omdenalore.satellite_imagery package

Submodules

omdenalore.satellite_imagery.find_resolution module

omdenalore.satellite_imagery.find_resolution.resolution(input_tiff: str) → Tuple[int, int]

Calculate the resolution of the given TIF file.

Parameters

input_tiff (str) – Path to TIF file.

Returns

resolution as a (x, y) tuple

omdenalore.satellite_imagery.geom_to_json module

omdenalore.satellite_imagery.geom_to_json.geom_to_json(df: pandas.DataFrame, i: int) → Tuple[ee.geometry.Geometry, str]

Converts shapefile or any geometry input into JSON format for GEE.

Parameters

  • df – panda dataframe containing shapefile contents

  • i – integer value indicating which row in the dataframe to process

Returns

Tuple of (JSON string, label)

omdenalore.satellite_imagery.preprocess_tiff module

class omdenalore.satellite_imagery.preprocess_tiff.SatelliteImage(path: str, parts: int)

Bases: object

SatelliteImage class is used to preprocess satellite images

Parameters

  • path (str) – Path of the input .tif image

  • parts – Splits the single image into smaller images with

a total of (parts * parts) images returned :type parts: int

Example

from omdenalore.satellite_imagery.preprocess_tiff import SatelliteImage >>> si = SatelliteImage( >>> path=”sample.tif”, parts=5 >>> ) >>> si.split()

split() → None

Splits a single .tif image into multiple parts in the same directory

This makes it easier for visualising and storing in memory compared to having a single large image

omdenalore.satellite_imagery.raster_summary module

omdenalore.satellite_imagery.raster_summary.raster_summary(raster_dir: str, raster_file: str) → None

Reads a raster file and outputs the metadata and image.

Parameters

  • raster_dir – dir path location of the raster file

  • raster_file – raster file name

omdenalore.satellite_imagery.reprojecct_crs module

omdenalore.satellite_imagery.reprojecct_crs.reproject_raster(rf: sphinx.ext.autodoc.mock.rioxarray.__qualname__, new_crs: str) → sphinx.ext.autodoc.mock.rioxarray.__qualname__

Reproject the raster from a current CRS to a user specified CRS.

Parameters

  • rf – (r)aster (f)ile sent for reprojection

  • new_crs – user defined new CRS projection

Example

from omdenalore.satellite_imagery.reproject_crs import reproject_raster new_raster_object = reproject_raster(current_raster, ‘EPSG:4326’)

omdenalore.satellite_imagery.utils module

omdenalore.satellite_imagery.utils.add_padding(image: numpy.ndarray, padding_height: int, padding_width: int, color: Tuple[int, int, int] = (0, 0, 0)) → numpy.ndarray

Add padding around the image so that buildings touching the corners can be bought to center. Creates a new image of the desired size and color (black) for padding This is necessary as corner buildings will not be detected by contour detection algorithm if the building edges are outsite of the image

Parameters

  • images – Image that you want padding around

  • padding_height (int) – Amount of padding in y direction

  • padding_width – Amount of padding in x direction

  • padding_width – int

  • color (1x3 tuple) – RGB color of the padding

Returns

Image with padding

omdenalore.satellite_imagery.utils.convert_pixels_to_geocoordinates(source_tiff_path: str, list_of_coords: List[Tuple[int, int]]) → numpy.ndarray

This function convert pixel coordinates into geo coordinates according to the transform information inside the source tiff file

Parameters

source_tiff_path – The source tiff file path from which the

bounding box was predicted. Tiff file should contain geo-transform information :type source_tiff_path: string :param list_of_coords: list of coordinates in the form of [(x1, y1)] :type bbox: List[List[int]]

Returns

Numpy array

omdenalore.satellite_imagery.utils.create_contours(coords: List[Tuple[int, int]]) → numpy.ndarray

Converting the coordinates of the masks to numpy array

Parameters

coords (List of lists) – List of Coordinates in the form the form of [(x1, y1)]

Returns

list of contours in numpy format

omdenalore.satellite_imagery.utils.zoom_to_fill(image: numpy.ndarray, mask: numpy.ndarray, padding_val: int = 0) → numpy.ndarray

Use the mask of the object to center the object in the image and zoom to that object

Parameters

  • image (numpy array) – Input Image

  • mask (numpy array) – Mask of the object that you want to center

  • padding_val – number of pixel padding to add around

the centered object :type padding_val: int

Returns

Numpy image with centered object

Module contents