|
Imaging through Fog
Navy SBIR 2016.1 - Topic N161-055 ONR - Ms. Lore-Anne Ponirakis - [email protected] Opens: January 11, 2016 - Closes: February 17, 2016 N161-055 TITLE: Imaging through Fog TECHNOLOGY AREA(S): Information Systems, Sensors ACQUISITION PROGRAM: Combined Electrooptic Surveillance and Response System (CESARS), FY16-02 FN OBJECTIVE: Develop and demonstrate a passive Electro-Optical (EO)/Infrared (IR) - EO-IR - imaging system that employs coordinated, jointly optimized, acquisition and processing of multi-modal (spectral, temporal, polarization, quantum etc.) data to enhance the operational range by 10X compared to a single-mode imaging system in the presence of obscurant. DESCRIPTION: The US Fleet Forces are often present in congested waterways throughout the world for a variety of humanitarian and military purposes. To maintain situational awareness (SA) and to support target detection, tracking, and identification, electro-optical (EO) and infrared (IR) sensors could be employed for their superior resolution and image-forming mode of operation, in contrast to radar. However, the short wavelengths associated with EO-IR make imaging far more susceptible to performance degradation from scattering by ubiquitous water-based aerosols, which typically generate a large, non-information carrying, background radiation that overwhelms the ballistic signals that do carry information about the scene. Imaging through dense fog is the intrinsic hard problem, as a strongly scattering medium fills the entire working volume. Imaging through cloud layers [1] or haze [2] or fog [3] can be improved by exploiting prior information in processing, but the deleterious effect of the scattering medium on the signal-to-background ratio remains the key limitation, which depends on the image acquisition mode as well as the optical properties of the water-based aerosols. While active imaging techniques, which typically employ structured laser-light illumination and/or temporal gating, have been employed somewhat successfully to enhance image acquisition, passive imaging provides fewer degrees of freedom to manipulate. Changes in atmospheric conditions, ambient illumination conditions, and orientation-dependent scene reflection/emission characteristics further conspire to complicate passive-imaging enhancement. However, by employing, for example, a judicious choice of spectral band(s), polarization diversity, high-speed multi-frame acquisition, or other mode of acquisition together with advanced processing techniques, significant improvements can potentially be achieved, especially for a specific class of obscurants. For example, select spectral bands may provide enhanced transmission, while ballistic photons may possess different average polarization states compared to scattered photons. Processing techniques can optimally exploit these enhancements to extract additional scene information. Although the gain in image quality associated with a single degree of freedom may be modest, an overall improvement obtained by combining multiple, optimized degrees of freedom, may be more substantial. Therefore, a multi modal hardware solution combined with coordinated processing techniques may enable an imaging system to be realized that provides a substantial overall operational improvement. This topic seeks to develop a passive EO-IR imaging system that employs jointly optimized multi-modal image acquisition and processing to increase operational range by 10X over baseline range (to be selected by performer) that corresponds to traditional single-mode image acquisition in the presence of obscurant. Solutions can exploit all or any portion of the electromagnetic spectrum ranging from the UV to the far IR, including the conventional bands referred to as electro-optic (EO), near-infrared (NIR), short-wave IR (SWIR), mid-wave IR (MWIR), and long-wave IR (LWIR), but excluding mm-wave bands. System designs that employ either innovative sensors or Commercial-off-the-shelf (COTS) components are both of interest, along with data fusion techniques and advanced algorithms. While systems having low size, weight, and power (e.g., < 1 cu. ft., <30 lbs., and <300 W) are of interest, larger systems will also be considered. The overall objective is to achieve a 10X enhancement to substantially improve situational awareness, target detection, tracking, and ID tasks in presence of strongly scattering medium. PHASE I: Determine feasibility, design and simulate a multi-modal EO-IR system with jointly optimized sensing and processing to achieve a 10X improvement in operational range compared to single-mode operation in the presence of obscurant. Identify key risk elements to achieving the 10X improvement objective and perform suitable simulations and/or experiments to mitigate these risk factors. Prepare a publication-quality technical document detailing the system design and performance characteristics, which should include an analysis of the proposed system relative to the current state-of-the art. PHASE II: Construct and demonstrate the multi-modal EO-IR system with associated processing designed in Phase I. Specifically, conduct quantitative measurements and analysis to verify the purported 10X improvement in operational range. The experimental validation can be performed in a laboratory simulated environment that is realistic representation of a shipboard environment. Prepare a document based on results from Phase II that follows the standards of a publication in refereed journal. PHASE III DUAL USE APPLICATIONS: Extend the technology to a full system prototype by optimizing the hardware and processing demonstrated in Phase II. Refine the design to minimize size, weight and power consumption while introducing mechanical robustness against shock and vibration. Detailed specifications will be provided by the Navy during Phase III. Provide support in transitioning the technology and qualifying its use for the Navy. The small business will provide user�s manuals and training materials to the Navy. This technology will have applications in all services of the military, in law enforcement, civilian, and commercial sectors. Anywhere enhanced long-range imagery is needed constitutes a relevant application of this technology. Applications might include unmanned aerial surveillance (UAS) systems, precision agriculture, oil and gas pipeline inspection, disaster relief, or search and rescue. Drug interdiction in the maritime environment via unmanned or manned surface or aerial vehicles is another example application. REFERENCES: 1. Sermsak Jaruwatanadilok, Akira Ishimaru, and Yasuo Kuga, "Optical Imaging Through Clouds and Fog," IEEE Trans. On Geoscience and Remote Sensing, vol. 41, no. 8, August 2003, pp. 1834-1843 2. Kaiming He, Jian Sun, Xiaoou Tang, "Single Image Haze Removal Using Dark Channel Prior," in Proc. Of Int�l. Conf. on Computer Vision and Pattern Recognition 2009. 3. Jing Yu and Qingmin Liao, "Fast Single Image Fog Removal Using Edge-Preserving Smoothing" IEEE Int�l conf. on Acoustics, Speech, and Signal Processing (ICASSP) 2011, Prague, Czech pp. 1245-1248. KEYWORDS: High dynamic range imaging, fog, electro-optical, infrared, polarization, multi-spectral, sensor fusion, autonomous, real-time, advanced processing, intelligence, surveillance, reconnaissance, situational awareness. TPOC-1: Ravindra Athale Email: [email protected] TPOC-2: James Waterman Email: [email protected] Questions may also be submitted through DoD SBIR/STTR SITIS website.
|