In the current implementation of bfopen.m it is possible, and probably preferable, to input the parameters x, y, w, h to set an arbitrary image tile to be read from the file.
However, given that the file contains several series, e.g. both subsampled versions of the original image but also label, macro and thumbnail images, the parameters are not relevant for the other series contained in the file. For subsampled images the parameters could simply be scaled according to the scale factor and for label, macro and thumbnail images they should not be applied at all. So far so good.
When I have been playing around with a couple of images from Apeiro and Hamamatsu, download from openslide.org, I don't seem to find any metadata that I can use to determine what kind of image that is contained in the current series. So far it seems that the first series is the original image, and then comes a number of subsampled images and finally the label, macro and thumbnail images. But I cannot determine the number of subsampled images.
Any suggestions?
/Daniel