Pixel is the pixel number (column as integer) of that particu-lar line andRSR is the range sampling rate as annotated in themetadata The interferometric phase can now be processedcorrectly

3 TerraSAR-X Spotlight Interferometry

Very high-resolution spaceborne SAR images generatedwith current satellites are only possible by sweeping the


Orbit pathAcquisition time

Antenna

Ground track

Ground samples

Illuminated ground area

Forward

Backward

Image footprint

Figure 5 Image acquisition geometry for spotlight mode The beam is steered backwards when the satellite passes the area of interest [4]

Read master Read slave

Crop master Crop slave

Coregistration

Shift kernel Resample

Removefrequency

Reinsertfrequency

InterferogramformationDORIS Matlab

Figure 6 General processing flow chart Two options implementa-tion in DORIS shifting the kernel (left) external scripts inMATLABto set the image spectrum to zero (right)

antenna beam in spotlight modeThe occurring problems arementioned above We argue that especially high-resolutioninterferograms would be explored more frequently if theprocessing could be done with more ease

31 Spotlight Image Formation in Repeat Pass Spotlightimage mode of TSX and TDX is a sliding azimuth beam

Range (pixel)

Azi

mut

h (ti

me)

Doppler records(109 polynomials)

Image framePixel 11

Pixel max max

sim+3700Hz

simminus3700Hz

Figure 7 Schematic sketch of interpolation values derived fromXML data The x-axis represents the range time calculated fromthe pixel number and the y-axis is interpolated from the annotatedDoppler polynomial records The recording of them starts and endsbefore and after the actual image footprint

steering technique [14] As shown in Figure 5 the satellitepasses the area of interest in a certain time The observationtime is increased by steering the antenna beam in azimuthSee Figure 5 for a diagram of this process The drift rate of


0

1000

2000

3000

4000

5000 6000minus4000

minus3000

minus2000

minus1000

0 1000 2000 3000 4000Lines

fDC

(Hz)

Actual frequencyModeled frequency

Figure 8 Modeled and actual Doppler frequency at pixel 4000(range) from the first to the last line

the sweeping motion is slower than the overpass velocityConsequently the beam is sliding over an area and not star-ring at a single location This results in more ground samplesand an increased ground resolution One shortcoming is thelimited image frame length A longer illumination time isachieved in azimuth but not in range High-range resolutionis accomplished with a bandwidth of 300MHz [13]

The Doppler frequency decreases over the whole imagelength in azimuth This behavior is different compared tothe standard stripmap mode The Doppler spectrum of aspotlight image follows a linear trend from positive Dopplerfrequencies around 3700Hz to negative values aroundminus3700Hz as visualized in Figure 7 Theoretically only onezero Doppler line exists in the center of the image This is thecase when the beam is steered exactly perpendicular to theorbit and the Earth Stripmap images do not have this lineartrendmdashall lines are centered around the Doppler frequencyProper treatment of the Doppler frequency is necessary forInSAR processingThe amplitude information is not affected

32 Implementation of Processing Method As discussed in[14] and mentioned in [15 22] there are two options to dealwith the different image spectra of master and slave Correctphase information is kept when the spectrum of the slaveis set to zero before resampling or the interpolation kernelis shifted during the resampling step We implemented thefirst option in MATLAB Both options are shown as a flowchart in Figure 6 Shifting the interpolation kernel is thesecond option and is described in detail in [14]The accordingDoppler frequency tomultiply the kernel with is retrieved byexploiting the annotated Doppler polynomials (see Figure 7)Eineder et al [14] propose a low-pass azimuth bandpass filterafter resampling We will focus solely on the generation ofthe frequency matrix for the pixel-by-pixel calculation of theinterpolation kernel

The chosen interpolation kernel is a six-point cubic con-volution kernel as described in [22] implemented in DORIS

and shown in (5) Hanssen and Bamler [22] recommenda six-point cubic convolution kernel for high-resolutionapplications that is also recommended by Eineder et al [14]and used in our implementation

119894 (119909) =

(120572 minus 120573 + 2) |119909|

3

minus (120572 minus 120573 + 3) |119909|

2

+ 1

0 le |119909| lt 1

120572|119909|

3

minus (5120572 minus 120573) |119909|

2

+ (8120572 minus 3120573) |119909| minus (4120572 minus 2120573)

1 le |119909| lt 2

120573|119909|

3

minus (8120573) |119909|

2

+ (21120573) |119909| minus (18120573)

2 le |119909| lt 3

0 3 le |119909|

(5)

where 120572 and 120573 are two parameters (120572 = minus12 and 120573 = 12)as defined by [22] The resulting interpolation kernel 119894(119909) isthen multiplied with the modeled Doppler frequency duringresampling

To calculate the frequency matrix the timing andDoppler rate metadata are evaluated for each product specifi-cally See appendix of [14 23] for amore elaborate descriptionFigure 7 helps to visualize the pixel-by-pixel calculation forthe frequency matrix The number of Doppler polynomialrecords is variable and not fixed The recording of thesediscrete measurements starts and ends before and after theactual image footprint

Each annotated Doppler polynomial is evaluated alongthe range time as described in [16] and shown as follows

fDC (pixel) = (fDCcoef1198910 lowast (119875 (pixel) minus Refpoint) 0)

+ (fDCcoef1198911 lowast (119875 (pixel) minus Refpoint)1)


(6)

119875(pixel) is the function pixel2119905range (pixel) which calculatesthe range time based on the pixel number (integer value)the range sampling frequency RSF and the start range time119905range0 The corresponding calculation is shown as follows

pixel2119905range (pixel) = 119905range0 +(pixel minus 10)

RSF

(7)

The timing information gathered from the metadata arein the system receive time (the actual uptime of the satellite)and has to be converted to zero-Doppler time system by

119905SSC = (119905RAW minus (

119891DC (119905RAW)

119865119872rate)) minus 119905zerominusDoppler Start (8)

119905RAW is the time UTC annotation in seconds and119905zerominusDoppler Start is the actual first zero Doppler time Thementioned 119865119872rate is calculated according to the followingfunction

119865119872rate =fDCcoef1198910 minus

fDCcoef1198911

2

(9)


Imag

e dist

ortio

n in

azim

uth

Distorted fringes

Incoherence dueto shadow effect

Last line

First lineLook direction

Figure 9 TerraSAR-X image pair (Uluru) with default frequency shift in DORIS The image is not corrected and gets distorted towards thelast line

