Data compression for light-field rendering

Two light-eld compression schemes are presented. The codecs are compared with regard to compression eÆciency and rendering performance. The rst proposed coder is based on video-compression techniques that have been modied to code the four-dimensional light-eld data structure eÆciently. The second coder relies entirely on disparity-compensated image prediction, establishing a hierarchical structure among the light-eld images. Coding performance of both schemes is evaluated using publicly available light elds of synthetic as well as real-world scenes. Compression ratios vary between 100:1 and 2000:1, depending on reconstruction quality and light-eld scene characteristics. I