Source code for cloudreg.scripts.download_data

from .util import imgResample
from cloudvolume import CloudVolume
from argparse import ArgumentParser
import numpy as np
import SimpleITK as sitk
import tifffile as tf


[docs]def get_mip_at_res(vol, resolution): """Find the mip that is at least a given resolution Args: vol (cloudvolume.CloudVoluem): CloudVolume object for desired precomputed volume resolution (int): Desired resolution in nanometers Returns: tuple: mip and resolution at that mip """ tmp_mip = 0 tmp_res = 0 for i, scale in enumerate(vol.scales): if (scale["resolution"] <= resolution).all(): tmp_mip = i tmp_res = scale["resolution"] elif i == 0: tmp_res = scale["resolution"] return tmp_mip, tmp_res return tmp_mip, tmp_res
[docs]def download_data(s3_path, outfile, desired_resolution, resample_isotropic=False, return_size=False): """Download whole precomputed volume from S3 at desired resolution and optionally resample data to be isotropic Args: s3_path (str): S3 path to precomputed volume outfile (str): Path to output file desired_resolution (int): Lowest resolution (in nanometers) at which to download data if desired_resolution isnt available. resample_isotropic (bool, optional): If true, resample data to be isotropic at desired_resolution. Returns: resolution: Resoluton of downloaded data in microns """ vol = CloudVolume(s3_path) mip_needed, resolution = get_mip_at_res(vol, np.array([desired_resolution] * 3)) vol = CloudVolume(s3_path, mip=mip_needed, parallel=True) # download img and convert to C order img = np.squeeze(vol[:, :, :]).T # save out as correct file type img_s = sitk.GetImageFromArray(img) # set spacing in microns resolution = np.divide(resolution, 1000.0).tolist() img_s.SetSpacing(resolution) if resample_isotropic: img_s = imgResample(img_s, np.divide([desired_resolution]*3,1000.)) resolution = np.divide([desired_resolution]*3,1000.) # if output is tiff, use tiffile if 'tif' in outfile.split('.')[-1]: tf.imwrite(outfile, sitk.GetArrayFromImage(img_s)) # else use SimpleITK else: sitk.WriteImage(img_s, outfile) # tf.imsave(outfile, img.T, compress=3) # return resolution in um if return_size: return (resolution, vol.scales[mip_needed]['size']) return resolution
if __name__ == "__main__": parser = ArgumentParser( description="Download volume from S3 for subsequent registration." ) parser.add_argument( "s3_path", help="S3 path to precomputed volume layer in the form s3://<bucket-name>/<path-to-precomputed-volume>", ) parser.add_argument("outfile", help="name of output file with associated file extension. eg. /path/to/image.tif") parser.add_argument( "desired_resolution", help="Desired minimum resolution for downloaded image in nanometers. Resolution assumed to be same in all 3 dimensions.", type=int ) parser.add_argument("--isotropic", help="Resample data to isotropic at desired_resolution.", action='store_true') args = parser.parse_args() download_data( args.s3_path, args.outfile, args.desired_resolution, resample_isotropic=args.isotropic )