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 NASA SBIR Program 1
Applied Research LLC 1. Contents 1. Contents 2 2. Research Objectives 3 3. Technical Approach 4-8 4. Results 9-20 2
Applied Research LLC 2. Research Objectives • 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. 3 3
Applied Research LLC 3. Technical Approach 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. [*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). 4
Applied Research LLC 3. Technical Approach • Radiance equation model parameter estimation using MODTRAN outputs [*1] A ρ B ρ A L = + + P − α D ρ 1 − ρ S 1 − ρ S A A L : Radiance ρ : Material reflectance ρ : Adjacent region reflectance A 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). 5
Applied Research LLC 3. Technical Approach • Radiance equation model parameter estimation using MODTRAN outputs [*1] DRCT_RFLT: Direct Reflectance (MODTRAN output) Atmospheric, GRND_RFLT: Ground Reflectance (MODTRAN output) geometric and solar SOL_SCAT: Solar Multiple Scattering (MODTRAN output) illumination conditions MODTRAN Output "DRCT_REFL(1)" MODTRAN MODTRAN Output Simulation ρ = 0.05 " GRND_RFLT(1) " 1 Estimation of (1) MODTRAN Output radiance equation " SOL_SCAT (1) " parameters (A,B, D,P and S) MODTRAN Output A ρ B ρ A L = + + P − α D ρ "DRCT_REFL(2)" 1 − ρ S 1 − ρ S MODTRAN A A MODTRAN Output Simulation ρ = 0.6 " GRND_RFLT(2)" 2 (2) MODTRAN Output " SOL_SCAT(2) " [*1] “Hyperspectral material identification on radiance data using single-atmosphere or multiple-atmosphere 6 modeling," Adrian V. Mariano ; John M. Grossmann, J. Appl. Remote Sens. 4(1), 043563 (November 23, 2010).
Applied Research LLC 3. Technical Approach • Radiance equation model parameter estimation using MODTRAN outputs [*1] A ρ B ρ A L = + + P − α D ρ 1 − ρ S 1 − ρ S A A C = SOL _ SCAT (1) and G = GRND _ RFLT (1) Suppose , and ρ 1 ρ 1 C = SOL _ SCAT (2) and G = GRND _ RFLT (2) ρ 2 ρ 2 A ρ B ρ A ρ B ρ 2 2 1 1 Then, and G = , C = + P G = , C = + P 2 2 ρ ρ ρ 1 ρ 1 1 − ρ S 1 − ρ S 1 − ρ S 1 − ρ S 2 2 1 1 The radiance model parameters can then be found as: ( ) ( ) D = DRCT_RFLT 1 / ρ = DRCT_RFLT 2 / ρ 1 2 G / ρ − G / ρ 1 ρ 2 2 ρ 1 1 B = ( C − P )( − S ) S = ρ 1 ρ G − G 1 ρ 2 ρ 1 S C ( − C ) + C / ρ − C / ρ G ρ 1 ρ 2 ρ 2 2 ρ 1 1 ρ 2 P = A = − G S ρ 2 1/ ρ − 1/ ρ ρ 2 1 2 [*1] “Hyperspectral material identification on radiance data using single-atmosphere or multiple-atmosphere 7 modeling," Adrian V. Mariano ; John M. Grossmann, J. Appl. Remote Sens. 4(1), 043563 (November 23, 2010).
Applied Research LLC 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, occlusion 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 8 Proprietary Information - ARLLC
Applied Research LLC 4. Results • Demonstration of radiance model parameter estimation and simulating radiance profiles with the model parameter estimates Model Tropical Atmospheric, solar IHAZE Rural illumination and geometric VIS 5 km location parameters used in H2OSTR 0.5 two MODTRAN runs to Altitude 1 km estimate model parameters Approximate observer position Latitude: 39.3305º (N), Longitude: 76.2879 (W) (S, A, P, D, B) Date August 29, 1995 Time data collected 18:37 UTC 0.4 16 S A S A 0.35 14 Plots of 0.3 12 estimated model 0.25 10 Amplitude parameters Amplitude 0.2 8 (S, A, P, D, B) 0.15 6 0.1 4 0.05 2 0 0 0 500 1000 1500 2000 2500 3000 0 500 1000 1500 2000 2500 3000 Frequency (nm) Frequency (nm) 7 18 3.5 D P P B B D 16 6 3 14 5 2.5 12 2 4 Amplitude Amplitude Amplitude 10 1.5 3 8 6 1 2 4 0.5 1 2 0 0 500 1000 1500 2000 2500 3000 0 0 0 500 1000 1500 2000 2500 3000 9 Frequency (nm) 0 500 1000 1500 2000 2500 3000 Frequency (nm) Frequency (nm)
Applied Research LLC 4. Results • Comparing the simulated radiance profiles (from the model parameter estimates) with the MODTRAN results 25 MODTRAN generated radiance Case 1 (fixed reflectance of 0.6) Radiance estimated with model parameters 20 1 0.8 MODTRAN 15 Radiance Reflectance 0.6 0.4 10 Radiance model with 0.2 estimated parameters 5 0 0.4 0.6 0.8 1 1.2 1.4 1.6 1.8 2 2.2 2.4 Wavelength ( µ m) 0 0 500 1000 1500 2000 2500 3000 Wavelength (nm) 12 Case 2 (reflectance of a green tree) MODTRAN generated radiance Radiance estimated with model parameters Reflectance Data 10 1 tree1000010x2Easd0x2Eref 0.5 8 MODTRAN 0.4 Radiance R e f le c ta n c e 6 0.3 0.2 Radiance model with 4 estimated parameters 0.1 2 0 0.4 0.6 0.8 1 1.2 1.4 1.6 1.8 2 2.2 2.4 Wavelength ( µ m) 0 0 500 1000 1500 2000 2500 3000 Wavelength (nm) The results are almost identical which shows that radiance model parameter estimation is successful ! 10
Applied Research LLC 4. Results • 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 km 2 ) 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]. AVIRIS images acquired on October 6, 2009 (fire is mostly over) 11 [*2] http://en.wikipedia.org/wiki/2009_California_wildfires
Applied Research LLC 4. Results • Getting groundtruth of burned locations for AVIRIS data using MODIS MCD45A1 product Zoom into the region for the LA MCD45A1 burned area product for MODIS H8-V5 tile (Los Station Fire in MCD45A1 product this tile for Sep 2009 (red pixels Angeles region falls into indicate burned areas) this tile) 10 20 500 30 40 1000 50 60 1500 70 80 2000 90 20 40 60 80 100 120 500 1000 1500 2000 Based on AVIRIS image data coverage for each strip the groundtruth maps for burned area are extracted 12
Applied Research LLC 4. Results • Burnscar pixel selections from AVIRIS image strips 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 occlusion) Direct sunlight: With occlusion: r09 r10 r11 r12 r13 r14 r15 13
Recommend
More recommend
