Fast Target Detection Framework for Onboard Processing of - - PowerPoint PPT Presentation

▶
fast target detection framework for onboard processing of
SMART_READER_LITE
LIVE PREVIEW

Fast Target Detection Framework for Onboard Processing of - - PowerPoint PPT Presentation

Aug 18, 2022 193 likes •410 views

Applied Research LLC Fast Target Detection Framework for Onboard Processing of Multispectral and Hyperspectral Images B. Ayhan and C. Kwan June 3, 2015 Applied Research LLC 9605 Medical Center Dr., Rockville, MD20850 Research supported by

slide-1
SLIDE 1

Applied Research LLC 1

Fast Target Detection Framework for Onboard Processing of Multispectral and Hyperspectral Images

  • B. Ayhan and C. Kwan

June 3, 2015 Applied Research LLC 9605 Medical Center Dr., Rockville, MD20850 Research supported by NASA SBIR Program

slide-2
SLIDE 2

Applied Research LLC 2

  • 1. Contents
  • 1. Contents

2

  • 2. Research Objectives

3

  • 3. Technical Approach

4-8

  • 4. Results

9-20

slide-3
SLIDE 3

Applied Research LLC 3

  • 2. Research Objectives

3

  • Develop a robust, automated, and real-time target detection

system under varying illumination, atmospheric conditions and target/sensor viewing geometry.

  • Demonstrate the feasibility of the system using actual and/or

simulated data.

slide-4
SLIDE 4

Applied Research LLC 4

Fast target detection framework

  • Conventional target detection is done in the reflectance domain: a lot of

computations due to atmospheric compensation, not suitable for onboard processing, difficult to change mission goals during mission.

  • Our approach is done in the radiance domain. Only a few target signatures

(reflectance) need to be transformed to the radiance domain. This is very suitable for onboard processing such as search and rescue missions.

  • JHU/APL developed a similar approach that uses MODTRAN and AFWA MM5.

Our approach was motivated by [1], which is a hybrid framework that uses MODTRAN and a nonlinear analytical model.

  • 3. Technical Approach

[*1] “Hyperspectral material identification on radiance data using single-atmosphere or multiple-atmosphere modeling," Adrian V. Mariano ; John M. Grossmann, J. Appl. Remote Sens. 4(1), 043563 (November 23, 2010).

slide-5
SLIDE 5

Applied Research LLC 5

  • Radiance equation model parameter estimation using

MODTRAN outputs [*1]

  • 3. Technical Approach

1 1

A A A

B A L P D S S ρ ρ α ρ ρ ρ = + + − − −

ρ

: Material reflectance

A

ρ

: Adjacent region reflectance

S : Spherical albedo

A,B : Coefficients that depend on atmospheric, geometric and solar illumination conditions P

: Path radiance

D

: Radiance due to direct solar illumination

α : Amount of solar occlusion

[*1] “Hyperspectral material identification on radiance data using single-atmosphere or multiple-atmosphere modeling," Adrian V. Mariano ; John M. Grossmann, J. Appl. Remote Sens. 4(1), 043563 (November 23, 2010).

L: Radiance

slide-6
SLIDE 6

Applied Research LLC 6

  • Radiance equation model parameter estimation using

MODTRAN outputs [*1]

  • 3. Technical Approach

1

0.05 ρ =

2

0.6 ρ =

MODTRAN Simulation (1)

MODTRAN Output "DRCT_REFL(1)"

Atmospheric, geometric and solar illumination conditions

MODTRAN Output " GRND_RFLT(1) "

Estimation of radiance equation parameters (A,B, D,P and S)

1 1

A A A

B A L P D S S ρ ρ α ρ ρ ρ = + + − − −

MODTRAN Output " SOL_SCAT (1) "

MODTRAN Simulation (2)

MODTRAN Output "DRCT_REFL(2)" MODTRAN Output " GRND_RFLT(2)" MODTRAN Output " SOL_SCAT(2) "

