First commit and upload to repository

Author: Nicolás Guardo

image_processing.py

@@ -0,0 +1,53 @@
+import os
+import glob
+from pathlib import Path
+import numpy as np
+import rasterio
+from matplotlib import pyplot as plt
+
+
+def load_image_bands(image_folder: str, bands: list, pattern_str: str, file_ext: str = "TIF") -> dict:
+    """
+    Function fot loading the individual bands of a raster.
+    :param image_folder:
+    :param bands:
+    :param pattern_str:
+    :param file_ext:
+    :return: dict
+    """
+    band_dict = {}
+    path = Path(image_folder)
+    for band in bands:
+        file = next(path.glob(f"{pattern_str}{band}.{file_ext}"))
+        dataset = rasterio.open(file)
+        band_dict.update({band: dataset.read(1)})
+        dataset.close()
+
+    return band_dict
+
+
+def display_rgb(bands, red_channel, green_channel, blue_channel):
+    rgb = np.stack([bands[red_channel], bands[green_channel], bands[blue_channel]], axis=-1)
+    rgb = rgb / rgb.max()
+
+
+def stack_bands(image_path: str, image_folder: str, file_pattern: str, output_raster: str, metadata: dict, sort=True):
+    os.chdir(image_path)
+
+    files = glob.glob(f"{image_folder}/{file_pattern}")
+
+    if sort:
+        files.sort()
+    else:
+        files.sort(reverse=True)
+
+    with rasterio.open(output_raster, "w", metadata) as output:
+        for index, filepath in enumerate(files, start=1):
+            source = rasterio.open(filepath)
+            output.write(source.read(1), index)
+            source.close()
+        
+
+
+stack_bands("C:\\Users\\orden\\Desktop\\Teledeteccion_QGIS\\ImagenesLandsat",
+            "LT05_L1TP_229092_19860116_20200918_02_T1", "*_B*.TIF")

