Accelerating volume rendering by frame indexing

Volume visualization is the technique of displaying two dimensional projections of three dimensional (volume) data. Data may be acquired from a medical scanner, such as an MRI, CT, SPECT, or US scanner. Visualizing a given three dimensional dataset can be done by surface rendering algorithms or by volume rendering algorithms. Surface rendering algorithms require an intermediate geometric representation and are therefore less attractive. Several volume rendering techniques are known. The most important distinction between the techniques is the order in which the voxels, smallest volume element, are processed to create an image, image-order or object-order algorithms. The most often used technique is ray tracing, an image-order algorithm, which will also be used in this thesis.The major problem of volume rendering methods is the time required to generate images. To improve the processing time, special care should be taken. Achieving interactive rendering rates requires hardware implementation, parallel rendering or other assistance models. This paper describes a volume rendering method for visualization of scientific data with commercially available personal computers, without requiring dedicated systems.