DRCT_RFLT: Direct Reflectance (MODTRAN output) SOL_SCAT: Solar Multiple Scattering (MODTRAN output) GRND_RFLT: Ground Reflectance (MODTRAN output)

[*1] “Hyperspectral material identification on radiance data using single-atmosphere or multiple-atmosphere modeling," Adrian V. Mariano ; John M. Grossmann, J. Appl. Remote Sens. 4(1), 043563 (November 23, 2010).

slide-7
SLIDE 7

Applied Research LLC 7

  • Radiance equation model parameter estimation using

MODTRAN outputs [*1]

  • 3. Technical Approach

1 1

A A A

B A L P D S S ρ ρ α ρ ρ ρ = + + − − −

( ) ( )

1 2

DRCT_RFLT 1 / DRCT_RFLT / 2 D ρ ρ = =

1 1 1 1 1 1

, 1 1 A B G C P S S

ρ ρ

ρ ρ ρ ρ = = + − −

2 2 2 2 2 2

, 1 1 A B G C P S S

ρ ρ

ρ ρ ρ ρ = = + − −

2 2 1 1 2 1

/ / G G S G G

ρ ρ ρ ρ

ρ ρ − = −

2 2 2

G A G S

ρ ρ

ρ = −

1 2 2 2 1 1 2 1

( ) / / 1/ 1/ S C C C C P

ρ ρ ρ ρ

ρ ρ ρ ρ − + − = −

1 1

1 ( )( ) B C P S

ρ

ρ = − − Suppose , and

1 1

_ (1) and _ (1) C G SOL SCAT GRND RFLT

ρ ρ

= =

2 2

_ (2) and _ (2) C G SOL SCAT GRND RFLT

ρ ρ

= = Then, and The radiance model parameters can then be found as:

[*1] “Hyperspectral material identification on radiance data using single-atmosphere or multiple-atmosphere modeling," Adrian V. Mariano ; John M. Grossmann, J. Appl. Remote Sens. 4(1), 043563 (November 23, 2010).

slide-8
SLIDE 8

Applied Research LLC

Proprietary Information - ARLLC

8

  • 3. Technical Approach

Advantages of the proposed system

  • Eliminates the need of applying atmospheric correction on the whole

image cube and instead simulates the variants of the radiance signature of the target of interest and searches for these signatures in the test radiance image cube

  • The effects of different illumination, atmospheric conditions,
  • cclusion and varying sensor/target viewing geometries are taken

into effect during the simulation of the radiance spectral profiles of the target of interest

  • Allows generation of look-up tables for several radiance signature

variants of the target which will reduce computation/processing time for target detection in operations like “search and rescue” that require quick on-board decisions

slide-9
SLIDE 9

Applied Research LLC 9

  • Demonstration of radiance model parameter estimation and

simulating radiance profiles with the model parameter estimates

Model Tropical IHAZE Rural VIS 5 km H2OSTR 0.5 Altitude 1 km Approximate observer position Latitude: 39.3305º (N), Longitude: 76.2879 (W) Date August 29, 1995 Time data collected 18:37 UTC

Atmospheric, solar illumination and geometric location parameters used in two MODTRAN runs to estimate model parameters (S, A, P, D, B)

500 1000 1500 2000 2500 3000 0.05 0.1 0.15 0.2 0.25 0.3 0.35 0.4 Frequency (nm) Amplitude S 500 1000 1500 2000 2500 3000 2 4 6 8 10 12 14 16 Frequency (nm) Amplitude A 500 1000 1500 2000 2500 3000 0.5 1 1.5 2 2.5 3 3.5 Frequency (nm) Amplitude P 500 1000 1500 2000 2500 3000 1 2 3 4 5 6 7 Frequency (nm) Amplitude D 500 1000 1500 2000 2500 3000 2 4 6 8 10 12 14 16 18 Frequency (nm) Amplitude B

