True 4D Image Denoising on the GPU

Algorithm 1

//  Code  that  is  executed  before  the  kernel  is  launched

int threadsInX  =  32;

int threadsInY  =  16;

int blocksInX  =  DATA_W/threadsInX;

int blocksInX  =  DATA_H/threadsInY;

dimGrid  =  dim3(blocksInX,  blocksInY);

dimBlock  =  dim3(threadsInX,  threadsInY,  1);

//  Code  that  is  executed  inside  the  kernel

int x  =  blockIdx.x  *  blockDim.x  +  threadIdx.x;

int y  =  blockIdx.y  *  blockDim.y  +  threadIx.yd;