ABSTRACT
The extra information within an HDR image means that the resultant data files are large. Floating point representations, which were introduced in Chapter 2, can achieve a reduction down to 32/24 bpp (i.e., RGBE and LogLuv) from 96 bpp of an uncompressed HDR pixel. However, this memory reduction is not enough and not practical for easily distributing HDR content or storing large databases of images or video. For example, a minute of a high definition movie (1920× 1080) at 24 fps encoded using 24 bpp LogLuv requires more than 8.3 GB of space, which is nearly double the space of a single layer DVD. Researchers have been working on more sophisticated compression schemes in the last few years to make storing of HDR content more practical. The main strategy has been to modify and/or adapt current compression standards and techniques such as JPEG, MPEG, and block truncation coding (BTC) to HDR content. This chapter presents a review of the state-of-the-art of these compression schemes for HDR images, textures, and videos.