vector_conversion.py

@@ -0,0 +1,154 @@
+# Imports
+import numpy as np
+import pandas as pd
+from shapely.geometry import Point, MultiPoint, LineString, Polygon
+import geopandas as gpd
+
+
+def extract_coordinates(line: str) -> tuple:
+    """
+    Function used to extract coordinates from a string in the form x,y\n read from a txt file
+    1. Convert the line into a list using the split method
+    2. Select the first element in the list (which should be the x coordinate) and convert it into a float type
+    3. Select the second element in the list (which should be the y coordinate), slice it until the last element (-1 - \n)
+    and convert it into a float number
+
+    The function returns a tuple in the form (x,y)
+    """
+    pair = line.split(",")
+    x = float(pair[0])
+    y = float(pair[1][:-1])
+    coordinates = (x, y)
+    return coordinates
+
+
+def txt_to_vector_layer(txt_filepath: str, dst_filepath: str = None, driver: str = None, geom_type="Point",
+                        epsg=4326) -> gpd.GeoDataFrame:
+    """
+    Function created to convert a txt file to a simple vector layer
+
+    :param txt_filepath: Filepath to the txt file
+    :param dst_filepath: None The vector layer's output filepath in case of saving
+    :param driver: None The vector layer's driver in case of saving
+    :param geom_type: Point The geometry type of the vector layer
+    :param epsg: 4326 The Coordinate Reference System of the layer
+    :return: GeoPandas GeoDataFrame of and output file
+    """
+    # Open the txt file
+    with open(txt_filepath) as txt_file:
+        lines = txt_file.readlines()
+        txt_file.close()
+
+    coords_list = []  # empty list
+
+    for line in lines:
+        coord_pair = extract_coordinates(line)
+        coords_list.append(coord_pair)
+
+    if geom_type == "Point":  # Point geometry vector layer
+
+        geometry = pd.Series(coords_list).apply(Point)
+        df = pd.DataFrame({"geometry": geometry})
+        gdf = gpd.GeoDataFrame(df, crs=epsg)
+
+        return gdf.to_file(dst_filepath, driver=driver) if dst_filepath else gdf
+
+    elif geom_type == "MultiPoint":  # MultiPoint geometry vector layer
+
+        geometry = MultiPoint(coords_list)
+        df = pd.DataFrame({"geometry": [geometry]})
+        gdf = gpd.GeoDataFrame(df, crs=epsg)
+
+        return gdf.to_file(dst_filepath, driver=driver) if dst_filepath else gdf
+
+    elif geom_type == "LineString":  # LineString geometry vector layer
+
+        geometry = LineString(coords_list)
+        df = pd.DataFrame({"geometry": [geometry]})
+        gdf = gpd.GeoDataFrame(df, crs=epsg)
+
+        return gdf.to_file(dst_filepath, driver=driver) if dst_filepath else gdf
+
+    elif geom_type == "Polygon":  # Poygon geometry vector layer
+
+        if coords_list[0] == coords_list[-1]:
+            geometry = Polygon(coords_list)
+            df = pd.DataFrame({"geometry": [geometry]})
+            gdf = gpd.GeoDataFrame(df, crs=epsg)
+
+            return gdf.to_file(dst_filepath, driver=driver) if dst_filepath else gdf
+
+        else:
+            raise Exception("To construct a Polygon geometry, make sure the first and last point are identical")
+
+
+def csv_to_vector_layer(csv_filepath: str, x_column: str, y_column: str, dst_filepath=None, driver=None,
+                        geom_type="Point", epsg=4326):
+
+    """
+    Function to create a vector layer from a csv file.
+    :param csv_filepath: path corresponding to the csv file
+    :param x_column: column containing longitude values
+    :param y_column: column containing latitude values
+    :param dst_filepath: If wanted, path to export the file
+    :param driver: driver to export the file
+    :param geom_type: geometry type
+    :param epsg: EPSG code of the geometry
+    :return:
+    """
+    if geom_type == "Point":
+        df = pd.read_csv(csv_filepath)
+        geometry = gpd.points_from_xy(df[x_column], df[y_column], crs=epsg)
+        gdf = gpd.GeoDataFrame(df, geometry=geometry)
+
+        return gdf
+
+    elif geom_type == "LineString":
+        df = pd.read_csv(csv_filepath)
+        geometry = gpd.points_from_xy(df[x_column], df[y_column], crs=epsg) # GeometryArray 1D
+        geometry = LineString(geometry)
+
+        gdf = gpd.GeoDataFrame({"tipo":["cuenca"], "geometry":[geometry]})
+
+        return gdf
+
+# txt_to_vector_layer(
+#     txt_filepath="C:\\Users\\orden\\Desktop\\Codigo\\data\\Cuenca_Parana.txt",
+#     dst_filepath="C:\\Users\\orden\\Desktop\\Codigo\\data\\Polygon_Cuenca_Parana.shp",
+#     driver="Shapefile",
+#     geom_type="Polygon"
+# )
+
+gdf = csv_to_vector_layer("C:\\Users\\orden\\Desktop\\Codigo\\data\\Cuenca_Parana.csv", "Longitud", "Latitud", geom_type="LineString")
+
+print(gdf)
+
+# df_cuenca_parana_point["geometry"] = df_cuenca_parana_point["geometry"].apply(Point)
+#
+# gdf_cuenca_parana_point = gpd.GeoDataFrame(df_cuenca_parana_point, geometry="geometry", crs=4326)
+
+# POLYLINE (LINESTRING)
+
+# 1° opcion
+# df_cuenca_parana_line = pd.DataFrame({"nombre":"Cuenca Paraná", "geometry":[cuenca_parana]})
+# df_cuenca_parana_line["geometry"] = df_cuenca_parana_line["geometry"].apply(LineString)
+
+# 2° opcion
+# cca_parana_linestring = LineString(cuenca_parana)
+# df_cuenca_parana_line = pd.DataFrame({"nombre":"Cuenca Paraná", "geometry":[cca_parana_linestring]})
+
+# gdf_cuenca_parana_line = gpd.GeoDataFrame(df_cuenca_parana_line, geometry="geometry", crs=4326)
+# gdf_cuenca_parana_line.to_file("C:/Users/Orden/Downloads/Cuenca_Parana.shp")
+
+# POLYGON
+
+# 1° opcion
+# df_cuenca_parana_polyg = pd.DataFrame({"nombre":"Cuenca Paraná", "geometry":[cuenca_parana]})
+# df_cuenca_parana_polyg["geometry"] = df_cuenca_parana_polyg["geometry"].apply(Polygon)
+
+# 2° opcion
+# cca_parana_polyg = Polygon(cuenca_parana)
+# df_cuenca_parana_polyg = pd.DataFrame({"nombre":"Cuenca Paraná", "geometry":[cca_parana_polyg]})
+
+# gdf_cuenca_parana_polyg = gpd.GeoDataFrame(df_cuenca_parana_polyg, geometry="geometry", crs=4326)
+# gdf_cuenca_parana_polyg.to_file("C:/Users/Orden/Downloads/Cuenca_Parana_Polyg.shp")