Repair TIFF image metadata that have MPP in XResolution and YResolution fields.

repair_tiff_resolution.py

"""Repair the TIFF metadata for images where XResolution and YResolution contain MPP.

This script updates the following tags:
- ResolutionUnit --> to centimeter
- XResolution --> from mppx to pixels per centimeter in x dimension
- YResolution --> from mppy to pixels per centimeter in y dimension
"""

import argparse
from pathlib import Path
from typing import Tuple

import tifftools


def get_mpp(path: str) -> Tuple[float, float]:
    """Return tuple of (mpp_x, mpp_y).

    MPP = micrometers per pixel.
    """
    info = tifftools.read_tiff(path)
    # [1490525311, 4294967295]  --> 0.34703997
    xs = info["ifds"][0]["tags"][int(tifftools.Tag.XRESOLUTION)]["data"]
    ys = info["ifds"][0]["tags"][int(tifftools.Tag.YRESOLUTION)]["data"]
    assert len(xs) == 2
    assert len(ys) == 2
    mpp_x = xs[0] / xs[1]
    mpp_y = ys[0] / ys[1]
    return mpp_x, mpp_y


def mpp_to_resolution_cm(mpp: float):
    # (pixel / microns) * (microns / centimeter)
    # 10_000 microns per centimeter.
    return 1 / mpp * 10_000


def repair(path: str, output_dir: str):
    """Repair the resolution metadata and write to 'output_dir' with the same name as
    the original file.
    """
    output = Path(output_dir) / Path(path).name
    mppx, mppy = get_mpp(path)
    print(f"{Path(path).name}\tmppx={mppx}\tmppy={mppy}")
    if mppx > 1 or mppy > 1:
        raise ValueError(
            f"MPP_X and/or MPP_Y are greater than 1: {mppx} and {mppy}. They probably"
            " do not represent microns per pixel."
        )
    xresolution = mpp_to_resolution_cm(mppx)
    yresolution = mpp_to_resolution_cm(mppy)
    tifftools.tiff_set(
        path,
        output,
        overwrite=False,
        setlist=[
            (
                tifftools.Tag.RESOLUTIONUNIT,
                tifftools.constants.ResolutionUnit.CENTIMETER.value,
            ),
            (tifftools.Tag.XRESOLUTION, xresolution),
            (tifftools.Tag.YRESOLUTION, yresolution),
        ],
    )


if __name__ == "__main__":
    p = argparse.ArgumentParser()
    p.add_argument("--output-dir", required=True, help="Path to output directory.")
    p.add_argument("tiff", help="Path to input TIFF file.")
    args = p.parse_args()
    repair(path=args.tiff, output_dir=args.output_dir)