Merge ML Indices

app.py

@@ -91,13 +91,21 @@ container = st.container()
 
 with st.expander("Advanced Settings"):
     st.write("Select the vegetation indices to calculate:")
-    all_veg_indices = ["NDVI", "EVI", "EVI2"]
+    all_veg_indices = ["GujEVI", "NDVI", "EVI", "EVI2", "RandomForest", "GujVDI", "MNDWI", "SAVI", "MVI", "NBR", "GCI"]
     formulas = {
+        "GujEVI": r"$0.5 \times \frac{NIR - Red}{NIR + 6 \times Red - 8.25 \times Blue + 0.01}, \text{(Optimized EVI for Gujarat)}$",
         "NDVI": r"$\frac{NIR - Red}{NIR + Red}$",
         "EVI": r"$G \times \frac{NIR - Red}{NIR + C1 \times Red - C2 \times Blue + L}$",
         "EVI2": r"$G \times \frac{NIR - Red}{NIR + L + C \times Red}$",
+        "RandomForest": "ML based Classification",
+        "GujVDI": r"$2.29 \times \frac{-3.98 \left(\frac{Blue}{NIR}\right) + 12.54 \left(\frac{Green}{NIR}\right) - 5.49 \left(\frac{Red}{NIR}\right) - 0.19}{-21.87 \left(\frac{Blue}{NIR}\right) + 12.4 \left(\frac{Green}{NIR}\right) + 19.98 \left(\frac{Red}{NIR}\right) + 1}$",
+        "MNDWI": r"$\frac{Green - SWIR}{Green + SWIR}$",
+        "SAVI": r"$\frac{(1 + L) \times (NIR - Red)}{NIR + Red + L}, \text{ where } L=0.5$",
+        "MVI": r"$\frac{NIR - (Green + SWIR)}{NIR + (Green + SWIR)}$",
+        "NBR": r"$\frac{NIR - SWIR2}{NIR + SWIR2}$",
+        "GCI": r"$\frac{NIR - Green}{Green}$",
     }
-    defaults = [True, False, False]
+    defaults = [True, False, False, False, False, False, False, False, False, False, False]
     veg_indices = []
     for veg_index, default in zip(all_veg_indices, defaults):
         if st.checkbox(f"{veg_index} = {formulas[veg_index]}", value=default):
@@ -338,10 +346,11 @@ if submit:
             "scale": 10,
             "collection": ee.ImageCollection("COPERNICUS/S2_SR_HARMONIZED")
             .select(
-                ["B2", "B4", "B8", "MSK_CLDPRB", "TCI_R", "TCI_G", "TCI_B"],
-                ["Blue", "Red", "NIR", "MSK_CLDPRB", "R", "G", "B"],
+                ["B2", "B3", "B4", "B8", "B11", "B12", "MSK_CLDPRB", "TCI_R", "TCI_G", "TCI_B"],
+                ["Blue", "Green", "Red", "NIR", "SWIR", "SWIR2", "MSK_CLDPRB", "R", "G", "B"],
             )
-            .map(lambda image: add_indices(image, nir_band="NIR", red_band="Red", blue_band="Blue", evi_vars=evi_vars)),
+            .map(lambda image: add_indices(image, nir_band="NIR", red_band="Red", blue_band="Blue", green_band="Green", 
+                                           swir_band="SWIR", swir2_band = "SWIR2", evi_vars=evi_vars)),
         },
     }
     satellite = list(satellites.keys())[0]
@@ -444,11 +453,6 @@ if "result" in st.session_state:
             "Year 2", result_df.index, index=len(result_df.index) - 1, format_func=lambda x: daterange_str_to_year(x)
         )
 
-    vis_params = {"min": 0, "max": 1, "palette": ["white", "green"]}  # Example visualisation for Sentinel-2
-
-    # Create a colormap and name it as NDVI
-    colormap = cm.LinearColormap(colors=vis_params["palette"], vmin=vis_params["min"], vmax=vis_params["max"])
-
     for veg_index in veg_indices:
         st.write(f"<h3><div style='text-align: center;'>{veg_index}</div></h3>", unsafe_allow_html=True)
         cols = st.columns(2)
@@ -469,9 +473,9 @@ if "result" in st.session_state:
                 else:
                     raise ValueError(f"Unknown satellite: {satellite}")
 
-                if veg_index == "NDVI":
+                if veg_index == "Test":
                     bins = [-1, 0, 0.1, 0.2, 0.3, 0.4, 0.5, 1]
-                    histogram, bin_edges = get_histogram(mosaic.select(veg_index), ee_geometry, bins)
+                    histogram, bin_edges = get_histogram(veg_index, mosaic.select(veg_index), ee_geometry, bins)
                     total_pix = np.sum(histogram)
                     formatted_histogram = [f"{h*100/total_pix:.2f}" for h in histogram]
                     print(histogram, bin_edges, bins, formatted_histogram)
@@ -496,12 +500,94 @@ if "result" in st.session_state:
                     }
                     m.add_layer(mosaic.select(veg_index).clip(outer_ee_geometry), ndvi_vis_params)
 
