Reading and re sizing a raw 32 bit float image in c with opencv

edit

I have no idea why anyone would like to do this in the old and depricated C-API, nor have I the idea if this extra constant is even available in C-API. If you use the now actively developed C++ API this should do it

Mat image = imread("/data/test.tif", CV_LOAD_IMAGE_ANYDEPTH);

It reads images with 16/32 bit depth and it is available in OpenCV 2.4.9.