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;