-                # add colorbar
-                # m.add_colorbar(colors=["#000000", "#00FF00"], vmin=0.0, vmax=1.0)
+                elif veg_index in ["NDVI", "RandomForest", "GujVDI", "GujEVI", "EVI", "EVI2", "SAVI"]:
+                    bins = [0,0.2,0.4,0.6,0.8,1]
+                    histogram, bin_edges = get_histogram(veg_index, mosaic.select(veg_index), ee_geometry, bins)
+                    total_pix = np.sum(histogram)
+                    formatted_histogram = [f"{h*100/total_pix:.2f}" for h in histogram]
+                    print(histogram, bin_edges, bins, formatted_histogram)
+                    m.add_legend(
+                        title=f"{veg_index} Class/Value",
+                        legend_dict={
+                            #"<0:Waterbody ({}%)".format(formatted_histogram[0]): "#0000FF",
+                            "0-0.2: Open/Sparse Vegetation Density ({}%)".format(formatted_histogram[0]): "#FF0000",
+                            "0.2-0.4: Low Vegetation Density ({}%)".format(formatted_histogram[1]): "#FFFF00",
+                            "0.4-0.6: Moderate Vegetation Density ({}%)".format(formatted_histogram[2]): "#FFA500",
+                            "0.6-0.8: Dense Vegetation ({}%)".format(formatted_histogram[3]): "#00FE00",
+                            "0.8-1: Very Dense Vegetation ({}%)".format(formatted_histogram[4]): "#00A400",
+                            #">0.5: Very Dense ({}%)".format(formatted_histogram[6]): "#006D00",
+                        },
+                        position="bottomright",
+                        draggable=False,
+                    )
+                    ind_vis_params = {
+                        "min": 0,
+                        "max": 1,
+                        "palette": ["#FF0000", "#FFFF00", "#FFA500", "#00FE00", "#00A400"],
+                    }
+                    m.add_layer(mosaic.select(veg_index).clip(outer_ee_geometry), ind_vis_params)
 
-                if veg_index != "NDVI":
+                elif veg_index in ["MNDWI"]:
+                    bins = [-0.8, -0.6, -0.4, -0.2, 0.0, 0.2, 1]
+                    histogram, bin_edges = get_histogram(veg_index, mosaic.select(veg_index), ee_geometry, bins)
+                    total_pix = np.sum(histogram)
+                    formatted_histogram = [f"{h*100/total_pix:.2f}" for h in histogram]
+                    print(histogram, bin_edges, bins, formatted_histogram)
+                    m.add_legend(
+                        title=f"{veg_index} Class/Value",
+                        legend_dict={
+                            #"<0:Waterbody ({}%)".format(formatted_histogram[0]): "#0000FF",
+                            "-0.8 to -0.6:  ({}%)".format(formatted_histogram[0]): "#FF0000",
+                            "-0.6 to -0.4: ({}%)".format(formatted_histogram[1]): "#FFFF00",
+                            "-0.4 to -0.2:  ({}%)".format(formatted_histogram[2]): "#FFA500",
+                            "-0.2 to 0.0: ({}%)".format(formatted_histogram[3]): "#00FE00",
+                            "0.0 to 0.2:  ({}%)".format(formatted_histogram[4]): "#00A400",
+                            ">0.2: Very Dense ({}%)".format(formatted_histogram[5]): "#006D00",
+                        },
+                        position="bottomright",
+                        draggable=False,
+                    )
+                    ind_vis_params = {
+                        "min": -0.8,
+                        "max": 1,
+                        "palette": ["#FF0000", "#FFFF00", "#FFA500", "#00FE00", "#00A400", "#006D00"],
+                    }
+                    m.add_layer(mosaic.select(veg_index).clip(outer_ee_geometry), ind_vis_params)
+
+                elif veg_index in ["NBR"]:
+                    bins = [-1, 0.1, 0.275, 0.45, 0.65, 1]
+                    histogram, bin_edges = get_histogram(veg_index, mosaic.select(veg_index), ee_geometry, bins)
+                    total_pix = np.sum(histogram)
+                    formatted_histogram = [f"{h*100/total_pix:.2f}" for h in histogram]
+                    print(histogram, bin_edges, bins, formatted_histogram)
+                    m.add_legend(
+                        title=f"{veg_index} Class/Value",
+                        legend_dict={
+                            #"<0:Waterbody ({}%)".format(formatted_histogram[0]): "#0000FF",
+                            "-1 to 0.1: Unburned ({}%)".format(formatted_histogram[0]): "#00A400" ,
+                            "0.1 to 0.275: Low-severity burn ({}%)".format(formatted_histogram[1]): "#00FE00",
+                            "0.275 to 0.45: Moderate-to-low severity burn ({}%)".format(formatted_histogram[2]): "#FFA500",
+                            "0.45 to 0.66: Moderate-to-high severity burn ({}%)".format(formatted_histogram[3]): "#FFFF00",
+                            "0.66 to 1:  High-severity burn ({}%)".format(formatted_histogram[4]): "#FF0000",
+                        },
+                        position="bottomright",
+                        draggable=False,
+                    )
+                    ind_vis_params = {
+                        "min": -1,
+                        "max": 1,
+                        "palette": ["#FF0000", "#FFFF00", "#FFA500", "#00FE00", "#00A400", "#006D00"],
+                    }
+                    m.add_layer(mosaic.select(veg_index).clip(outer_ee_geometry), ind_vis_params)
+
+                
+                else:
+                    # For GCI
+                    vis_params = {"min": 1, "max": 10, "palette": ["white", "green"]}  
+                    colormap = cm.LinearColormap(colors=vis_params["palette"], vmin=vis_params["min"], vmax=vis_params["max"])
                     m.add_layer(mosaic.select(veg_index).clip(outer_ee_geometry), vis_params)
                     m.add_child(colormap)
+                
                 add_geometry_to_maps([m], geometry_gdf, buffer_geometry_gdf)
                 m.to_streamlit()