13 km

One fringe = one fullcolor circle two pointssim118 km length

Bend profile with 95fringes (color circles)sim093km length

8x

Runway

20 m


Look direction

Figure 10 Interferogram generated with DORIS using the cc6p SP interpolation kernel Temporal baseline is 11 days and perpendicularbaseline is 2337m with a height ambiguity of 34m Master acquired on 2009-02-12 and slave on 2009-02-23

fDCcoef1198910 marks the Doppler rate at the start of the scene andfDCcoef1198911 at the end of the sceneThe 119865119872rate is the average ofthese values To interpolate along azimuth the fDC and 119905SSCvalues are used

Comparing the actual frequency and the interpolatedmodeled frequency derived from the metadata is shownin Figure 8 The spectrum for a single line in azimuth iscolored in red and follows an almost linear trend fromapprox3500Hz in the first line to around minus3500Hz in the last line

of the image To shift the kernel the modeled frequency wasused (black line)

4 Results and Discussion

Results for one bistatic TDM product and several monostaticimage pairs from TSX and TDX have been produced ATDM pair from Las Vegas in high-resolution spotlight modewas used to test the bistatic configuration of DORIS and


M 2010-03-24 S 2010-02-19

M 2010-03-24 S 2010-05-29

M 2010-03-24 S 2010-03-02

M 2010-03-24 S 2010-06-20 M 2010-03-24 S 2010-07-12 M 2010-03-24 S 2010-08-03 M 2010-03-24 S 2010-08-25

M 2010-03-24 S 2010-05-07M 2010-03-24 S 2010-04-15

M 2010-03-24 S 2010-12-24M 2010-03-24 S 2010-11-10

M 2010-03-24 S 2011-04-13 M 2010-03-24 S 2011-05-16 M 2010-03-24 S 2011-05-27

M 2010-03-24 S 2011-02-06M 2010-03-24 S 2010-10-08

M 2010-03-24 S 2011-03-22

M 2010-03-24 S 2010-09-16

M 2010-03-24 S 2011-02-28

Master2010-03-24

minus7∘C

+5∘C

+11∘C minus7

∘Cminus7∘C0

∘C

0∘C

+13∘C

+7∘C

+1∘C

+16∘C

0∘C

+3∘C

+2∘C

+13∘C

minus7∘C +5

∘C +5∘C

+5∘C

+13∘C

Figure 11 Time series of TerraSAR-X ascending stack showing flattened interferograms Area Convention Center in Las Vegas USATemperature annotations indicate the relative temperature difference to the master image (2010-03-24) source [18]

the result is shown in Figure 13 Our implementation of thecc6p SP processing option in the resampling stage of DORIS(version 406 beta) was tested with the following TerraSAR-X HS source images ascending and descending data stacksfrom Las Vegas (20 images each) and one data stack fromLuneburg (16 images) In addition a sample dataset fromUluru Australia provided by Infoterra Germany [24] wasused All scenes were acquired in high-resolution spotlightmode The process chain includes all steps listed in DORISmanual [21] until coherence estimation We present flattenedinterferograms flattened and filtered interferograms andcoherence images

The implementation of bistatic image processing wasdone in DORIS and the results are shown in Figures 13 and15 By taking care of the float16-bit data format the referencephase and the effective baseline a coherent interferogramwas generated

To demonstrate the effect of running DORIS with thedefault parameters in the input card ldquoresamplerdquo that isfrequency shift in azimuth as described in [21] during

interpolation leads to the results shown in Figure 9 Applyingthe default frequency shift results in an interferogram wherethe phase seems to be correct in the first lines gettingmore and more distorted and corrupted towards the lastline This is due to the not applied linear kernel shift inazimuth Using a six-point cubic convolution kernel with alinear frequency shift as described above and in [21] yields aflattened interferogram as shown in Figure 10

There are a few things to pay attention to in Figure 10for example fringes along steep slopes and distorted fringesdue to low coherence caused by shadow effects The resultsare not DEM corrected The phase follows the topographyAlong the bend profile (bottom to top of the rock) 95 fringescan be counted Evidence for a relatively flat surroundingarea is given by the distance between two points (sim118 Kmannotation on the right) with one phase circle

Stacking images is a commonly used technique for timeseries analysis A TSX ascending image stack over Las Vegaswas used to generate interferograms flattened the phaseand then cropped to the extend of the Convention Center


M 2010-07-16 S 2009-12-08 M 2010-07-16 S 2010-01-10 M 2010-07-16 S 2010-04-08

M 2010-07-16 S 2010-06-13

M 2010-07-16 S 2010-09-20

M 2010-07-16 S 2010-12-17

M 2010-07-16 S 2011-02-21

M 2010-07-16 S 2010-06-02

M 2010-07-16 S 2010-08-29

M 2010-07-16 S 2010-11-25

M 2010-07-16 S 2011-01-30

M 2010-07-16 S 2010-05-22

M 2010-07-16 S 2010-06-24

M 2010-07-16 S 2010-11-03

M 2010-07-16 S 2010-01-08

Figure 12 Time series of 15 full scene TerraSAR-X spotlight interferograms over Luneburg GermanyThe color intensity has been increasedextremely

(36∘071015840524410158401015840North 115∘091015840048510158401015840West) see Figure 11 Eachpair was processed with exactly the same parameter set Themaster image was chosen based on minimizing 119861perp Thesubsets are still in radar coordinates and not reprojected theillumination direction is from the left We show this resultbecause (a) the roof is a relatively flat surface with highbackscatter and (b) it is from the center of the image wherethe Doppler shift between both acquisitions is rather lowRelating the visualized phase to the temperature information(relative to the master) one can see that the roof gets moredeformed the greater the absolute temperature difference

The data stack of Luneburg with 16 images showsless coherence The interferograms for a full scene aredepicted in Figure 12 Coherent areas are found in theurban area where ground is covered with solid structureslike buildings and roads The surrounding fields are mostly

incoherent and therefore homogeneous phase informationcannot be derived for these areas The master image wasselected based on minimizing temporal and perpendicularbaselines

A comparison of bi- and monostatic images is illustratedin Figures 13 and 14 The higher brilliance of the TDM imagepair compared to theTSXTDX image pair is due to the smalltemporal decorrelation for bistatic image acquisitions

Looking at the hotels and the Convention Center inFigures 13 and 14 demonstrates two things First the phase orfringes following along the front of the hotels are similar inboth scenes Second the phase on the Convention Center isvery smooth in the TDM image pair On the other hand theTSX image pair shows more fringes on the roof indicatinga motion of the roof in LOS (line of sight) between bothacquisitions


Look direction

Figure 13 TanDEM-X flattened interferogram in radar coordinates with amplitude image in background Subsets show hotels and the LasVegas Convention Center Temporal baseline 0 days effective baseline sim87m

