Pattern recognition for high performance imaging

High performance imaging technology generates images with high spectral and spatial resolution, high dynamic range, and/or at high speed. Hyperspectral images contain tens or hundreds of contiguous wavelength indexed bands that are related to material information. High spatial resolution images provide fine details on target objects. High dynamic range images present a great range of luminance levels to capture vivid lights or shadows. High speed cameras offer high frame rate to record fast moving objects. Sometimes, high performance imaging can also be achieved by combing the output of a large number of imaging devices. While high performance imaging has greatly expanded the sensing capability of cameras to capture scenes or phenomena that are beyond human vision, the processing, analysis and understanding of these imaging modalities are still challenging, with many unsolved problems. In particular, various types of high performance images have their unique properties, and are normally in very larger size. As a consequence, though the state-of-the-art pattern recognition techniques have achieved great success on traditional grayscale and color images, for example, in object detection and image classification, they cannot be directly applied to high performance images. On the other hand, this also brings new opportunities to the research community, as there are strong needs to develop effective and efficient methods for a variety of pattern recognition tasks on these images. The goal of this special issue is to provide a forum for researchers and practitioners in the broad computer vision and pattern recognition community to present their novel and original pattern recognition research for high performance imaging. We hope this special issue will become an enlightening and useful source on high performance imaging research, and also for wider research community in pattern recognition.