#!/usr/bin/env python3
# -*- coding: utf-8 -*-
"""
Convenience functions for handling IO with hdf5 files of frequently used data
structures.
@author: Nicola VIGANÒ, Computational Imaging group, CWI, The Netherlands,
and ESRF - The European Synchrotron, Grenoble, France
Created on Wed Mar 1 16:09:53 2017
"""
import numpy as np
try:
import imageio as iio
except ImportError as ex:
print("WARNING: error importing Imageio, using matplotlib instead")
print("Error message:\n", ex)
import matplotlib.image as iio
from . import lightfield
import h5py
[docs]def save_refocused_image(img2d, filename, ind=None):
if ind is not None:
iio.imsave(filename, img2d[ind, ...], vmin=0.0, vmax=1.0)
else:
iio.imsave(filename, img2d, vmin=0.0, vmax=1.0)
[docs]def save_field_toh5(filename, dset_name, data, verbose=False, append=False, compression_lvl=7, to_uint8=False):
if append:
mode = "a"
else:
mode = "w"
if to_uint8:
data = ((data - np.min(data)) / (np.max(data) - np.min(data)) * 255).astype(np.uint8)
with h5py.File(filename, mode) as f:
if compression_lvl is not None:
f.create_dataset(dset_name, data=data, compression="gzip", compression_opts=compression_lvl)
else:
f.create_dataset(dset_name, data=data)
if verbose:
print('Saved dataset "%s" to file: %s' % (dset_name, filename))
[docs]def load_field_fromh5(filename, dset, verbose=False):
mode = "r"
with h5py.File(filename, mode) as f:
data = f[dset][()]
if verbose:
print('Loaded dataset "%s" from file: %s' % (dset, filename))
return data
[docs]def save_refocused_stack(refocus_stack, filename, verbose=False, zs=None):
save_field_toh5(filename, "refocus_stack", refocus_stack, verbose=verbose)
if zs is not None:
save_field_toh5(filename, "zs", zs, append=True)
[docs]def save_lightfield(filename, lf: lightfield.Lightfield):
with h5py.File(filename, "w") as f:
c = lf.camera.__dict__
for k, v in c.items():
f["camera/%s" % k] = v
f["mode"] = lf.mode
f.create_dataset("data", data=lf.data, compression="gzip")
if lf.flat is not None:
f.create_dataset("flat", data=lf.flat, compression="gzip")
if lf.mask is not None:
f.create_dataset("mask", data=lf.mask, compression="gzip")
if (
lf.shifts_vu is not None
and isinstance(lf.shifts_vu, (tuple, list))
and lf.shifts_vu[0] is not None
and lf.shifts_vu[1] is not None
):
f.create_dataset("shifts_vu", data=lf.shifts_vu, compression="gzip")
[docs]def load_lightfield(filename):
with h5py.File(filename, "r") as f:
camera = lightfield.Camera()
c = camera.__dict__
for k in c:
setattr(camera, k, f["camera/%s" % k][()])
lf = lightfield.Lightfield(camera_type=camera)
lf.mode = f["mode"][()]
lf.data = f["data"][()]
if "/flat" in f:
lf.flat = f["flat"][()]
if "/mask" in f:
lf.mask = f["mask"][()]
if "/shifts_vu" in f:
lf.shifts_vu = f["shifts_vu"][()]
return lf