Look direction

Figure 14 TerraSAR-X flattened interferogram in radar coordinates with amplitude image in background Subsets show hotels and the LasVegas Convention Center Temporal baseline 110 days perpendicular baseline 481m

On the subject of coherencemdashwhich is a strong indicatorfor the quality of an interferogrammdashthe effect of bistaticacquisitions compared to repeat pass interferometry becomesapparent In the bistatic TDM image pair over Las Vegaswe observe a mean coherence for the whole scene of about07048 see Figure 15 However comparing this with a pairof TerraSAR-X acquisitions with a temporal baseline of 11days gives a mean coherence of 06486 see Figure 16 Thementioned pair was taken on 2010-09-19 for the master and2010-09-08 for the slave Picking two images with a longertemporal baseline of 110 days for 2010-03-24 (master) and2010-07-12 (slave) gives a mean coherence of 05206 seeFigure 17 for a histogram as well

Comparing the relatively dense urbanized scene of LasVegaswith a relatively rural setting in Luneburg illustrates thefollowing The coherence for any scene in Luneburg is ratherlow for example a temporal baseline of 22 days in summerbetween the images taken on the 2010-07-16 and 2010-06-24 gives a mean coherence of 04029 There are two specialcases one in Las Vegas and one in Luneburg where very low

coherence was noticeable by visual interpretation The imagepairs in question are 2010-03-24 with 2010-08-03 (Las Vegas)and 2010-07-16 with 2010-01-10 (Luneburg) which show amean coherence value of 04071 and 03506 respectively Inthe case of Luneburgwewere able to retrace the possible causeby checking the weather data that was snowfall [25] on thatparticular day

5 Conclusion

In this paper we explained the steps necessary to processbistatic and monostatic TanDEM-X and TerraSAR-X imageacquisitions utilizing a public domain tool DORIS DORISwas modified and extended for this purpose We havealso clarified the basic concepts of bistatic acquisitions andillustrated linear Doppler drift which occurs during thebeam steering technique used in spotlight image creationWespecifically focused on the generation of a Doppler frequencymatrix that is used to shift the interpolation kernel during


Figure 15 TanDEM-X coherence image The brighter the more coherent Parks and water become visible due to the effects of wind

Figure 16 TerraSAR-X repeat pass coherence image Master 2010-07-12 slave 2010-03-24

resampling Furthermore experimental results based on sev-eral input datasets demonstrated that our implementationgives consistently good output A comparison of bistaticTDM and monostatic TSXTDX interferograms shows thedifference in phase and general coherence

The character of the paper is rather instructional with afocus on the explanation of bi- and monostatic very high-resolution SAR imagesThe key point lies within the guidanceto understand and produce TSX spotlight interferogramsWe also show that advanced image acquisition techniqueslike spotlight and or bistatic spotlight can be processedwith already available open-source software solutions Addi-tionally it demonstrates the quality and consistency of thedata which has been used in the experiments This papercan serve as a basis for future papers by illustrating thepotential of the very high-resolution repeat or single pass SARinterferometry

A TanDEM-X interferogram can be retrieved by just afew modifications in the source code of DORIS The overall

coherence of the image is relatively high given the veryshort temporal baseline in bistatic acquisitions Experimentalresults showed that the roof of the Las Vegas ConventionCenter or other buildings can exhibit thermal deformationThis is observable due to the phase information In the caseof TDM the phase is very homogenous because of the con-temporaneous image acquisitionThat indicates nomotion atleast not detectable with SAR within the acquisition period

Particularly interesting is the comparison of the bistaticand monostatic images The overall coherence of a TDM isas expected very high Even water does not show up withthe significant dark signature The TSX image pair with 110days temporal baseline is less coherent and objects in quickmotion like treesleafs and water show very low coherence

Spotlight InSAR in DORIS requires some modifica-tions to calculate consistent interferograms The annotatedDoppler information in the XML metafile have to be eval-uated and a Doppler frequency matrix was created It wasthen used to shift the six-point cubic convolution kernel


0

05

1

15

2

25

3

35times104

0 02 04 06 08 1Coherence

Cou

nt

TSX

TDM

Figure 17 Coherence histogram comparison between TerraSAR-X(TSX) with 110 days 119861temp and a TanDEM-X product (TDM)

during the resampling stage to match both image spec-tra A new ldquocc6p SPrdquo option in the processing input cardldquoRS METHODrdquo in DORIS has been added to accomplishthis We have compared the result to a calculation that usedthe default kernel shift in DORIS The result was expectantlydistorted Additionally we compared to a result made by acommercially available software with very high agreement toour result

Interpreting the phase information on the roof of theLas Vegas Convention Center gave evidence of correlationbetween temperature and number of fringes or distancebetween fringes respectively The higher the temperaturedifferences to the master image the greater the deformationThis was qualitatively assessed by looking at the fringesappearing for separate roof sections

The initial goal was to provide an implementation of analready known algorithm to process TerraSAR-X spotlightimages in freely available software DORIS As seen fromthe results the integrated processing option to shift theinterpolation kernel according to the Doppler frequencyworks seamlessly and effectively It enables researchers toprocess high-resolution spotlight interferograms

The code is directly available upon request from thecorresponding author and will probably be included in thenext release of DORIS

Acknowledgments

This work is financially supported by the National KeyBasic Research Program of China (Grant no 2013CB733205)the National Natural Science Foundation of China (Grantno 41174120) and the German Academic Exchange Service(DAAD) The authors would like to thank the German

Aerospace Center (DLR) for providing the test datasets viathe DLR AO LAN1301 and LAN0634

References

[1] M Eineder H Runge E Borner et al ldquoSAR interferometrywith TerraSAR-Xrdquo in Proceedings of FRINGE Workshop Fras-cati Italy December 2003

[2] RHanssenRadar Interferometry Data Interpretation and ErrorAnalysis 2001

[3] IWoodhouse Introduction toMicrowave Remote Sensing 2005[4] G Krieger and A Moreira ldquoTanDEM-X a satellite formation

for high-resolution SAR interferometryrdquo IEEE Transactions onGeoscience and Remote Sensing vol 45 no 11 pp 3317ndash33412007

[5] U Soergel U Thoennessen A Brenner and U Stilla ldquoHigh-resolution SAR data new opportunities and challenges for theanalysis of urban areasrdquo IEE Proceedings Radar Sonar andNavigation vol 153 no 3 pp 294ndash300 2006

[6] P Gamba B Houshmand and M Saccani ldquoDetection andextraction of buildings from interferometric SAR datardquo IEEETransactions on Geoscience and Remote Sensing vol 38 no 1pp 611ndash618 2000

