reading pixels as floating point values

Hello,

Is it possible to read in an image where the pixels have normalized floating point values?

If yes what is the difference between reading image using imread function and reading it in a normalized floating point pixel value?

Please suggest me the MATLAB image processing functions that read in the pixel values as normalized floating point.

Thank you.

[David Barry]

Why do you want to do this would be my first question?

Answer 1

[Walter Roberson]

The MATLAB Tiff class can handle reading and writing 32 bit floating point (single precision). See https://www.mathworks.com/matlabcentral/newsreader/view_thread/314008 for some example code on writing it. If I recall correctly, imread() should be able to read those kinds of images without having to use Tiff class, but you do need to use the Tiff class to write them from MATLAB.

The Tiff format does not offer a 64 bit double as a native image format.

[Walter Roberson]

The thread can be found at https://groups.google.com/g/comp.soft-sys.matlab/c/RWWqqLkTcHU/m/RfNsWeTGAwAJ

Answer 2

[Image Analyst]

If you have some custom/proprietary image format (not like the Tiff Walter described) you'll either have to get an already-made reader m-file for it (perhaps from the File Exchange or your instrument or camera anufacturer) or write it yourself using fread().