Naslov
4D Data Compression Methods for Modelling Virtual Medical Reality

Autori
Žagar, Martin ; Knezović, Josip ; Mlinarić, Hrvoje

Izvornik
Annual ... of the Croatian Academy of Engineering (1332-3482) 2009 (2009); 177-191

Vrsta, podvrsta i kategorija rada
Radovi u časopisima, članak, znanstveni

Ključne riječi
reality; active model; region of interest; 4D data compression

Sažetak
Virtual reality is based on sequences of volumetric images whose motion is captured in time. Temporal 3D visualizations of various anatomical parts have been used for education of medical students. Various organ models have been developed using this data. These data sets are typically very large in size and demand a great amount of resources for storage and transmission. Therefore it is necessary to compress such data both fast and efficiently. Also, medical datasets usually contain a region representing the part of the body under investigation, and noisy background with no diagnostic interest. Therefore, it is important to identify contained sub-volumes as different regions of interest. In this paper we propose combination of lossless and lossy compression models to obtain toset demands. Coding of 4D data models should be both fast and efficient. In this paper is shown that with defining the regions of interest there can be achieved higher compression rates. Therefore we proposed lossless coding of areas of high interest and lossy coding areas of small interest.

Izvorni jezik
Engleski

Znanstvena područja
Računarstvo

POVEZANOST RADA