S P A D B Plots of estimated model parameters (S, A, P, D, B)

  • 4. Results
slide-10
SLIDE 10

Applied Research LLC 10

  • Comparing the simulated radiance profiles (from the model

parameter estimates) with the MODTRAN results

0.4 0.6 0.8 1 1.2 1.4 1.6 1.8 2 2.2 2.4 0.1 0.2 0.3 0.4 0.5 Reflectance Data Wavelength (µm) R e f le c ta n c e 1 tree1000010x2Easd0x2Eref

Case 2 (reflectance of a green tree) Case 1 (fixed reflectance of 0.6)

0.4 0.6 0.8 1 1.2 1.4 1.6 1.8 2 2.2 2.4 0.2 0.4 0.6 0.8 1 Wavelength (µm) Reflectance

MODTRAN Radiance model with estimated parameters MODTRAN Radiance model with estimated parameters

500 1000 1500 2000 2500 3000 5 10 15 20 25 Wavelength (nm) Radiance MODTRAN generated radiance Radiance estimated with model parameters 500 1000 1500 2000 2500 3000 2 4 6 8 10 12 Wavelength (nm) Radiance MODTRAN generated radiance Radiance estimated with model parameters

The results are almost identical which shows that radiance model parameter estimation is successful !

  • 4. Results
slide-11
SLIDE 11

Applied Research LLC 11

  • AVIRIS data for Los Angeles Station Fire (Aug 2009) to detect

burnscar

The Station Fire took place between August 26 and October 16, and a total of 160,577 acres (251 sq mi; 650 km2) were affected, 209 structures had been destroyed, including 89 homes [*2]. It first started in the Angeles National Forest near the U.S. Forest Service ranger station on the Angeles Crest Highway (State Highway 2) [*2].

[*2] http://en.wikipedia.org/wiki/2009_California_wildfires

AVIRIS images acquired on October 6, 2009 (fire is mostly over)

  • 4. Results
slide-12
SLIDE 12

Applied Research LLC 12

  • Getting groundtruth of burned locations for AVIRIS data using

MODIS MCD45A1 product

500 1000 1500 2000 500 1000 1500 2000 20 40 60 80 100 120 10 20 30 40 50 60 70 80 90

MODIS H8-V5 tile (Los Angeles region falls into this tile) MCD45A1 burned area product for this tile for Sep 2009 (red pixels indicate burned areas) Zoom into the region for the LA Station Fire in MCD45A1 product Based on AVIRIS image data coverage for each strip the groundtruth maps for burned area are extracted

  • 4. Results
slide-13
SLIDE 13

Applied Research LLC 13

  • Burnscar pixel selections from AVIRIS image strips

Direct sunlight: With occlusion:

Based on extracted groundtruth maps and also visual examination, burnscar pixels are selected from each AVIRIS image data (pixels with direct sunlight and under shadow with some

  • cclusion)

r09 r10 r11 r12 r13 r14 r15

  • 4. Results
slide-14
SLIDE 14

Applied Research LLC 14

  • Radiance and reflectance signatures of selected burnscar pixels

Actual radiance signatures of selected pixels (in DN)

