Image processing improves tank vision
Smart fusing of data from multiple image sensors can improve military driving, said a European consortium.
Several image sensor types are used in military applications: daylight sensors (mostly CCD), image intensifier tubes for low light, and thermal (short, medium and long wavelength).
Each has its own advantages and disadvantages, and some sort of mixing can be useful.
“The typical solution is to blend daylight and thermal imaging, but this results in a loss of contrast and definition relative to the original daylight image,” said Isle of Wight DSP firm RFEL, which is providing an FPGA-based processing box (‘Halo’, pictured) and image processing algorithms. “Image fusion between two sensors is completely different from simple image blending. A fusion algorithm processes the images captured by two sensors, generally from different wavelengths, and is able to maintain the best attributes from both input images for the resulting output. This, in combination with image stabilisation, provides the user with the best view possible for informed decisions.”
RFEL has tied up with KTK Kommunikationstechnik, GuS and Rheinmetall Defence to offer a mix-and-match modular system called ‘fused driver vision enhancement’ (fDVE) – based on a standardised mechanical and electric interfaces to cater for different kinds of image sensors.
As it plugs together, units can be shared around vehicles, and combined differently for different missions.
Other algorithms allow: video scaling, contrast enhancement, picture-in-picture and lens distortion correction.
See fDVE at Eurosatory.