3D Data Denoising via Nonlocal Means Filter by Using Parallel GPU Strategies

Algorithm 4

CUDA code of NLM algorithm with full unrolling strategy.

(1)  int  const  i_1  =  threadIdx.x  +  blockDim.x*blockIdx.x;

(2)  int  const  i_2  =  threadIdx.y  +  blockDim.y*blockIdx.y;

(3)  int  const  i_3  =  threadIdx.z  +  blockDim.z*blockIdx.z;

(4)     /*  localstatements  */

(5)         if  ((i_1  ≥  0)  &&  (i_1  <  X_Dim)  &&  (i_2  ≥  0)  &&  (i_2  <  Y_Dim)  &&

   (i_3  ≥  0)  &&  (i_3  <  Z_Dim))  {

(6)         /*  compute  out_img [i_1  +  i_2*X_Dim +  i_3*X_Dim*Y_Dim] using

   in_img [i_1  +  i_2*X_Dim  +  i_3*X_Dim*Y_Dim]   */

(7)     }