5 1 1 5 2 2 5 3 2 4 6 8 1 1 2 1 4 1 6 1 8 2 2 2 W a v e le n g th (n m ) R a d ia n ce r0 9 (S a m p le :3 8 9 , L in e :1 9 2 9 ) r0 9 (S a m p le :5 7 5 , L in e :1 9 5 5 ) r1 (S a m p le :3 4 3 7 , L in e :2 8 8 7 ) r1 (S a m p le :6 6 5 L in e :2 7 7 8 ) r1 1 (S a m p le :4 9 8 , L in e :2 4 3 ) r1 1 (S a m p le :5 3 2 , L in e :2 2 9 8 ) r1 2 (S a m p le :3 7 7 , L in e :1 9 9 7 ) r1 2 (S a m p le :3 9 8 , L in e :1 9 5 3 ) r1 3 (S a m p le :1 8 2 , L in e :2 1 9 7 ) r1 3 (S a m p le :1 6 3 , L in e :2 2 3 5 ) r1 4 (S a m p le :5 1 9 , L in e :3 2 3 2 ) r1 4 (S a m p le :5 3 7 , L in e :3 2 8 ) r1 5 (S a m p le :6 1 9 , L in e :4 1 3 ) 2 3 4 5 6 7 5 1 1 5 2 M O D IS b a n d in d e x R adiance (7 M O D IS bands) r0 9 (Sa m p le :3 8 9 , L in e :1 9 2 9 ) r0 9 (Sa m p le :5 7 5 , L in e :1 9 5 5 ) r1 (Sa m p le :3 4 3 7 , L in e :2 8 8 7 ) r1 (Sa m p le :6 6 5 L in e :2 7 7 8 ) r1 1 (Sa m p le :4 9 8 , L in e :2 4 3 ) r1 1 (Sa m p le :5 3 2 , L in e :2 2 9 8 ) r1 2 (Sa m p le :3 7 7 , L in e :1 9 9 7 ) r1 2 (Sa m p le :3 9 8 , L in e :1 9 5 3 ) r1 3 (Sa m p le :1 8 2 , L in e :2 1 9 7 ) r1 3 (Sa m p le :1 6 3 , L in e :2 2 3 5 ) r1 4 (Sa m p le :5 1 9 , L in e :3 2 3 2 ) r1 4 (Sa m p le :5 3 7 , L in e :3 2 8 ) r1 5 (Sa m p le :6 1 9 , L in e :4 1 3 )

Actual radiance signatures of selected pixels (in DN) for MODIS bands

  • nly

MODIS bands: 648 nm, 858 nm, 470 nm, 555 nm, 1240 nm, 1640 nm, and 2130 nm (In target detection investigations in this work, we used the 7 bands that are closest to these MODIS bands )

5 1 1 5 2 2 5 .1 .2 .3 .4 .5 .6 .7 .8 .9 1 W a v e le n g th (n m ) R e fle ctan ce r0 9 (S a m p le :3 8 9 , L in e :1 9 2 9 ) r0 9 (S a m p le :5 7 5 , L in e :1 9 5 5 ) r1 (S a m p le :3 4 3 7 , L in e :2 8 8 7 ) r1 (S a m p le :6 6 5 L in e :2 7 7 8 ) r1 1 (S a m p le :4 9 8 , L in e :2 4 3 ) r1 1 (S a m p le :5 3 2 , L in e :2 2 9 8 ) r1 2 (S a m p le :3 7 7 , L in e :1 9 9 7 ) r1 2 (S a m p le :3 9 8 , L in e :1 9 5 3 ) r1 3 (S a m p le :1 8 2 , L in e :2 1 9 7 ) r1 3 (S a m p le :1 6 3 , L in e :2 2 3 5 ) r1 4 (S a m p le :5 1 9 , L in e :3 2 3 2 ) r1 4 (S a m p le :5 3 7 , L in e :3 2 8 ) r1 5 (S a m p le :6 1 9 , L in e :4 1 3 )

Reflectance signatures of selected pixels (after QUAC atmospheric correction)

1 2 3 4 5 6 7 0.1 0.2 0.3 0.4 0.5 0.6 0.7 Wavelength (nm) Reflectance r09 (Sample:389, Line:1929) r09 (Sample:575, Line:1955) r10 (Sample:3437, Line:2887) r10 (Sample:665 Line:2778) r11 (Sample:498, Line:2043) r11 (Sample:532, Line:2298) r12 (Sample:377, Line:1997) r12 (Sample:398, Line:1953) r13 (Sample:182, Line:2197) r13 (Sample:163, Line:2235) r14 (Sample:519, Line:3232) r14 (Sample:537, Line:3280) r15 (Sample:619, Line:4013)

