Study Of Key Issues On The Hardware Architecture RCAA Of Ray-casting Arithmetic

Direct Volume Rendering is a branch and heat of Computer Graphics and it is also a core technique of Computer Visualization,which is applied abroad in the field of medicine,geology science,hydromechanics etc.Recently,the Ray-casting arithmetic is the most popular volume rendering technique in the international boundary.Following the development of the visualization techniques, the scale of volume data is larger and larger.It leads the conflict between the speed of rendering and the need of users(real time and interactive display) deeply.Ensuring the quality of rending,speed becomes the bottleneck of the development of volume rendering.Based on this problem, this paper proposes the hardware supported architecture RCAA(Ray-casting Arithmetic Accelerator) on Ray-casting arithmetic.Ray-casting is a typical pipeline structural arithmetic.Because of this point this paper presents the computation model (High speed rendering model and High quantity rendering model) and the hardware supported architecture RCAA. This paper also describes the function of modules of RCAABecause the memory system is the bottleneck of RCAA,this paper presents that the conflict-free low-bit crossover parallel memory mode can improve the troughput of memory system.Based the point of parallel memory,this paper propose how to deposit the voxels in the memory and how to use the Page Mode of DRAM sufficiently to improve the of Cache.According to the distribution of voxels in the parallel memory,this paper presents a method to calculate the address in order to confirm the location of voxels.This method can convert the three-dimensional address to the one-dimensional.The paper also proposes how to use the Look-Up Table struction to complete the coordinate convertion in order to confirm the location of resamples.This paper improves on the linear interpolation and designs a tri-linear interpolator for the restruction of volume rendering.It makes use of emulational tool and synthetical tool along with software to ensure the validity of logic.Finally, this paper complete the design of Shading Unit and Compositing Unit.And it makes use of emulational tool and synthetical tool along with software to ensure the validity of logic.Furthermore based on the point of shading,it proposes a new assumption in order to solve the speed of computation and the area of chip.