Store Results in HDF5

Save processed images and their intermediate states to HDF5 for efficient storage and retrieval.

Save a Processed Image

import phenotypic as pht

image = pht.Image.imread("plate.png")
# ... process image ...
image.save2hdf5("processed_plate.h5")

Save Pipeline Intermediates

Capture the image state after each pipeline operation:

pipeline = pht.ImagePipeline(ops=[...])
result = pipeline.apply_with_intermediates(
    image,
    output_dir="intermediates/"
)

Each intermediate is saved as a separate HDF5 file in output_dir/, named after the operation that produced it.

Why HDF5

• Compact: Compressed binary format, much smaller than TIFF/PNG for processed images

• Fast I/O: Random access to any data component without loading the entire file

• Self-describing: Metadata, image arrays, masks, and object maps stored together in one file