Reflectance signatures of selected pixels after QUAC atmospheric correction (MODIS bands

  • nly)
  • 4. Results
slide-15
SLIDE 15

Applied Research LLC

500 1000 1500 2000 2500 0.1 0.2 0.3 0.4 0.5 0.6 0.7 Wavelength (nm) Reflectance r09 (Sample:389, Line:1929) r09 (Sample:575, Line:1955) r10 (Sample:3437, Line:2887) r10 (Sample:665 Line:2778) r11 (Sample:498, Line:2043) r11 (Sample:532, Line:2298) r12 (Sample:377, Line:1997) r12 (Sample:398, Line:1953) r13 (Sample:182, Line:2197) r13 (Sample:163, Line:2235) r14 (Sample:519, Line:3232) r14 (Sample:537, Line:3280) r15 (Sample:619, Line:4013)

15

  • Atmospheric, illumination and geometric location conditions

used in radiance model parameter estimation and simulation of radiance signatures for the selected reflectance profiles

Geometric locations of the AVIRIS image strips and corresponding UTC values

AVIRIS Image Strip Min Lat Min Long Max Lat Max Long UTC Ground Elevation (km) r09 34.1436

  • 118.376

34.5639

  • 118.276

18:51 0.67 r10 34.1488

  • 118.31

34.6222

  • 118.208

19:01 0.93 r11 34.152

  • 118.236

34.5625

  • 118.133

19:11 0.98 r12 34.1571

  • 118.165

34.5788

  • 118.066

19:21 1.11 r13 34.1603

  • 118.091

34.5626

  • 117.99

19:29 1.22 r14 34.1575

  • 118.022

34.5984

  • 117.924

19:39 1.17 r15 34.1634

  • 117.943

34.563

  • 117.847

19:48 1.18 Model Mid Latitude Summer (MODEL = 2) VIS (Visibility) 20 km H2OSTR (Water vapor) 0.270 IHAZE Rural (IHAZE =1) Altitude 14 km Observer position (see above table for observer positions) Data collection day Oct 6, 2009 Data collection time (see above table for UTC times)

Atmospheric and time/date conditions Variants of selected burnscar reflectance profiles (based on QUAC)

  • 4. Results
slide-16
SLIDE 16

Applied Research LLC 16

  • Simulated radiance signatures with estimated radiance model

parameters and target detection results

5 1 1 5 2 2 5 5 1 1 5 2 2 5 3 W a v e le n g th (n m ) R a d ia n ce r0 9 (S a m p le :3 8 9 , L in e :1 9 2 9 ) r0 9 (S a m p le :5 7 5 , L in e :1 9 5 5 ) r1 (S a m p le :3 4 3 7 , L in e :2 8 8 7 ) r1 (S a m p le :6 6 5 L in e :2 7 7 8 ) r1 1 (S a m p le :4 9 8 , L in e :2 4 3 ) r1 1 (S a m p le :5 3 2 , L in e :2 2 9 8 ) r1 2 (S a m p le :3 7 7 , L in e :1 9 9 7 ) r1 2 (S a m p le :3 9 8 , L in e :1 9 5 3 ) r1 3 (S a m p le :1 8 2 , L in e :2 1 9 7 ) r1 3 (S a m p le :1 6 3 , L in e :2 2 3 5 ) r1 4 (S a m p le :5 1 9 , L in e :3 2 3 2 ) r1 4 (S a m p le :5 3 7 , L in e :3 2 8 ) r1 5 (S a m p le :6 1 9 , L in e :4 1 3 ) 1 2 3 4 5 6 7 5 1 1 5 2 2 5 3 3 5 M O D IS b a n d in d e x n

