Extrapolative Lightspace Method for HDR Video Exposure Selection
Sarang Nerkar, Ryan E. Janzen, Pete Scourboutakos, Steve Mann · 2016
We propose a method to automatically adjust multiple exposure-value settings for HDR video compositing. The method uses imagespace-to-lightspace post-conversion and tonal extrapolation to iteratively select optimal exposure settings, as opposed to present systems, which use manually set settings. The limiting factor becomes the image sensor as opposed to the implementation of the algorithm, as in present systems. By choosing the exposure settings with the described algorithm, the high dynamic range sampling process can be adapted to various lighting environments. This algorithm is especially useful for ultra low power capture of optimally selected exposures which can be processed later using well-known HDR compositing methods.