SBIR-STTR Award

Improved quality & comprehensiveness of measured UV-VIS-IR threat signatures of military targets (vehicles, ordnance, and weapons) is needed by developing an expert technology that improves the performance of conventional multispectral and advanced hyperspectral (HS) imager systems. Threat signatures must be characterized in all three of the spatial + temporal + spectral domains thus the large DoD investment in multispectral imagers, which measure all three domains in a single instrument. However, as requirements for increased resolution in time, space, and wavelength have been pushed to ever higher values, the performance of conventional multispectral (MS) imagers has reached the photon starvation limit (Refs. 1-3). A dramatically new technology is needed. This topic solicits an expert technology to combine the signature data stream from a conventional MS sensor with adjunct sensors (imagers, radiometers, CVF spectrometer, etc.) in order to achieve enhanced spatial, temporal, and spectral performance that is substantially beyond the limits achievable by a stand-alone MS imager. Open-source literature and prior SBIR efforts (Refs. 4-8) have addressed only the spatial resolution. To ensure widespread application to DoD and allow for growth as electro-optical (EO) sensor technology evolves, the approach must not be hard-wired to a specific type of MS/HS imager or to a specific suite of adjunct EO sensors. Instead, the technology must adapt to rapidly reconfigure existing EO assets (UV-Vis-IR cameras, MS imager) to meet new test requirements as they arise. This comprehensive level of sensor fusion for multispectral imagers (spatial + spectral + temporal), together with the requirement for flexibility, have never been attempted. The benefit to the war fighter is comprehensive high fidelity target signature measurements (spatial, temporal, and spectral) for battlefield simulation, target recognition, scientific and technical data collections, and improved threat signature measurements for aircraft warning receivers (missiles and hostile fire). The benefit to the DoD test measurement community is the ability to significantly enhance the performance of existing multispectral imager systems - regardless of design, and without the need to purchase expensive new hardware.

Keywords:
Fusion, Remote Sensing, Hyperspectral, Sensors, Pan-Sharpening, Registration

Field measurement campaigns typically deploy numerous sensors having different spatial, temporal, and spectral sampling characteristics. This makes it difficult to process experiment data when a phenomenon of interest spans multiple sensors. There is a clear need for a new software tool that can process sequences of sensor imagery in disparate formats and produce a time sequence of data accurately resampled to a common grid. We demonstrated end-to-end processing performance: 1) cross-sensor spatial registration via scene-based methods; 2) temporal upsampling to common time base via optical flow interpolation; and 3) advanced “fused” image products (e.g. pan sharpened imagery). Our proposed Phase II effort will culminate in a functional prototype software tool.

Benefit:
The first planned product that will incorporate the proposed technology is a Windows Application for ingestion, registering, resampling and fusing multiple imager camera data. This prototype will ingest common format data sets and produce industry standard output data products, including ENVI™-compatible data files and standard image and video format imagery (e.g. JPEG, MPEG). This product is intended to demonstrate the capabilities and provide a lead-in to development of custom solutions. We anticipate that probable customers for this initial product include researchers in Government laboratories, Prime Government contractors, Academic researchers, medical imaging device OEM manufacturers and research facilities, biotechnology researchers and commercial providers of manufacturing control and monitoring equipment. Commercial applications include monitoring of manufacturing processes, factories and equipment as well as diagnostic imaging equipment used in the biotechnology and medical device community. This work will benefit the Government directly by providing the processing needed to combine multiple sensor data sets from imagers typically used by AEDC and other government facilities to monitor system and device testing, as well as monitoring and launches.

Keywords:
Multisensor, Fusion, Multispectral, Registration, Optical Flow