  • R

a d ia n ce M O D T R A N

  • g

e n e ra te d ra d ia n ce sig n a tu re s (M O D IS b a n d s o n ly) r0 9 (S a m p le :3 8 9 , L in e :1 9 2 9 ) r0 9 (S a m p le :5 7 5 , L in e :1 9 5 5 ) r1 (S a m p le :3 4 3 7 , L in e :2 8 8 7 ) r1 (S a m p le :6 6 5 L in e :2 7 7 8 ) r1 1 (S a m p le :4 9 8 , L in e :2 4 3 ) r1 1 (S a m p le :5 3 2 , L in e :2 2 9 8 ) r1 2 (S a m p le :3 7 7 , L in e :1 9 9 7 ) r1 2 (S a m p le :3 9 8 , L in e :1 9 5 3 ) r1 3 (S a m p le :1 8 2 , L in e :2 1 9 7 ) r1 3 (S a m p le :1 6 3 , L in e :2 2 3 5 ) r1 4 (S a m p le :5 1 9 , L in e :3 2 3 2 ) r1 4 (S a m p le :5 3 7 , L in e :3 2 8 ) r1 5 (S a m p le :6 1 9 , L in e :4 1 3 )

  • =

− j i j i j i

SAM r s r s r s , cos ) , (

1 1 2

M-SAM( ,burnscar) min( ( , ), ( , ),..., ( , ))

i i i i K

SAM SAM SAM = s s r s r s r

1 2

where burnscar { , ,..., }

K

= r r r

Simulated radiance signatures for the selected pixels Simulated radiance signatures for the selected pixels (MODIS bands

  • nly)

Target detection score images with M-SAM for "r09" using simulated radiance signatures, using actual radiance signatures and using actual reflectance signatures with QUAC (groundtruth map for “r09” image is also shown). In target detection, we used the 7 MODIS bands only.

test pixel spectral profile

i =

s burnscar signature variant

j =

r

  • 4. Phase 1 Results
slide-17
SLIDE 17

Applied Research LLC 17

  • ROC curves with M-SAM detection technique for AVIRIS images

.1 .2 .3 .4 .5 .6 .7 .8 .9 1 .1 .2 .3 .4 .5 .6 .7 .8 .9 1 F a lse a la rm ra te P ro b a b ility o f d e te ctio n r0 9 A ctu a l ra d ia n ce sig n a tu re s w ith M

  • S

A M A ctu a l re fle cta n ce sig n a tu re s w ith M

  • S

A M M O D T R A N

  • sim

u la te d ra d ia n ce sig n a tu re s w ith M

  • S

A M .1 .2 .3 .4 .5 .6 .7 .8 .9 1 .1 .2 .3 .4 .5 .6 .7 .8 .9 1 F a lse a la rm ra te P ro b a b ility o f d e te ctio n r1 A ctu a l ra d ia n ce sig n a tu re s w ith M

  • S

A M A ctu a l re fle cta n ce sig n a tu re s w ith M

  • S

A M M O D T R A N

  • sim

u la te d ra d ia n ce sig n a tu re s w ith M

  • S

A M .1 .2 .3 .4 .5 .6 .7 .8 .9 1 .1 .2 .3 .4 .5 .6 .7 .8 .9 1 F a lse a la rm ra te P ro b a b ility o f d e te ctio n r1 1 A ctu a l ra d ia n ce sig n a tu re s w ith M

  • S

A M A ctu a l re fle cta n ce sig n a tu re s w ith M

  • S

A M M O D T R A N

  • sim

u la te d ra d ia n ce sig n a tu re s w ith M

  • S

A M .1 .2 .3 .4 .5 .6 .7 .8 .9 1 .1 .2 .3 .4 .5 .6 .7 .8 .9 1 F a lse a la rm ra te P ro b a b ility o f d e te ctio n r1 2 A ctu a l ra d ia n ce sig n a tu re s w ith M

