论文部分内容阅读
We implemented accurate FFD in terms of triangular Bézier surfaces as matrix multiplications in CUDA and rendered them via Open GL. Experimental results show that the proposed algorithm is more efficient than the previous GPU acceleration algorithm a