[7] T Balz andM Liao ldquoBuilding-damage detection using postseis-mic high-resolution SAR satellite datardquo International Journal ofRemote Sensing vol 31 no 13 pp 3369ndash3391 2010

[8] Y Dong Q Li A Dou and X Wang ldquoExtracting damagescaused by the 2008Ms 80 Wenchuan earthquake from SARremote sensing datardquo Journal of Asian Earth Sciences vol 40no 4 pp 907ndash914 2011

[9] D Ehrlich H D Guo K Molch J W Ma and M PesaresildquoIdentifying damage caused by the 2008Wenchuan earthquakefromVHR remote sensing datardquo International Journal of DigitalEarth vol 2 no 4 pp 309ndash326 2009

[10] T Moon I Joughin B Smith and I Howat ldquo21st-centuryevolution of greenland outlet glacier velocitiesrdquo Science vol336 no 6081 pp 576ndash578 2012

[11] D Perissin ZWang andH Lin ldquoShanghai subway tunnels andhighways monitoring through Cosmo-SkyMed Persistent Scat-terersrdquo ISPRS Journal of Photogrammetry and Remote Sensingvol 73 pp 58ndash67 2012

[12] T Wang and D Perissin ldquoDeformation monitoring by longterm D-InSAR analysis in Three Gorges area Chinardquo IEEEGeoscience and Remote Sensing Letters vol 4 pp 7ndash10 2008

[13] German Aerospace Agency DLR ldquoTerraSAR-X ground seg-ment basic product specification documentrdquo in Working PaperTX-GS-DD-3 no 17 2010

[14] M Eineder N Adam R Bamler N Yague-Martinez andH Breit ldquoSpaceborne spotlight SAR interferometry withTerraSAR-Xrdquo IEEE Transactions on Geoscience and RemoteSensing vol 47 no 5 pp 1524ndash1535 2009

[15] B Kampes R Hanssen and Z Perski ldquoRadar interferometrywith public domain toolsrdquo in Proceedings of FRINGEWorkshopFrascati Italy December 2003

[16] H Fiedler G Krieger M Werner and K Reiniger ldquoTheTanDEM-X mission design and data acquisition planrdquo inProceedings of the European Conference on Synthetic ApertureRadar (EUSAR rsquo06) Dresden Gemany May 2006

[17] German Aerospace Agency DLR ldquoTanDEM-X ground seg-ment experimental product descriptionrdquo inWorking Paper TD-GS-PS-3028 no 12 2007


[18] J Tu D Gu Y Wu and D Yi ldquoError modeling and analysisfor InSAR spatial baseline determination of satellite formationflyingrdquo Mathematical Problems in Engineering vol 2012 pp 1ndash23 2012

[19] J van der Zijp ldquoFast half float conversionsrdquo Working paper2012ftpwwwfox-toolkitorgpubfasthalffloatconversionpdf

[20] C Rau ldquoHalf-precision floating point libraryrdquo sourceforgenet2012 httphalfsourceforgenet

[21] Technical University Delft ldquoDelft Object-oriented Radar Inter-ferometric Software Userrsquos manual and technical documenta-tionrdquo Tech Rep Delft The Netherlands 2008

[22] R Hanssen and R Bamler ldquoEvaluation of interpolation kernelsfor SAR interferometryrdquo IEEE Transactions on Geoscience andRemote Sensing vol 37 no 1 pp 318ndash321 1999

[23] German Aerospace Agency DLR ldquoTerraSAR-X ground seg-ment level 1b product format specificationrdquo in Working PaperTX-GS-DD-3307 no 13 2007

[24] ASTRIUM ldquoSample Imagery Australia UrulumdashInSARrdquo 2009httpwwwastrium-geocomen23-sample-imagery

[25] The Weather Channel Companies ldquoWeather Undergroundrdquo2010 httpwwwwundergroundcom

International Journal of

AerospaceEngineeringHindawi Publishing Corporationhttpwwwhindawicom Volume 2014

RoboticsJournal of

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014


Active and Passive Electronic Components

Control Scienceand Engineering

Journal of



RotatingMachinery


Hindawi Publishing Corporation httpwwwhindawicom

Journal ofEngineeringVolume 2014

Submit your manuscripts athttpwwwhindawicom

VLSI Design



Shock and Vibration


Civil EngineeringAdvances in

Acoustics and VibrationAdvances in



Electrical and Computer Engineering

Journal of

Advances inOptoElectronics


Volume 2014

The Scientific World JournalHindawi Publishing Corporation httpwwwhindawicom Volume 2014

SensorsJournal of


Modelling amp Simulation in EngineeringHindawi Publishing Corporation httpwwwhindawicom Volume 2014


Chemical EngineeringInternational Journal of Antennas and

Propagation




Navigation and Observation



DistributedSensor Networks



Look direction

Figure 1 TerraSAR-X Spotlight image (full scene) of Luneburg Germany Left subset shows a railway station right subset shows buildingstructures and a pond copy DLR 2010

frequencies towards the last line Matching the two imagespectra is the challenge in interferometric processing ofspotlight data

The purpose of this paper is (i) to explain high-resolutionspotlight interferogram generation for bi- and monostaticacquisitions (ii) the implementation of a resampling algo-rithm described in detail in Eineder et al [14] and (iii) mak-ing it available to the research community by incorporatingthe source code into DORIS [15]This will enable researchersto generate coherent high-resolution interferograms with anopen software solutionThe key point of this paper lies in theinstruction to process bi- and monostatic acquisitions

In this paper we showTanDEM-X single pass interferom-etry (bistatic) and TerraSAR-X (TSX) repeat pass interferom-etry (monostatic) We abbreviate the TerraSAR-X satellite asTSX TanDEM-X as TDX and the bistatic product generatedby the satellite constellation as TDM In the first (bistatic)case one satellite serves as a sender to illuminate the groundBoth antennas on satellite 1 (TSX) and 2 (TDX) are usedfor receiving [16] In order to form images in this modethe spacecrafts have to orbit in close formation and must belinked together to acquire two images at almost the same timefrom slightly different positions [4] In the second case onesatellite serves as a sending and receiving antenna during oneimage acquisition Images are recorded every time the satellitepasses the area of interest

This article is divided into four parts starting fromsimple to more complex operations It will first discuss TDMspotlight products (TanDEM-X bistatic) and the additionalprocessing steps required to generate a bistatic interferogramThe second part illustrates how proper spotlight interfero-grams are derived by describing how spotlight images areacquired and subsequently treated during the resamplingprocess The last two parts show examples and discuss theresults which are achieved with the presented implementa-tion

