September 2006 1

SAR Processing

Jeremy Nicoll

September 2006 2

Outline

• Sonar Echoes • Definitions• What do we measure?• Target detection• Convolution• Range compression• Azimuth Compression

September 2006 3

Hello!

Sonar Echoes

September 2006 4

Some definitions

Azimuth

Range

Swath

September 2006 5

What do we measure?

• At the satellite– Radar signal strength as a function of time

• After processing– Radar Cross Section per piece of dirt

September 2006 6

Target detection

September 2006 7

Target detection:Short pulse

• Power problem– We need a short pulse for resolution– We need a lot of power for detection

September 2006 8

Target detection:Long pulse

• Power problem– We need a short pulse for resolution– We need a lot of power for detection

September 2006 9

Target detection:Chirp

• Can go from low to high or high to low frequency• Can be much more complicated

September 2006 10

Target detection:Chirp pulse

September 2006 11

Convolution

• Run simulation

September 2006 12

Convolution:Constructive interference

27 28 29 30 31 32

ReceiveConvolutionTransmit

September 2006 13

Convolution:Destructive interference

27 28 29 30 31 32

ReceiveConvolutionTransmit

September 2006 14

Convolution

ReceiveConvolutionTransmit

September 2006 15

Range Compression

• Convolve the received range signal with the chirp you sent out.

• Range line looks similar to that of a short pulse –We have “compressed” the chirps in the signal to short pulses.

September 2006 16

Azimuth:Real Aperture Radar

• Good resolution in range, poor resolution in azimuth

September 2006 17

Azimuth:The Doppler effect

September 2006 18

Azimuth compression

• Carl Wiley, in 1951, observed that two point targets at different azimuth positions, will have distinct Doppler frequency shifts. [Curlander, McDonough, 1991]

• We can compress the pulse in azimuth in the same fashion we compressed it in range.

• Convolve an azimuth line with the azimuth reference function:

September 2006 19

Synthetic Aperture Radar

• Raw data

• Range compressed

• Azimuth compressed

September 2006 20

Range compression

• each element of the raw data is multiplied by the complex conjugate of the corresponding element of the range reference function

September 2006 21

Range compression

September 2006 22

Azimuth reference function

September 2006 23

Range cell migration

• range cell migration realigns all the returns for a single target into an appropriate single line of data

September 2006 24

Azimuth compression

• each element of the data is multiplied by the complex conjugate of the corresponding element of the azimuth reference function

September 2006 25

Range compression

September 2006 26

SAR training processor

September 2006 27

Input file selection menu

Choice of start lineprocessing data patch

Choice of Doppler parameters for processing

September 2006 28

Choice of performing steps during processing

Select to save individualintermediate product

Select to save allintermediate products

September 2006 29

Processing flow withdetailed description of respective processing step

September 2006 30

Preview of intermediateproduct