  • S

A M A ctu a l re fle cta n ce sig n a tu re s w ith M

  • S

A M M O D T R A N

  • sim

u la te d ra d ia n ce sig n a tu re s w ith M

  • S

A M

r09 r11 r10 r12

  • 4. Results
slide-18
SLIDE 18

Applied Research LLC 18

  • ROC curves with M-SAM detection technique

.1 .2 .3 .4 .5 .6 .7 .8 .9 1 .1 .2 .3 .4 .5 .6 .7 .8 .9 1 F a lse a la rm ra te P ro b a b ility o f d e te ctio n r1 3 A ctu a l ra d ia n ce sig n a tu re s w ith M

  • S

A M A ctu a l re fle cta n ce sig n a tu re s w ith M

  • S

A M M O D T R A N

  • sim

u la te d ra d ia n ce sig n a tu re s w ith M

  • S

A M .1 .2 .3 .4 .5 .6 .7 .8 .9 1 .1 .2 .3 .4 .5 .6 .7 .8 .9 1 F a lse a la rm ra te P ro b a b ility o f d e te ctio n r1 4 A ctu a l ra d ia n ce sig n a tu re s w ith M

  • S

A M A ctu a l re fle cta n ce sig n a tu re s w ith M

  • S

A M M O D T R A N

  • sim

u la te d ra d ia n ce sig n a tu re s w ith M

  • S

A M

r13 r14

  • 4. Results

.1 .2 .3 .4 .5 .6 .7 .8 .9 1 .1 .2 .3 .4 .5 .6 .7 .8 .9 1 F a lse a la rm ra te P ro b a b ility o f d e te ctio n r1 5 A ctu a l ra d ia n ce sig n a tu re s w ith M

  • S

A M A ctu a l re fle cta n ce sig n a tu re s w ith M

  • S

A M M O D T R A N

  • sim

u la te d ra d ia n ce sig n a tu re s w ith M

  • S

A M

r15

slide-19
SLIDE 19

Applied Research LLC 19

  • AUC (Area under curve) measures from ROC curves

AVIRIS Filename Analysis Type Using actual radiance signatures (target detection in radiance domain) Using QUAC reflectance signatures (target detection in reflectance domain) Using simulated radiance signatures (reflectances are from QUAC) (target detection in radiance domain ) r09 AUC(overall) 0.8958 0.8856 0.8935 AUC(partial) 0.0761 0.0750 0.0745 r10 AUC(overall) 0.9395 0.9183 0.9400 AUC(partial) 0.0655 0.0612 0.0653 r11 AUC(overall) 0.9028 0.8814 0.9087 AUC(partial) 0.0427 0.0472 0.0471 r12 AUC(overall) 0.8696 0.8487 0.8783 AUC(partial) 0.0473 0.0544 0.0514 r13 AUC(overall) 0.8772 0.8601 0.8771 AUC(partial) 0.0691 0.0679 0.0676 r14 AUC(overall) 0.8370 0.8055 0.8415 AUC(partial) 0.0648 0.0631 0.0659 r15 AUC(overall) 0.8536 0.7794 0.8713 AUC(partial) 0.0676 0.0585 0.0687

Overall: Whole ROC curve is used in AUC computation Partial: ROC curve up to a false alarm rate of 0.1 is used in AUC computation

  • 4. Results
slide-20
SLIDE 20

Applied Research LLC 20

  • Observations
  • With the simulated radiance signature library using estimated

radiance models, we observed detection performances close to the detection performance obtained with using actual radiance signatures that are retrieved from actual images and in some cases slightly better than those.

  • It is possible that if a better set of MODTRAN parameters has

been selected in the radiance model parameter estimation that fully complies with the atmospheric conditions during the actual data acquisition, or a more realistic set of reflectance signature variations for burnscar has been used, the detection performances could perhaps be further improved.

  • These results are found highly promising in demonstrating the

feasibility and effectiveness of the proposed fast target detection idea.

  • 4. Results