Fusion of Radar and Electro-Optical/Infrared (EO/IR) for Ship Classification and Identification

TECHNOLOGY AREA(S): Weapons OBJECTIVE: Develop an innovative approach that exploits new methodology in machine learning and modern mobile computing devices to fuse information obtained from different sensor types in order to achieve dramatic improvement in target classification and identification capability for space, weight and power (SWaP) constrained platforms. DESCRIPTION: Radar, electronic support measures (ESM), a.k.a. anti-radiation homing (ARH), and electro-optical (EO)/imaging infrared (IIR)/laser detection and ranging (LIDAR) currently provide different sensor phenomenology that can lead to different salient feature manifestation that depends on operating conditions (e.g., acquisition geometry) and scene content type. Current technology approaches develop automatic target recognition (ATR) systems for a single sensor, each designed to exploit the salient features specific to each sensor type, which leads to suboptimal classification performance for each sensor type and not a higher confidence performance by combining independent sensor data into a single solution. The capability to combine the salient feature information from the different sensors to get improved target classification, and possibly identification, of the ships is needed. Recent advances in machine learning can be explored to discover and to fuse the different feature information inherent within the different sensor types while advances in mobile computing processors enables these machine learning approaches to work efficiently and robustly in real-time. The algorithms should be designed for execution on mobile processors, including multi-core system-on-a-chip (SoC) systems, combining general purpose computing elements (multi-core Advanced Reduced-Instruction-Set-Computer Machines (ARM) processor), with on-chip co-processors as multi-core graphical processing units (GPUs) and/or field-programmable gate arrays (FPGAs). Work produced in Phase II may become classified. Note: The prospective contractor(s) must be U.S. owned and operated with no foreign influence as defined by DoD 5220.22-M, National Industrial Security Program Operating Manual, unless acceptable mitigating procedures can and have been implemented and approved by the Defense Security Service (DSS). The selected contractor and/or subcontractor must be able to acquire and maintain a secret level facility and Personnel Security Clearances, in order to perform on advanced phases of this project as set forth by DSS and NAVAIR in order to gain access to classified information pertaining to the national defense of the United States and its allies; this will be an inherent requirement. The selected company will be required to safeguard classified material IAW DoD 5220.22-M during the advanced phases of this contract. PHASE I: Develop an efficient and robust approach based on state-of-the art machine learning technology to extract and fuse information from radar, ESM, and EO/IIR/LIDAR. Define the data specifics desired for each sensor type and provide a format including meta data needs. The principle modes of interest are radar Inverse Synthetic Aperture Radar (ISAR) images with IIR images. Demonstrate the feasibility of your approach utilizing a laptop computer for the algorithm to obtain a 5 Hz solution rate, and analyze the detailed mapping and simulation of the new algorithm onto candidate military compatible processors as dictated by the PMA. PHASE II: Develop and optimize the real-time embedded software code of the machine learning fusion algorithm developed in Phase I for the candidate processor selected. Work with the government team to test the algorithms against data collected from candidate sensors relevant to the Navy. Pertinent information will be provided to performer if necessary. PHASE III: Develop the modifications to the algorithm and real-time code to be hosted in the transition Program of Record as desired by the Navy. Support modeling and simulation efforts as well as software integration, field testing and performance analysis in the specific application. Maritime activities such as the Coast Guard, Shipping monitoring, Homeland Security, that have the need to know what ship traffic exists can benefit from this technology. The basic core of the algorithms and fusion may apply to land-based commercial vehicle tracking as well. REFERENCES: 1. Li, H. & Zhou, Y.T., (1996). SAR/IR Sensor Fusion and Real-time Implementation. 1996 29th Asilomar Conference on Signals, Systems and Computers (2 Volume Set (Asilomar Conference on Signals, Systems and Computers//Conference Record). https://www.amazon.com/Asilomar-Conference-Signals-Systems-Computers/dp/08186737022. Recognition of SAR Target Based on Multilayer Auto-Encoder and SNN; by Sun et al; International Journal of Innovative Computing, Information and Control Vol 9, Number 11, November 2013 ISSN 1349-4198. http://www.ijicic.org/ijicic-12-11029.pdf KEYWORDS: Target Recognition; Multi-sensor Fusion; Machine Learning; Radar; IIR; Maritime Identification