load_geo_img_dir#

Module : eotorchloader.dataset.utils

load_geo_img_dir(img_dir: pathlib.Path, img_suffix_list=['.tif', '.jp2']) → pandas.DataFrame#

list info/metadata of images from a directory into a dataframe

list all images which are in a img_dir and with suffix present in img_suffix_list. All image are open with rasterio to load their geographic metadata and all the informations are written in a pandas Dataframe with one row by image and with columns :

“name” : name (str) of the image without extension
“path” : full path of the image (dir+name+extension)
“width” : width of the image in pixel (int)
“height” : height of the image in pixel (int)
“res_x” : resolution in x axis (float)
“res_y” : resolution in y axis (float)
“ul_x” : upper left coordinate of the image in x axis (float)
“ul_y” : upper left coordinate of the image in y axis (float)
“transform” : full affine transform to convert from pixel coordinate to grund coordinate. follow the rasterio convention.

Note

The rasterio transform convention is :

transform = (
    resolution/scale x,
    rotation row/x,
    upper_left/origin x,
    rotation col/y,
    resolution/scale y,
    upper_left/origin y)

Parameters

img_dir (Path) – Path of the directory where images will be search
img_suffix_list (List[str]) – list of image extension to filter files in img_dir.

Returns

Pandas dataframe with a row by images. empty if no images found

Return type

pd.DataFrame

Dataset

Transforms