Case StudiesWe respect the privacy of our clients, most of which do not want their applications discussed publicly. Some, however, have graciously allowed us to provide details of our work for them.
Project: Enhanced Long-Range Imaging Problem: Long-range imaging is a critical component to many NASA applications including range surveillance, launch tracking, and astronomical observation. However, significant degradation occurs when imaging through the Earth’s atmosphere. The subsequent effects of poor image quality range from inconvenient to dangerous depending on the application. Solution: EM Photonics is developing an embedded system based on field-programmable gate array (FPGA) technology capable of enhancing long-range images and videos by compensating for atmosphere induced distortions. This solver processes incoming video streams in real-time for a variety of formats, including the high-definition version used by NASA. The resulting device is light-weight and low-power and can be integrated with current video collection, viewing, and recording equipment. It can be used to process data 1) as it is collected (in real-time) or 2) from previously recorded imagery and deployed with camera systems or 3) in data centers depending on the application. Additionally, since this processing unit is built on FPGA technology, it can easily be extended to perform a variety of other tasks such as compression, encryption, or further processing.
Project: Optical Device Design Problem: Next-generation night-vision devices require the incorporation of advanced optical and photonic components. As these components get larger, more complicated, and include smaller scale features, the design and simulation times become incredibly long. In some cases, designs cannot be analyzed because of the computational resources required to perform the simulations. Solution: EM Photonics built an accelerated simulation engine for optical and photonic devices based on the finite-difference time-domain (FDTD) method. This method allowed for a rigorous analysis of the devices of interest – a key requirement since reduced order or approximate methods did not provide the necessary level of accuracy. In order to reduce simulation times to allow for reasonable designs to be analyzed, EM Photonics built the solver on a field-programmable gate array (FPGA) platform. The custom designed device readily integrates with a standard desktop PC and is currently sold packaged with Optiwave’s award-winning CAD software. This solver has since been transitioned to run on graphics processing units (GPUs) – the computational heart of modern graphics cards, further improving performance.
|
|||||