Study Of Accelerated Algorithm Based On Ray Casting Volume Rendering

The ray casting is a classical direct volume rendering algorithm. With rendering in high quality, it has been gotten a lot of practical applications. The improvements of the algorithm have never been interrupted. Due to the huge volume data and the ray casting directly on the volume data operation, the computational complexity is very high. For its improvement is mainly by reducing the number of sampling points inner circulation, reducing the complexity and improving the way of interpolation algorithm to enhance rendering speed. Along with the development of computer hardware, especially the appear of parallel, programmable GPU and large Graphics memory card, We can make full use of GPU fast floating-point calculations ability and large-scale parallel computing characteristic, making volume rendering speed basically reaches real-time rendering.This paper uses the software method and combines with the MITK algorithm and 3DMeD system to realize the improvement of the ray casting, firstly. Core ideology is through the invalid light judgment reasonably reduced the number of sampling light, using Proximity Clouds algorithm skip empty volume element, and using three linear interpolation method who based on look-up table, the computational time is reduced and the rendering speed is improved; Then a GPU-Parallel computing-based ray casting algorithm is presented. This algorithm make full use of GPU fast floating-point calculations ability, large-scale parallel computing characteristic, and support of GPU to Clips processor inner circulation and Parallel computing function. This algorithm moves the CPU re-sampling process to GPU in parallel processing. Finally, traditional accelerated techniques of early ray terminal, empty space skipping and invalid projection light judgment are combined into the algorithm on the GPU platform. The result is shown that the rendering speed is enhanced effectively.