2 TanDEM-X Spotlight Interferometry

TanDEM-X (TDX) is the twin satellite of TerraSAR-X (TSX)and the mission purpose of this constellation is to acquire aglobal homogeneous DEM [4] Using InSAR to generate aDEM (digital elevation model) is well known since NASArsquosSRTM (Shuttle Radar Topography Mission) However thereare certain aspects that have to be considered tomaximize thequality of the interferogram and therefore the resultingDEMIn order to minimize phase decorrelation (due to temporalchanges or surface deformation) both satellites have torecord the images within a very short time delay Anotheraspect in DEM generation is the height ambiguity whichis determined by choosing an appropriate baseline Bothsatellites orbit consistently in very close formation recordingimages at almost the same time with a defined baselineImages taken in this setup show very high coherence and aretherefore particularly suitable for interferogram generationas shown in Figure 15

21 Mission Design for Bistatic Image Acquisition The pri-mary mission goal of TanDEM-X is the generation of aDEM of unprecedented accuracy with a spatial resolution of12m and a relative vertical accuracy of 2m [16] The globalcoverage will be larger towards the poles compared to theSRTM [16]

Themission design is a constellation of TSX and TDX as atwin concept It is based on two almost identical SAR systemsorbiting in close formation When TSX was launched italready incorporated all necessary equipment and proceduresto operate with a second satellite Both satellites need tobe able to synchronize the phase therefore X-band hornantennas are installed on both satellites The dual frequencyGPS receiver estimates the orbit parameters precisely Thepulse repetition frequency (PRF) is synchronized by GPSmeasurements An additional cold gas propulsion system


line = 119861perp

Equa

tor

Equa

tor

Equa

tor

Sout

h po

le

Nor

th p

ole


Orbit tube

TDX orbitTSX orbit











119868 = 119872 sdot 119878

lowast

sdot 119877

lowast

(1)




M M

Ground target


ReceiveSend

MS

Bistatic


around master orbit


Hei

ght o

fam

bigu

ity


Sat 1










120582Master


120582Master

(3)



(pixel minus 10)RSR

(4)






Antenna

Ground track

Ground samples


Forward

Backward

Image footprint




Coregistration


Removefrequency

Reinsertfrequency





Range (pixel)

Azi

mut

h (ti

me)


Image framePixel 11

Pixel max max

sim+3700Hz

simminus3700Hz




0

1000

2000

3000

4000

5000 6000minus4000

minus3000

minus2000

minus1000

0 1000 2000 3000 4000Lines

fDC

(Hz)








119894 (119909) =

(120572 minus 120573 + 2) |119909|

3

minus (120572 minus 120573 + 3) |119909|

2

+ 1

0 le |119909| lt 1

120572|119909|

3

minus (5120572 minus 120573) |119909|

2


1 le |119909| lt 2

120573|119909|

3

minus (8120573) |119909|

2

+ (21120573) |119909| minus (18120573)

2 le |119909| lt 3

0 3 le |119909|

(5)







(6)



RSF

(7)


119905SSC = (119905RAW minus (

119891DC (119905RAW)




fDCcoef1198911

2

(9)


Imag

e dist

ortio

n in

azim

uth

Distorted fringes


Last line



13 km



8x

Runway

20 m


Look direction








M 2010-03-24 S 2010-02-19

M 2010-03-24 S 2010-05-29

M 2010-03-24 S 2010-03-02

M 2010-03-24 S 2010-06-20 M 2010-03-24 S 2010-07-12 M 2010-03-24 S 2010-08-03 M 2010-03-24 S 2010-08-25

M 2010-03-24 S 2010-05-07M 2010-03-24 S 2010-04-15

M 2010-03-24 S 2010-12-24M 2010-03-24 S 2010-11-10

M 2010-03-24 S 2011-04-13 M 2010-03-24 S 2011-05-16 M 2010-03-24 S 2011-05-27

M 2010-03-24 S 2011-02-06M 2010-03-24 S 2010-10-08

M 2010-03-24 S 2011-03-22

M 2010-03-24 S 2010-09-16

M 2010-03-24 S 2011-02-28

Master2010-03-24

minus7∘C

+5∘C

+11∘C minus7

∘Cminus7∘C0

∘C

0∘C

+13∘C

+7∘C

+1∘C

+16∘C

0∘C

+3∘C

+2∘C

+13∘C

minus7∘C +5

∘C +5∘C

+5∘C

+13∘C









M 2010-07-16 S 2009-12-08 M 2010-07-16 S 2010-01-10 M 2010-07-16 S 2010-04-08

M 2010-07-16 S 2010-06-13

M 2010-07-16 S 2010-09-20

M 2010-07-16 S 2010-12-17

M 2010-07-16 S 2011-02-21

M 2010-07-16 S 2010-06-02

M 2010-07-16 S 2010-08-29

M 2010-07-16 S 2010-11-25

M 2010-07-16 S 2011-01-30

M 2010-07-16 S 2010-05-22

M 2010-07-16 S 2010-06-24

M 2010-07-16 S 2010-11-03

M 2010-07-16 S 2010-01-08








Look direction


Look direction





5 Conclusion












0

05

1

15

2

25

3

35times104

0 02 04 06 08 1Coherence

Cou

nt

TSX

TDM






Acknowledgments



References





























RoboticsJournal of





Journal of



RotatingMachinery





VLSI Design



Shock and Vibration







Journal of



Volume 2014


SensorsJournal of





Propagation










line = 119861perp

Equa

tor

Equa

tor

Equa

tor

Sout

h po

le

Nor

th p

ole


Orbit tube

TDX orbitTSX orbit











119868 = 119872 sdot 119878

lowast

sdot 119877

lowast

(1)




M M

Ground target


ReceiveSend

MS

Bistatic


around master orbit


Hei

ght o

fam

bigu

ity


Sat 1










120582Master


120582Master

(3)



(pixel minus 10)RSR

(4)






Antenna

Ground track

Ground samples


Forward

Backward

Image footprint




Coregistration


Removefrequency

Reinsertfrequency





Range (pixel)

Azi

mut

h (ti

me)


Image framePixel 11

Pixel max max

sim+3700Hz

simminus3700Hz




0

1000

2000

3000

4000

5000 6000minus4000

minus3000

minus2000

minus1000

0 1000 2000 3000 4000Lines

fDC

(Hz)








119894 (119909) =

(120572 minus 120573 + 2) |119909|

3

minus (120572 minus 120573 + 3) |119909|

2

+ 1

0 le |119909| lt 1

120572|119909|

3

minus (5120572 minus 120573) |119909|

2


1 le |119909| lt 2

120573|119909|

3

minus (8120573) |119909|

2

+ (21120573) |119909| minus (18120573)

