Søren Hauberg skrev: > tir, 23 02 2010 kl. 21:43 +0100, skrev David Grundberg: > >> I don't know if the underlaying tiff library supports reading single frames. >> >> One way to work-around any frame loading problem is to make sure that >> all frames are loaded at once in one imread call. Maybe imread should >> read all frames available if there is only one argument. Maybe imread >> should be made to accept ":" as a frame index and load all available >> frames only then. >> > > If I understand things correctly, you can currently do > > I = imread ("myfile.tif", 7); > > after which 'I' will contain the 7th frame of "myfile.tif". Hence, the > class of 'I' is an array. > > The obvious extension is to allow for > > J = imread ("myfile.tif", 3:5); >
This is already supported. The result J is a four-dimensional matrix. I think it makes perfect sense. > which would then return frames 3, 4 and 5 in a cell array. The problem > is then that the class of the output depends on how many frames the user > asks for. If the user asks for one frame, she would get an array (for > Matlab compatibility), but if she asks for more than one frame the > result is a cell array. This will most likely be the cause of great > confusion and many bugs. > > One solution would be to do something like > > [J1, J2, J3] = imread ("myfile.tif", 3:5); > > but then you have to know how many frames you are asking for in advance, > which really isn't that great. > > Any thoughts on how to deal with this? > > Søren > I tested Matlab and it doesn't support several frame indices, so I don't think we need to make any changes (like put frames into cell arrays) David ------------------------------------------------------------------------------ Download Intel® Parallel Studio Eval Try the new software tools for yourself. Speed compiling, find bugs proactively, and fine-tune applications for parallel performance. See why Intel Parallel Studio got high marks during beta. http://p.sf.net/sfu/intel-sw-dev _______________________________________________ Octave-dev mailing list Octave-dev@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/octave-dev