Lahaie P.,DRDC Valcartier
Proceedings of SPIE - The International Society for Optical Engineering | Year: 2014
The atmospheric correction of thermal hyperspectral imagery can be separated in two distinct processes: Atmospheric Compensation (AC) and Temperature and Emissivity separation (TES). TES requires for input at each pixel, the ground leaving radiance and the atmospheric downwelling irradiance, which are the outputs of the AC process. The extraction from imagery of the downwelling irradiance requires assumptions about some of the pixels' nature, the sensor and the atmosphere. Another difficulty is that, often the sensor's spectral response is not well characterized. To deal with this unknown, we defined a spectral mean operator that is used to filter the ground leaving radiance and a computation of the downwelling irradiance from MODTRAN. A user will select a number of pixels in the image for which the emissivity is assumed to be known. The emissivity of these pixels is assumed to be smooth and that the only spectrally fast varying variable in the downwelling irradiance. Using these assumptions we built an algorithm to estimate the downwelling irradiance. The algorithm is used on all the selected pixels. The estimated irradiance is the average on the spectral channels of the resulting computation. The algorithm performs well in simulation and results are shown for errors in the assumed emissivity and for errors in the atmospheric profiles. The sensor noise influences mainly the required number of pixels. © 2014 SPIE.
Ross V.,AEREX Avionique Inc. |
Dion D.,DRDC Valcartier
Proceedings of SPIE - The International Society for Optical Engineering | Year: 2011
A new C++ library for radiative transfer calculations in the visible and infrared bands which uses MODTRAN as a primary source for atmospheric optical parameters has been developed at Defense R&D Canada, Valcartier (DRDC Valcartier). The main benefit of the library is its capability to perform fast wide spectral band calculations with an appreciably high accuracy. Coherent calculations on wide bands are made possible by using a modified version of the correlated-k theory. The main features of the library are discussed, and comparisons with conventional spectral MODTRAN 4 calculations are presented. It is shown that the library is capable of producing band results that are usually within 5% of MODTRAN 4 with computation times that are thousands of times faster. © 2011 SPIE.
Boudreau S.,Laval University |
Levasseur S.,Laval University |
Perilla C.,Laval University |
Roy S.,DRDC Valcartier |
Genest J.,Laval University
Optics Express | Year: 2013
High-resolution spectral lidar measurements using dual frequency combs as a source is presented. The technique enables the rangeresolved measurement of fine spectral features, such as gas absorption lines, provided that a suitable scatterer is present in the scene. Measurements of HCN absorption lines at 20 meters are presented, with a water droplet cloud and a diffusely reflective surface as scatterers. © 2013 Optical Society of America.
Berger J.,DRDC Valcartier |
Lo N.,T OptLogic Ltd.
Computers and Operations Research | Year: 2015
Search and rescue path planning is known to be computationally hard, and most techniques developed to solve practical size problems have been unsuccessful to estimate an optimality gap. A mixed-integer linear programming (MIP) formulation is proposed to optimally solve the multi-agent discrete search and rescue (SAR) path planning problem, maximizing cumulative probability of success in detecting a target. It extends a single agent decision model to a multi-agent setting capturing anticipated feedback information resulting from possible observation outcomes during projected path execution while expanding possible agent actions to all possible neighboring move directions, considerably augmenting computational complexity. A network representation is further exploited to alleviate problem modeling, constraint specification, and speed-up computation. The proposed MIP approach uses CPLEX problem-solving technology in promptly providing near-optimal solutions for realistic problems, while offering a robust upper bound derived from Lagrangean integrality constraint relaxation. Modeling extension to a closed-loop environment to incorporate real-time action outcomes over a receding time horizon can even be envisioned given acceptable run-time performance. A generalized parameter-driven objective function is then proposed and discussed to suitably define a variety of user-defined objectives. Computational results reporting the performance of the approach clearly show its value. © 2014 Published by Elsevier Ltd. All rights reserved.
Dery B.,Laval University |
Buteau S.,DRDC Valcartier |
Simard J.-R.,DRDC Valcartier |
Bouchard J.-P.,LInstitut National dOptique |
Vallee R.,Laval University
IEEE Transactions on Geoscience and Remote Sensing | Year: 2010
A method has been developed to correlate spectral signatures obtained with various fluorescence LIght Detection And Ranging (LIDAR) systems. A calibrated fluorescence reference target was used to calibrate the spectral response of the LIDAR transmitter channels and obtain their transfer functions. Two LIDAR systems have been spectrally characterized, and corrected signatures for two bioaerosols are presented. The first LIDAR system is the Standoff Integrated Bioserosol Active Hyperspectral Detection field LIDAR developed by Defence R&D Canada. This standoff system uses a 351-nm pulsed laser in a monoaxial design. The second system is a lab-sized aerosol chamber designed to characterize fluorescent aerosols under controlled environmental conditions. The chamber was designed according to classical short-range biaxial LIDAR principles, with the purpose of duplicating the results obtained with field LIDAR systems. Aerosols generated within the chamber are probed by a 355-nm pulsed laser, and autofluorescence spectra are measured with a spectrometer and an intensified charge-coupled device camera. This chamber is used to collect the reference spectra of various fluorescing aerosols and simulants of biological agents. One of the main objectives in using this chamber is to produce and compile a library of instrument-free fluorescing spectra that can be transferred to other LIDAR-based bioaerosol sensors with known optical transfer functions. © 2006 IEEE.