This chapter presents multi-sensor fusion techniques for motion deblurring. With recent advances in digital imaging, the use of high resolution, high-speed, or high dynamic range cameras has become common practice. However, thus far no single image sensor can satisfy the diverse requirements of all the current industrial camera applications. For example, high-speed (HS) cameras can capture fast motion with little motion blur but require expensive sensors, bandwidth, and storage. The image resolution in HS cameras is often much lower than many commercial still cameras. This is mainly because the image resolution needs to scale linearly with the exposure time (Ben-Ezra & Nayar 2003) to maintain the signal-to-noise ratio (SNR), i.e. a higher speed maps to a lower resolution. In addition, the relatively low bandwidth on the usual interfaces such as USB 2.0 or FireWire IEEE 1394a restricts the image resolution especially when streaming videos at 100–200 fps.

The problem of acquiring high quality imagery with little motion under low light is particularly challenging. To guarantee enough exposures, one can choose to use either a wide aperture or a slow shutter. For example, by coupling a wide aperture with fast shutters, we can capture fast motions of scene objects with low noise. However, wide apertures lead to a shallow depth-of-field (DoF) where only parts of the scene can be clearly focused. In contrast, by coupling a slow shutter with a narrow aperture, one can capture all depth layers in focus.