void convolve(const ImageStackT& _image, const ImageStackT& _kernel,
ImageStackT& _result, const std::vector<DimT>& _offset) {
if (!_image.num_elements()) return;
std::vector<DimT> half_kernel(3);
for (unsigned i = 0; i < 3; ++i) half_kernel[i] = _kernel.shape()[i] / 2;
float image_value = 0;
float kernel_value = 0;
float value = 0;
for (int image_z = _offset[0]; image_z < int(_image.shape()[0] - _offset[0]);
++image_z) {
for (int image_y = _offset[1];
image_y < int(_image.shape()[1] - _offset[1]); ++image_y) {
for (int image_x = _offset[2];
image_x < int(_image.shape()[2] - _offset[2]); ++image_x) {
_result[image_z][image_y][image_x] = 0.f;
image_value = 0;
kernel_value = 0;
value = 0;
// TODO: adapt loops to storage order for the sake of speed (if so,
// prefetch line to use SIMD)
for (int kernel_z = 0; kernel_z < int(_kernel.shape()[0]); ++kernel_z) {
for (int kernel_y = 0; kernel_y < int(_kernel.shape()[1]);
++kernel_y) {
for (int kernel_x = 0; kernel_x < int(_kernel.shape()[2]);
++kernel_x) {
kernel_value = _kernel[_kernel.shape()[0] - 1 - kernel_z]
[_kernel.shape()[1] - 1 - kernel_y]
[_kernel.shape()[2] - 1 - kernel_x];
image_value = _image[image_z - half_kernel[0] + kernel_z]
[image_y - half_kernel[1] + kernel_y]
[image_x - half_kernel[2] + kernel_x];
value += kernel_value * image_value;
}
}
}
_result[image_z][image_y][image_x] = value;
}
}
}
}
void convolve(CImageStackT& _image,
CImageStackT& _kernel,
ImageStackT& _result,
const std::vector<DimT>& _offset){
if(!_image.num_elements())
return;
std::vector<DimT> half_kernel(3);
for(unsigned i = 0;i<3;++i)
half_kernel[i] = _kernel.shape()[i]/2;
float image_value = 0;
float kernel_value = 0;
float value = 0;
for(int image_z = _offset[2];image_z<int(_image.shape()[2]-_offset[2]);++image_z){
for(int image_y = _offset[1];image_y<int(_image.shape()[1]-_offset[1]);++image_y){
for(int image_x = _offset[0];image_x<int(_image.shape()[0]-_offset[0]);++image_x){
_result[image_x][image_y][image_z] = 0.f;
image_value = 0;
kernel_value = 0;
value = 0;
for(int kernel_z = 0;kernel_z<int(_kernel.shape()[2]);++kernel_z){
for(int kernel_y = 0;kernel_y<int(_kernel.shape()[1]);++kernel_y){
for(int kernel_x = 0;kernel_x<int(_kernel.shape()[0]);++kernel_x){
kernel_value = _kernel[_kernel.shape()[0]-1-kernel_x][_kernel.shape()[1]-1-kernel_y][_kernel.shape()[2]-1-kernel_z] ;
image_value = _image[image_x-half_kernel[0]+kernel_x][image_y-half_kernel[1]+kernel_y][image_z-half_kernel[2]+kernel_z] ;
value += kernel_value*image_value;
}
}
}
_result[image_x][image_y][image_z] = value;
}
}
}
}