This project started as way to show evidence on how much storage does it take to write the same image in different image formats.
We tested the following formats, and their compression style:
- TIFF, Lossless compression
- JPEG, Lossy compression
- PNG, Lossless compression
- GIF, Lossless compression
- BMP, Lossless compression
I had some techincal issue running this. I found easier to install
pip using:
sudo apt-get install python-pip --yes
The experiments are conducted with Python and the OpenSlide library.
To install OpenSlide in python use:
pip install openslide-python
We extract tiles with different sizes from an image in the TCGA project. We loaded the SVS image using the OpenSlide repository. Later we extract tiles from three different regions: start, middle and end. We extract six different tiles from each region, starting from a 32x32 size and doubling all the way up to tiles with 1024x1024 in size. Every tile generated, is stored in all five image formats, thus generating 90 tiles (5 formats x 6 dimensios x 3 regions).
