David R. Gerwe (Boeing), Paul Menicucci (Boeing)
Keywords: Imaging
Abstract:
An image processor is discussed that combines many types of image enhancement onto a single compact electronics card. The current enhancements include bad pixel compensation, focal plane array non-uniformity correction, and several stages of contrast enhancement, feature sharpening, superresolution, and image motion stabilization. Though there are certainly better algorithms for particular applications, this mixture of algorithms reliably enables the system to substantially improve image quality for a large variety of sensors, platforms, and imaging geometries. The card design hosted an FPGA and microprocessor facilitated rapid development by allowing many complicated algorithm elements to be quickly coded in C, with the FPGA providing horsepower for simpler but more computationally intensive elements. Examples show the quality improvement gained by compensating for image degradations including camera motion, atmospheric turbulence induced blur, focal plane imperfections, camera pixel density, and noise.
Date of Conference: September 1-4. 2009
Track: Imaging