Proof of Concept to Enhance Video Using Static Images
Video as a medium has been evolving quickly in the high-speed internet era. Handheld cameras are becoming more affordable as technology becomes cheaper and more widespread. However, limitations still exist which make capturing the perfect video difficult. Bad lighting, stabilization issues, unexpected objects in the video, and other video troubles can ruin an amazing shot - unless you have the expensive editing equipment to fix it.
It seems that an inexpensive and simple way to fix common issues with video shots has been developed. A demo was given Eurographics Symposium on Rendering in 2007. By developing a rendering system that is able to scan the depth of video and images - the team was able to use still images to enhance different aspects of a video; such as increasing the high dynamic range, exposure, and removing entire objects from the video.
http://www.vimeo.com/1513129These capabilities are enabled by two technical contributions presented in this paper. First, we make several improvements to a state-of-the-art multiview stereo algorithm in order to compute view-dependent depths using video, photographs, and structure-from-motion data. Second, we present a novel image-based rendering algorithm that can re-render the input video using the appearance of the photographs while preserving certain temporal dynamics such as specularities and dynamic scene lighting.
They even offer the source-code on their website for tweaking yourself. Hopefully some programs come out soon for users to utilize this technology for their video. This could be a huge boost in video quality for amateur directors or film students.
