Question

I am making a three dimensional array for image manipulation – X and Y dimensions and then a third dimension because each pixel has not one value but (bitsPerPixel/8) values.

I am frankly not very well versed in C++ but the API for the camera is strictly C++ and so I want to check up with somebody more skilled before I set out and ask “is this right?”:

vector<vector<vector<int>>> mask(m_nSizeX, vector<vector<int>>(m_nSizeY, vector<int>(m_nBitsPerPixel / 8)));

Answer 1

If you have C++11 support, then yes.

If not, you need some extra spaces in your templates:

vector<vector<vector<int> > >

instead of

vector<vector<vector<int>>>

and so on…

Although I personally wouldn’t use a vector for the last value, but rather a structure.