def download_sentinel_data(item, bands):
# get paths w.r.t. id
paths = file_paths_wrt_id(item._data["id"])
# get meta info on path, to be used by boto3
info_response = requests.get(item.assets["info"]["href"])
info_response_json = json.loads(info_response.text)
# save bands generically
for band in bands:
# pass band id in metadata
info_response_json["band_id"] = band
band_filename = paths["b%s" % band]
if not data_file_exists(band_filename):
save_to_file(
item.assets["B0{}".format(band)]["href"],
band_filename,
item._data["id"],
"✗ required data doesn't exist, downloading %s %s"
% (band_tag_map["b" + str(band)], "band"),
meta=info_response_json,
)
else:
rprint(
"[green] ✓ ",
"required data exists for {} band".format(
band_tag_map["b" + str(band)]
),
)
return item._data["id"]
def process_landsat_vegetation(id, bands):
# get paths of files related to this id
paths = file_paths_wrt_id(id)
# stack NIR, R, G bands
# open files from the paths, and save it as stack
b5 = rio.open(paths["b5"])
b4 = rio.open(paths["b4"])
b3 = rio.open(paths["b3"])
# read as numpy ndarrays
nir = b5.read(1)
r = b4.read(1)
g = b3.read(1)
with rio.open(
paths["stack"],
"w",
driver="Gtiff",
width=b4.width,
height=b4.height,
count=3,
crs=b4.crs,
transform=b4.transform,
dtype=b4.dtypes[0],
photometric="RGB",
) as rgb:
rgb.write(nir, 1)
rgb.write(r, 2)
rgb.write(g, 3)
rgb.close()
source_path_for_rio_color = paths["stack"]
rprint("Let's make our 🌍 imagery a bit more colorful for a human eye!")
# apply rio-color correction
ops_string = "sigmoidal rgb 20 0.2"
# refer to felicette.utils.color.py to see the parameters of this function
# Bug: number of jobs if greater than 1, fails the job
color(
1,
"uint16",
source_path_for_rio_color,
paths["vegetation_path"],
ops_string.split(","),
{"photometric": "RGB"},
)
# resize and save as jpeg image
print("Generated 🌍 images!🎉")
rprint("[yellow]Please wait while I resize and crop the image :) [/yellow]")
process_sat_image(paths["vegetation_path"], paths["vegetation_path_jpeg"])
rprint("[blue]GeoTIFF saved at:[/blue]")
print(paths["vegetation_path"])
rprint("[blue]JPEG image saved at:[/blue]")
print(paths["vegetation_path_jpeg"])
# display generated image
display_file(paths["vegetation_path_jpeg"])
def main(coordinates, location_name, pan_enhancement, no_preview, vegetation, version, product):
"""Satellite imagery for dummies."""
if version:
version_no = pkg_resources.require("felicette")[0].version
exit_cli(print, f"felicette {version_no}.")
if not coordinates and not location_name:
exit_cli(print, "Please specify either --coordinates or --location-name")
if location_name:
coordinates = geocoder_util(location_name)
# unless specified, cloud_cover_lt is 10
item = search_satellite_data(coordinates, 10, product=product)
# check if directory exists to save the data for this product id
check_sat_path(item._data["id"])
# if preview option is set, download and preview image
if not no_preview:
preview_satellite_image(item)
# set bands to process
bands = [2, 3, 4]
if pan_enhancement and (product != "sentinel"):
bands.append(8)
if vegetation:
bands = [3, 4, 5]
# NB: can't enable pan-enhancement with vegetation
# NB: can't enable pan-enhancement with sentinel
try:
trigger_download_and_processing(item, bands)
except RasterioIOError:
response = input(
"Local data for this location is corrupted, felicette will remove existing data to proceed, are you sure? [Y/n]"
)
if response in ["y", "Y", ""]:
# remove file dir
file_paths = file_paths_wrt_id(item._data["id"])
remove_dir(file_paths["base"])
# retry downloading and processing image with a clean directory
trigger_download_and_processing(item, bands)
elif response in ["n", "N"]:
exit_cli(print, "")
def preview_satellite_image(item):
paths = file_paths_wrt_id(item._data["id"])
# download image and save it in directory
if not data_file_exists(paths["preview"]):
save_to_file(
item.assets["thumbnail"]["href"],
paths["preview"],
item._data["id"],
"✗ preview data doesn't exist, downloading image",
)
else:
rprint("[green] ✓ ", "required data exists for preview image")
# print success info
rprint("[blue]Preview image saved at:[/blue]")
print(paths["preview"])
# prompt a confirm option
response = input(
"Are you sure you want to see an enhanced version of the image at the path shown above? [Y/n]"
)
return handle_prompt_response(response)
def download_landsat_data(landsat_item, bands):
# get paths w.r.t. id
paths = file_paths_wrt_id(landsat_item._data["id"])
# save bands generically
for band in bands:
band_filename = paths["b%s" % band]
if not data_file_exists(band_filename):
save_to_file(
landsat_item.assets["B{}".format(band)]["href"],
band_filename,
landsat_item._data["id"],
"✗ required data doesn't exist, downloading %s %s" %
(band_tag_map["b" + str(band)], "band"),
)
else:
rprint(
"[green] ✓ ",
"required data exists for {} band".format(
band_tag_map["b" + str(band)]),
)
return landsat_item._data["id"]
def process_landsat_rgb(id, bands):
# get paths of files related to this id
paths = file_paths_wrt_id(id)
# stack R,G,B bands
# open files from the paths, and save it as stack
b4 = rio.open(paths["b4"])
b3 = rio.open(paths["b3"])
b2 = rio.open(paths["b2"])
# read as numpy ndarrays
r = b4.read(1)
g = b3.read(1)
b = b2.read(1)
with rio.open(
paths["stack"],
"w",
driver="Gtiff",
width=b4.width,
height=b4.height,
count=3,
crs=b4.crs,
transform=b4.transform,
dtype=b4.dtypes[0],
photometric="RGB",
) as rgb:
rgb.write(r, 1)
rgb.write(g, 2)
rgb.write(b, 3)
rgb.close()
source_path_for_rio_color = paths["stack"]
# check if band 8, i.e panchromatic band has to be processed
if 8 in bands:
# pansharpen the image
rprint(
"Pansharpening image, get ready for some serious resolution enhancement! ✨"
)
gdal_pansharpen(["", paths["b8"], paths["stack"], paths["pan_sharpened"]])
# set color operation's path to the pansharpened-image's path
source_path_for_rio_color = paths["pan_sharpened"]
rprint("Let's make our 🌍 imagery a bit more colorful for a human eye!")
# apply rio-color correction
ops_string = "sigmoidal rgb 20 0.2"
# refer to felicette.utils.color.py to see the parameters of this function
# Bug: number of jobs if greater than 1, fails the job
color(
1,
"uint16",
source_path_for_rio_color,
paths["output_path"],
ops_string.split(","),
{"photometric": "RGB"},
)
# resize and save as jpeg image
print("Generated 🌍 images!🎉")
rprint("[yellow]Please wait while I resize and crop the image :) [/yellow]")
process_sat_image(paths["output_path"], paths["output_path_jpeg"])
rprint("[blue]GeoTIFF saved at:[/blue]")
print(paths["output_path"])
rprint("[blue]JPEG image saved at:[/blue]")
print(paths["output_path_jpeg"])
# display generated image
display_file(paths["output_path_jpeg"])
