Browse Source

norm kernel

master
anapt 7 years ago
parent
commit
f185b8ad31
  1. 3
      mean_shift_cuda/meanshift_kernels.cu

3
mean_shift_cuda/meanshift_kernels.cu

@ -88,14 +88,13 @@ __global__ void norm(Matrix mean_shift_vector, double *current_norm) {
int row = blockIdx.x * blockDim.x + threadIdx.x; int row = blockIdx.x * blockDim.x + threadIdx.x;
// performs calculations only if thread's indexes are within matrix bounds // performs calculations only if thread's indexes are within matrix bounds
if (row >= denominator.height){ if (row >= mean_shift_vector.height){
return; return;
} }
// for (int column = 0; column < kernel_matrix.width; ++column){ // for (int column = 0; column < kernel_matrix.width; ++column){
// cell_value += kernel_matrix.elements[row * kernel_matrix.width + column]; // cell_value += kernel_matrix.elements[row * kernel_matrix.width + column];
// } // }
denominator.elements[row] = cell_value;
// performs calculations only if thread's indexes are within matrix bounds // performs calculations only if thread's indexes are within matrix bounds
// if (row * mean_shift_vector.width + col >= mean_shift_vector.width * mean_shift_vector.height){ // if (row * mean_shift_vector.width + col >= mean_shift_vector.width * mean_shift_vector.height){
// return; // return;

Loading…
Cancel
Save