Argolid
uses Tensorstore
for reading and writing pixel data. So Tensorstore
build requirements are needed to be satisfied.
For Linux, these are the requirements:
-
GCC
10 or later -
Clang
8 or later -
Python
3.8 or later -
CMake
3.24 or later -
Perl
, for building libaom from source (default). Must be inPATH
. Not required if-DTENSORSTORE_USE_SYSTEM_LIBAOM=ON
is specified. -
NASM
, for building libjpeg-turbo, libaom, and dav1d from source (default). Must be inPATH
.Not required if-DTENSORSTORE_USE_SYSTEM_{JPEG,LIBAOM,DAV1D}=ON
is specified. -
GNU Patch
or equivalent. Must be inPATH
.
Here is an example of building and installing Argolid
in a Python virtual environment.
python -m virtualenv venv
source venv/bin/activate
pip install cmake
git clone --recurse-submodules https://github.com/sameeul/argolid.git
cd argolid
python setup.py install
Argolid can generate Pyramids from a single image or an image collection with a stitching vector provided. It can generate three different kind of pyramids:
- Neuroglancer compatible Zarr (NG_Zarr)
- Precomputed Neuroglancer (PCNG)
- Viv compatible Zarr (Viv)
Currently, three downsampling methods (mean
, mode_max
and mode_min
) are supported. A dictionary with channel id (integer) as key and downsampling method as value can be passed to specify downsampling method for specific channel. If a channel does not exist as a key in the
dictionary, mean
will be used as the default downsampling method
Here is an example of generating a pyramid from a single image.
from argolid import PyramidGenerartor
input_file = "/home/samee/axle/data/test_image.ome.tif"
output_dir = "/home/samee/axle/data/test_image_ome_zarr"
min_dim = 1024
pyr_gen = PyramidGenerartor()
pyr_gen.generate_from_single_image(input_file, output_dir, min_dim, "NG_Zarr", {0:"mode_max"})
Here is an example of generating a pyramid from a collection of images and a stitching vector.
from argolid import PyramidGenerartor
input_dir = "/home/samee/axle/data/intensity1"
file_pattern = "x{x:d}_y{y:d}_c{c:d}.ome.tiff"
output_dir = "/home/samee/axle/data/test_assembly_out"
image_name = "test_image"
min_dim = 1024
pyr_gen = PyramidGenerartor()
pyr_gen.generate_from_image_collection(input_dir, file_pattern, image_name,
output_dir, min_dim, "Viv", {1:"mean"})