 | 1 | initial version |
I'd like to do something like:
Mat foo(N, N, CV_MAKETYPE(CV_8U, 1), data);
typedef Matx<uint8_t, 4, 4> block;
auto iblock = foo.begin<block>();
auto done = foo.end<block>();
while (iblock != done)
{
Mat blk(*iblock);
// process my 4x4 block
}
I think this should be possible, but not sure how.
| | 2 | No.2 Revision |
I'd like to do something like:
Mat foo(N, N, CV_MAKETYPE(CV_8U, 1), data);
typedef Matx<uint8_t, 4, 4> block;
auto iblock = foo.begin<block>();
auto done = foo.end<block>();
while (iblock != done)
{
Mat blk(*iblock);
// process my 4x4 block
}
I think this should be possible, but not sure how.
| | 3 | No.3 Revision |
I'd like to do something like:
Mat foo(N, N, CV_MAKETYPE(CV_8U, 1), data);
typedef Matx<uint8_t, 4, 4> block;
auto iblock = foo.begin<block>();
auto done = foo.end<block>();
while (iblock != done)
{
Mat blk(*iblock);
// process my 4x4 block
++iblock;
}
I think this should be possible, but not sure how.
| | 4 | Tweaked code example for constistency with question title. |
I'd like to do something like:
Mat foo(N, N, foo(IMGSIZE, IMGSIZE, CV_MAKETYPE(CV_8U, 1), data);
typedef Matx<uint8_t, 4, 4> block;
block; // N=4
auto iblock = foo.begin<block>();
auto done = foo.end<block>();
while (iblock != done)
{
Mat blk(*iblock);
// process my 4x4 block
++iblock;
}
I think this should be possible, but not sure how.
| | 5 | retagged |
I'd like to do something like:
Mat foo(IMGSIZE, IMGSIZE, CV_MAKETYPE(CV_8U, 1), data);
typedef Matx<uint8_t, 4, 4> block; // N=4
auto iblock = foo.begin<block>();
auto done = foo.end<block>();
while (iblock != done)
{
Mat blk(*iblock);
// process my 4x4 block
++iblock;
}
I think this should be possible, but not sure how.