Interactive Editing of GigaSample Terrain Fields

Abstract

Figure 1: A terrain field of over 300 gigasamples (left). Direct editing using a paint and displacement brush (right) and simultaneous rendering of the resulting changes is performed at 60 fps on a 1920×1080 viewport using our approach. AbstractPrevious terrain rendering approaches have addressed the aspect of data compression and fast decoding for rendering, but applications where the terrain is repeatedly modified and needs to be buffered on disk have not been considered so far. Such applications require both decoding and encoding to be faster than disk transfer. We present a novel approach for editing gigasample terrain fields at interactive rates and high quality. To achieve high decoding and encoding throughput, we employ a compression scheme for height and pixel maps based on a sparse wavelet representation. On recent GPUs it can encode and decode up to 270 and 730 MPix/s of color data, respectively, at compression rates and quality superior to JPEG, and it achieves more than twice these rates for lossless height field compression. The construction and rendering of a height field triangulation is avoided by using GPU ray-casting directly on the regular grid underlying the compression scheme. We show the efficiency of our method for interactive editing and continuous level-of-detail rendering of terrain fields comprised of several hundreds of gigasamples.

Visualization of big SPH simulations via compressed octree grids

“…The recent survey by Balsa Rodriguez et al [19] provides a focused treatment of compression techniques used in the context of volume visualization. We use the wavelet-based method described by Treib et al [20], [21] for the compression of 3D scalar fields, which decodes the data directly on the GPU from a Huffman-encoded wavelet coefficient stream.…”

“…We employ the GPU for both compression and decompression, as it has been shown to achieve superior throughput compared to CPUs [20]. Our GPU compression scheme for volumetric data is similar in spirit to the one proposed by Treib et al [21]: First, the input floating point values are quantized with a userdefined quantization step.…”

Turbulence Visualization at the Terascale on Desktop PCs

“…We use the GPU compression scheme recently proposed by Treib et al [39], which has been tailored explicitly to support both encoding and decoding on current GPU architectures. The wavelet compression builds upon the fast CUDA implementation of the discrete wavelet transform (DWT) by van der Laan et al [40].…”

“…Compression schemes based on transform coding also have a long tradition in visualization, for instance to reduce memory and bandwidth limitations in volume visualization [16,33,43,45]. However, only with the possibility to perform the entire compression pipeline on the GPU [39]-including encoding and decoding-can the full potential of wavelet-based compression be employed for large data visualization. …”

“…This time is vastly dominated by the GPU decompression, and in particular the GPU compression of the intermediate and final results. Compression is about 3 times as expensive as decompression because the run-length and Huffman encoders are more complex than the respective decoders [39]. In particular, Huffman encoding requires a round-trip to the CPU, where the Huffman table is constructed.…”