Add reinhard color normalization support

[andreped]

I have added support for reinhard color normalization, as an alternative to Macenko, as discussed in #24.

Reinhard itself was straight forward to implement, but I had a hard time getting the RGB/LAB color converters to work as intended. After looking at what OpenCV, Scikit-Image, and other projects were doing, I managed to implement numpy-equivalents. In the end, from qualitative assessment, I get the same results as another commonly used library, StainTools.

Have not had the time to benchmark it yet, hence, experimental, but at least it runs and serves as an alternative method.

UPDATE: Now both Numpy and TensorFlow backends are supported!

What's changed:

• Added ReinhardNormalizer - currently supports both numpy and tensorflow backends.
• Added rgb2lab and lab2rgb color conversion implementations.
• Added convenience functions relevant to Reinhard in stats and split.
• Updated README regarding experimental Reinhard support with reference to original article.
• Added RGB->LAB->RGB unit test (currently fails - focused on getting it to work with Reinhard and not in the general case).
• Moved result.pngto data/directory and updated corresponding path in README.
• Minor README refactoring.

[andreped]

Note that I also added TF support now, as I need it for one of my studied.

Will not have time for PyTorch any time soon, but at least both Numpy and TF backends work as intended.

[carloalbertobarbano]

Thanks for the contribution! Merging it now