2 le |119909| lt 3

0 3 le |119909|

(5)







(6)



RSF

(7)


119905SSC = (119905RAW minus (

119891DC (119905RAW)




fDCcoef1198911

2

(9)


Imag

e dist

ortio

n in

azim

uth

Distorted fringes


Last line



13 km



8x

Runway

20 m


Look direction








M 2010-03-24 S 2010-02-19

M 2010-03-24 S 2010-05-29

M 2010-03-24 S 2010-03-02

M 2010-03-24 S 2010-06-20 M 2010-03-24 S 2010-07-12 M 2010-03-24 S 2010-08-03 M 2010-03-24 S 2010-08-25

M 2010-03-24 S 2010-05-07M 2010-03-24 S 2010-04-15

M 2010-03-24 S 2010-12-24M 2010-03-24 S 2010-11-10

M 2010-03-24 S 2011-04-13 M 2010-03-24 S 2011-05-16 M 2010-03-24 S 2011-05-27

M 2010-03-24 S 2011-02-06M 2010-03-24 S 2010-10-08

M 2010-03-24 S 2011-03-22

M 2010-03-24 S 2010-09-16

M 2010-03-24 S 2011-02-28

Master2010-03-24

minus7∘C

+5∘C

+11∘C minus7

∘Cminus7∘C0

∘C

0∘C

+13∘C

+7∘C

+1∘C

+16∘C

0∘C

+3∘C

+2∘C

+13∘C

minus7∘C +5

∘C +5∘C

+5∘C

+13∘C









M 2010-07-16 S 2009-12-08 M 2010-07-16 S 2010-01-10 M 2010-07-16 S 2010-04-08

M 2010-07-16 S 2010-06-13

M 2010-07-16 S 2010-09-20

M 2010-07-16 S 2010-12-17

M 2010-07-16 S 2011-02-21

M 2010-07-16 S 2010-06-02

M 2010-07-16 S 2010-08-29

M 2010-07-16 S 2010-11-25

M 2010-07-16 S 2011-01-30

M 2010-07-16 S 2010-05-22

M 2010-07-16 S 2010-06-24

M 2010-07-16 S 2010-11-03

M 2010-07-16 S 2010-01-08








Look direction


Look direction





5 Conclusion












0

05

1

15

2

25

3

35times104

0 02 04 06 08 1Coherence

Cou

nt

TSX

TDM






Acknowledgments



References





























RoboticsJournal of





Journal of



RotatingMachinery





VLSI Design



Shock and Vibration







Journal of



Volume 2014


SensorsJournal of





Propagation










M M

Ground target


ReceiveSend

MS

Bistatic


around master orbit


Hei

ght o

fam

bigu

ity


Sat 1










120582Master


120582Master

(3)



(pixel minus 10)RSR

(4)






Antenna

Ground track

Ground samples


Forward

Backward

Image footprint




Coregistration


Removefrequency

Reinsertfrequency





Range (pixel)

Azi

mut

h (ti

me)


Image framePixel 11

Pixel max max

sim+3700Hz

simminus3700Hz




0

1000

2000

3000

4000

5000 6000minus4000

minus3000

minus2000

minus1000

0 1000 2000 3000 4000Lines

fDC

(Hz)








119894 (119909) =

(120572 minus 120573 + 2) |119909|

3

minus (120572 minus 120573 + 3) |119909|

2

+ 1

0 le |119909| lt 1

120572|119909|

3

minus (5120572 minus 120573) |119909|

2


1 le |119909| lt 2

120573|119909|

3

minus (8120573) |119909|

2

+ (21120573) |119909| minus (18120573)

2 le |119909| lt 3

0 3 le |119909|

(5)







(6)



RSF

(7)


119905SSC = (119905RAW minus (

119891DC (119905RAW)




fDCcoef1198911

2

(9)


Imag

e dist

ortio

n in

azim

uth

Distorted fringes


Last line



13 km



8x

Runway

20 m


Look direction








M 2010-03-24 S 2010-02-19

M 2010-03-24 S 2010-05-29

M 2010-03-24 S 2010-03-02

M 2010-03-24 S 2010-06-20 M 2010-03-24 S 2010-07-12 M 2010-03-24 S 2010-08-03 M 2010-03-24 S 2010-08-25

M 2010-03-24 S 2010-05-07M 2010-03-24 S 2010-04-15

M 2010-03-24 S 2010-12-24M 2010-03-24 S 2010-11-10

M 2010-03-24 S 2011-04-13 M 2010-03-24 S 2011-05-16 M 2010-03-24 S 2011-05-27

M 2010-03-24 S 2011-02-06M 2010-03-24 S 2010-10-08

M 2010-03-24 S 2011-03-22

M 2010-03-24 S 2010-09-16

M 2010-03-24 S 2011-02-28

Master2010-03-24

minus7∘C

+5∘C

+11∘C minus7

∘Cminus7∘C0

∘C

0∘C

+13∘C

+7∘C

+1∘C

+16∘C

0∘C

+3∘C

+2∘C

+13∘C

minus7∘C +5

∘C +5∘C

+5∘C

+13∘C









M 2010-07-16 S 2009-12-08 M 2010-07-16 S 2010-01-10 M 2010-07-16 S 2010-04-08

M 2010-07-16 S 2010-06-13

M 2010-07-16 S 2010-09-20

M 2010-07-16 S 2010-12-17

M 2010-07-16 S 2011-02-21

M 2010-07-16 S 2010-06-02

M 2010-07-16 S 2010-08-29

M 2010-07-16 S 2010-11-25

M 2010-07-16 S 2011-01-30

M 2010-07-16 S 2010-05-22

M 2010-07-16 S 2010-06-24

M 2010-07-16 S 2010-11-03

M 2010-07-16 S 2010-01-08








Look direction


Look direction





5 Conclusion












0

05

1

15

2

25

3

35times104

0 02 04 06 08 1Coherence

Cou

nt

TSX

TDM






Acknowledgments



References





























RoboticsJournal of





Journal of



RotatingMachinery





VLSI Design



Shock and Vibration







Journal of



Volume 2014


SensorsJournal of





Propagation











Antenna

Ground track

Ground samples


Forward

Backward

Image footprint




Coregistration


Removefrequency

Reinsertfrequency





Range (pixel)

Azi

mut

h (ti

me)


Image framePixel 11

Pixel max max

sim+3700Hz

simminus3700Hz




0

1000

2000

3000

4000

5000 6000minus4000

minus3000

minus2000

minus1000

0 1000 2000 3000 4000Lines

fDC

(Hz)








