Abstract: A liver solid texture synthesis and mapping method based on Computer Unified Device Architecture acceleration (CUDA) was proposed to solve the problem of the overlong time consuming within the period of synthesizing liver solid texture in traditional way. The relevance in traditional serial texture synthesis was elim inated in the new method. The work of selecting and distributing blocks of space synthesis of the liver solid texture was processed by using the parallel processing of multiple threads based on CUDA. Both the procedures of tinting the surface grid nodes of liver model and internal point set traversal in mapping calculation were parallelized and these work were done by GPU. The experimental results show that the new method possesses a higher resultant velocity, simultaneously guaranteeing the realistic of liver solid texture.

Key words: solid texture, synthesis, mapping, CUDA, parallel computing