Object oriented definition of coordinates, voxels, matrix indexing etc.

[jwboth]

Currently, points in the image can be specified throughtheir coordinate or voxel location. In addition, voxel coordinates (in particular in 2d) are used in different formats (matrix indexing, cartesian indexing, etc.). As external tools as opencv (in particular) uses different formats, making an internally consisten choice for DarSIA is difficult. Thus, we need to stick to the different formats. To remove confusion, the aim however should be to use objects which uniquely identify the format, i.e., we need to define objects class Coordinate (Cartesian coordinate), Voxel (voxel coordinate, which can have different indexing modes and allows for transforming the indexing through helper routines).

[jwboth]

Still various transformation objects do not make use of Voxel, Coordinate etc. A unified use is envisioned.