119894 (119909) =

(120572 minus 120573 + 2) |119909|

3

minus (120572 minus 120573 + 3) |119909|

2

+ 1

0 le |119909| lt 1

120572|119909|

3

minus (5120572 minus 120573) |119909|

2


1 le |119909| lt 2

120573|119909|

3

minus (8120573) |119909|

2

+ (21120573) |119909| minus (18120573)

2 le |119909| lt 3

0 3 le |119909|

(5)







(6)



RSF

(7)


119905SSC = (119905RAW minus (

119891DC (119905RAW)




fDCcoef1198911

2

(9)


Imag

e dist

ortio

n in

azim

uth

Distorted fringes


Last line



13 km



8x

Runway

20 m


Look direction








M 2010-03-24 S 2010-02-19

M 2010-03-24 S 2010-05-29

M 2010-03-24 S 2010-03-02

M 2010-03-24 S 2010-06-20 M 2010-03-24 S 2010-07-12 M 2010-03-24 S 2010-08-03 M 2010-03-24 S 2010-08-25

M 2010-03-24 S 2010-05-07M 2010-03-24 S 2010-04-15

M 2010-03-24 S 2010-12-24M 2010-03-24 S 2010-11-10

M 2010-03-24 S 2011-04-13 M 2010-03-24 S 2011-05-16 M 2010-03-24 S 2011-05-27

M 2010-03-24 S 2011-02-06M 2010-03-24 S 2010-10-08

M 2010-03-24 S 2011-03-22

M 2010-03-24 S 2010-09-16

M 2010-03-24 S 2011-02-28

Master2010-03-24

minus7∘C

+5∘C

+11∘C minus7

∘Cminus7∘C0

∘C

0∘C

+13∘C

+7∘C

+1∘C

+16∘C

0∘C

+3∘C

+2∘C

+13∘C

minus7∘C +5

∘C +5∘C

+5∘C

+13∘C









M 2010-07-16 S 2009-12-08 M 2010-07-16 S 2010-01-10 M 2010-07-16 S 2010-04-08

M 2010-07-16 S 2010-06-13

M 2010-07-16 S 2010-09-20

M 2010-07-16 S 2010-12-17

M 2010-07-16 S 2011-02-21

M 2010-07-16 S 2010-06-02

M 2010-07-16 S 2010-08-29

M 2010-07-16 S 2010-11-25

M 2010-07-16 S 2011-01-30

M 2010-07-16 S 2010-05-22

M 2010-07-16 S 2010-06-24

M 2010-07-16 S 2010-11-03

M 2010-07-16 S 2010-01-08








Look direction


Look direction





5 Conclusion












0

05

1

15

2

25

3

35times104

0 02 04 06 08 1Coherence

Cou

nt

TSX

TDM






Acknowledgments



References





























RoboticsJournal of





Journal of



RotatingMachinery





VLSI Design



Shock and Vibration







Journal of



Volume 2014


SensorsJournal of





Propagation










0

1000

2000

3000

4000

5000 6000minus4000

minus3000

minus2000

minus1000

0 1000 2000 3000 4000Lines

fDC

(Hz)








119894 (119909) =

(120572 minus 120573 + 2) |119909|

3

minus (120572 minus 120573 + 3) |119909|

2

+ 1

0 le |119909| lt 1

120572|119909|

3

minus (5120572 minus 120573) |119909|

2


1 le |119909| lt 2

120573|119909|

3

minus (8120573) |119909|

2

+ (21120573) |119909| minus (18120573)

2 le |119909| lt 3

0 3 le |119909|

(5)







(6)



RSF

(7)


119905SSC = (119905RAW minus (

119891DC (119905RAW)




fDCcoef1198911

2

(9)


Imag

e dist

ortio

n in

azim

uth

Distorted fringes


Last line



13 km



8x

Runway

20 m


Look direction








M 2010-03-24 S 2010-02-19

M 2010-03-24 S 2010-05-29

M 2010-03-24 S 2010-03-02

M 2010-03-24 S 2010-06-20 M 2010-03-24 S 2010-07-12 M 2010-03-24 S 2010-08-03 M 2010-03-24 S 2010-08-25

M 2010-03-24 S 2010-05-07M 2010-03-24 S 2010-04-15

M 2010-03-24 S 2010-12-24M 2010-03-24 S 2010-11-10

M 2010-03-24 S 2011-04-13 M 2010-03-24 S 2011-05-16 M 2010-03-24 S 2011-05-27

M 2010-03-24 S 2011-02-06M 2010-03-24 S 2010-10-08

M 2010-03-24 S 2011-03-22

M 2010-03-24 S 2010-09-16

M 2010-03-24 S 2011-02-28

Master2010-03-24

minus7∘C

+5∘C

+11∘C minus7

∘Cminus7∘C0

∘C

0∘C

+13∘C

+7∘C

+1∘C

+16∘C

0∘C

+3∘C

+2∘C

+13∘C

minus7∘C +5

∘C +5∘C

+5∘C

+13∘C









M 2010-07-16 S 2009-12-08 M 2010-07-16 S 2010-01-10 M 2010-07-16 S 2010-04-08

M 2010-07-16 S 2010-06-13

M 2010-07-16 S 2010-09-20

M 2010-07-16 S 2010-12-17

M 2010-07-16 S 2011-02-21

M 2010-07-16 S 2010-06-02

M 2010-07-16 S 2010-08-29

M 2010-07-16 S 2010-11-25

M 2010-07-16 S 2011-01-30

M 2010-07-16 S 2010-05-22

M 2010-07-16 S 2010-06-24

M 2010-07-16 S 2010-11-03

M 2010-07-16 S 2010-01-08








Look direction


Look direction





5 Conclusion












0

05

1

15

2

25

3

35times104

0 02 04 06 08 1Coherence

Cou

nt

TSX

TDM






Acknowledgments



References





























RoboticsJournal of





Journal of



RotatingMachinery





VLSI Design



Shock and Vibration







Journal of



Volume 2014


SensorsJournal of





Propagation










Imag

e dist

ortio

n in

azim

uth

Distorted fringes


Last line



13 km



8x

Runway

20 m


Look direction








M 2010-03-24 S 2010-02-19

M 2010-03-24 S 2010-05-29

M 2010-03-24 S 2010-03-02

M 2010-03-24 S 2010-06-20 M 2010-03-24 S 2010-07-12 M 2010-03-24 S 2010-08-03 M 2010-03-24 S 2010-08-25

M 2010-03-24 S 2010-05-07M 2010-03-24 S 2010-04-15

M 2010-03-24 S 2010-12-24M 2010-03-24 S 2010-11-10

M 2010-03-24 S 2011-04-13 M 2010-03-24 S 2011-05-16 M 2010-03-24 S 2011-05-27

M 2010-03-24 S 2011-02-06M 2010-03-24 S 2010-10-08

M 2010-03-24 S 2011-03-22

M 2010-03-24 S 2010-09-16

M 2010-03-24 S 2011-02-28

Master2010-03-24

minus7∘C

+5∘C

+11∘C minus7

∘Cminus7∘C0

∘C

0∘C

+13∘C

+7∘C

+1∘C

+16∘C

0∘C

+3∘C

+2∘C

+13∘C

minus7∘C +5

∘C +5∘C

+5∘C

+13∘C









M 2010-07-16 S 2009-12-08 M 2010-07-16 S 2010-01-10 M 2010-07-16 S 2010-04-08

M 2010-07-16 S 2010-06-13

M 2010-07-16 S 2010-09-20

M 2010-07-16 S 2010-12-17

M 2010-07-16 S 2011-02-21

M 2010-07-16 S 2010-06-02

M 2010-07-16 S 2010-08-29

M 2010-07-16 S 2010-11-25

M 2010-07-16 S 2011-01-30

M 2010-07-16 S 2010-05-22

M 2010-07-16 S 2010-06-24

M 2010-07-16 S 2010-11-03

M 2010-07-16 S 2010-01-08








Look direction


Look direction





5 Conclusion












0

05

1

15

2

25

3

35times104

0 02 04 06 08 1Coherence

Cou

nt

TSX

TDM






Acknowledgments



References





























RoboticsJournal of





Journal of



RotatingMachinery





VLSI Design



Shock and Vibration







Journal of



Volume 2014


SensorsJournal of





Propagation










M 2010-03-24 S 2010-02-19

M 2010-03-24 S 2010-05-29

M 2010-03-24 S 2010-03-02

M 2010-03-24 S 2010-06-20 M 2010-03-24 S 2010-07-12 M 2010-03-24 S 2010-08-03 M 2010-03-24 S 2010-08-25

M 2010-03-24 S 2010-05-07M 2010-03-24 S 2010-04-15

M 2010-03-24 S 2010-12-24M 2010-03-24 S 2010-11-10

M 2010-03-24 S 2011-04-13 M 2010-03-24 S 2011-05-16 M 2010-03-24 S 2011-05-27

M 2010-03-24 S 2011-02-06M 2010-03-24 S 2010-10-08

M 2010-03-24 S 2011-03-22

M 2010-03-24 S 2010-09-16

M 2010-03-24 S 2011-02-28

Master2010-03-24

minus7∘C

+5∘C

+11∘C minus7

∘Cminus7∘C0

∘C

0∘C

+13∘C

+7∘C

+1∘C

+16∘C

0∘C

+3∘C

+2∘C

+13∘C

minus7∘C +5

∘C +5∘C

+5∘C

+13∘C









M 2010-07-16 S 2009-12-08 M 2010-07-16 S 2010-01-10 M 2010-07-16 S 2010-04-08

M 2010-07-16 S 2010-06-13

M 2010-07-16 S 2010-09-20

M 2010-07-16 S 2010-12-17

M 2010-07-16 S 2011-02-21

M 2010-07-16 S 2010-06-02

M 2010-07-16 S 2010-08-29

M 2010-07-16 S 2010-11-25

M 2010-07-16 S 2011-01-30

M 2010-07-16 S 2010-05-22

M 2010-07-16 S 2010-06-24

M 2010-07-16 S 2010-11-03

M 2010-07-16 S 2010-01-08








Look direction


Look direction





5 Conclusion












0

05

1

15

2

25

3

35times104

0 02 04 06 08 1Coherence

Cou

nt

TSX

TDM






Acknowledgments



References





























RoboticsJournal of





Journal of



RotatingMachinery





VLSI Design



Shock and Vibration







Journal of



Volume 2014


SensorsJournal of





Propagation










M 2010-07-16 S 2009-12-08 M 2010-07-16 S 2010-01-10 M 2010-07-16 S 2010-04-08

M 2010-07-16 S 2010-06-13

M 2010-07-16 S 2010-09-20

M 2010-07-16 S 2010-12-17

M 2010-07-16 S 2011-02-21

M 2010-07-16 S 2010-06-02

M 2010-07-16 S 2010-08-29

M 2010-07-16 S 2010-11-25

M 2010-07-16 S 2011-01-30

M 2010-07-16 S 2010-05-22

M 2010-07-16 S 2010-06-24

M 2010-07-16 S 2010-11-03

M 2010-07-16 S 2010-01-08








Look direction


Look direction





5 Conclusion












0

05

1

15

2

25

3

35times104

0 02 04 06 08 1Coherence

Cou

nt

TSX

TDM






Acknowledgments



References





























RoboticsJournal of





Journal of



RotatingMachinery





VLSI Design



Shock and Vibration







Journal of



Volume 2014


SensorsJournal of





Propagation










Look direction


Look direction





5 Conclusion












0

05

1

15

2

25

3

35times104

0 02 04 06 08 1Coherence

Cou

nt

TSX

TDM






Acknowledgments



References





























RoboticsJournal of





Journal of



RotatingMachinery





VLSI Design



Shock and Vibration







Journal of



Volume 2014


SensorsJournal of





Propagation



















0

05

1

15

2

25

3

35times104

0 02 04 06 08 1Coherence

Cou

nt

TSX

TDM






Acknowledgments



References





























RoboticsJournal of





Journal of



RotatingMachinery





VLSI Design



Shock and Vibration







Journal of



Volume 2014


SensorsJournal of





Propagation










0

05

1

15

2

25

3

35times104

0 02 04 06 08 1Coherence

Cou

nt

TSX

TDM






Acknowledgments



References





























RoboticsJournal of





Journal of



RotatingMachinery





VLSI Design



Shock and Vibration







Journal of



Volume 2014


SensorsJournal of





Propagation




















RoboticsJournal of





Journal of



RotatingMachinery





VLSI Design



Shock and Vibration







Journal of



Volume 2014


SensorsJournal of





Propagation











RoboticsJournal of





Journal of



RotatingMachinery





VLSI Design



Shock and Vibration







Journal of



Volume 2014


SensorsJournal of





Propagation









Documents

Research Article Using Open-Source Components to Process ...downloads.hindawi.com/journals/ijap/2013/275635.pdf · Using Open-Source Components to Process Interferometric TerraSAR-X