napari-musa 1.0.0__py3-none-any.whl

napari_musa/Widgets_Fusion.py

@@ -0,0 +1,458 @@
+""" """
+
+import sys
+from os.path import dirname
+
+sys.path.append(dirname(dirname(__file__)))
+import napari
+from magicgui.widgets import (
+    CheckBox,
+    ComboBox,
+    Container,
+    PushButton,
+    SpinBox,
+)
+from napari.utils.notifications import show_info
+from qtpy.QtCore import QTimer
+from qtpy.QtWidgets import (
+    QGroupBox,
+    QScrollArea,
+    QVBoxLayout,
+    QWidget,
+)
+
+from napari_musa.modules.functions import (
+    HSI2RGB,
+    datasets_fusion,
+)
+
+
+class Fusion(QWidget):
+    """ """
+
+    def __init__(self, viewer: napari.Viewer, data):
+        """ """
+        super().__init__()
+        self.viewer = viewer
+        self.data = data
+        self.init_ui()
+
+    def init_ui(self):
+        """ """
+        scroll = QScrollArea()
+        scroll.setWidgetResizable(True)
+        content_widget = QWidget()
+        content_layout = QVBoxLayout(content_widget)
+
+        # fusion_group = self.build_fusion_group()
+        # content_layout.addWidget(fusion_group)
+        content_layout.addWidget(self.create_fusion_controls_box())
+        content_layout.addWidget(self.create_midlevel_fusion_controls())
+        content_layout.addStretch()
+
+        scroll.setWidget(content_widget)
+        main_layout = QVBoxLayout(self)
+        main_layout.addWidget(scroll)
+        self.setLayout(main_layout)
+
+    def create_fusion_controls_box(self):
+        """ """
+        fusion_box = QGroupBox("Low-Level Fusion")
+        layout = QVBoxLayout()
+
+        self.reduced_dataset_checkbox = CheckBox(
+            text="Fuse the reduced datasets (Only if you have both reduced datasets)"
+        )
+
+        self.modes_fusion = ComboBox(
+            choices=[
+                "Frobenius norm",
+                "Z score",
+                "Z score - dataset",
+                "Z score - spectrum",
+                "SNV",
+                "Sum to one",
+                "Global min-max",
+                "Robust min-max",
+                "Pixel min-max",
+            ],
+            label="Select fusion modality",
+        )
+
+        self.modes_combobox_1 = ComboBox(
+            choices=self.data.modes, label="Select the first dataset"
+        )
+        self.modes_combobox_2 = ComboBox(
+            choices=self.data.modes,
+            label="Select the second dataset",
+            value="PL",
+        )
+        self.modes_combobox_3 = ComboBox(
+            choices=self.data.modes,
+            label="Select the third dataset",
+            value="-",
+        )
+
+        fusion_perform_btn = PushButton(text="Fuse the chosen datasets")
+        fusion_perform_btn.clicked.connect(self.fusion_perform_btn_f)
+
+        layout.addWidget(
+            Container(
+                widgets=[
+                    self.reduced_dataset_checkbox,
+                    self.modes_fusion,
+                    self.modes_combobox_1,
+                    self.modes_combobox_2,
+                    self.modes_combobox_3,
+                    fusion_perform_btn,
+                ]
+            ).native
+        )
+        fusion_box.setLayout(layout)
+        return fusion_box
+
+    def create_midlevel_fusion_controls(self):
+        """ """
+        fusion_box = QGroupBox("Mid-Level Fusion - PCA")
+        layout = QVBoxLayout()
+
+        self.PCA_components = SpinBox(
+            min=1, max=100, value=1, step=1, name="PCA components"
+        )
+
+        self.modes_combobox_1_MLF = ComboBox(
+            choices=self.data.modes, label="Select the first dataset"
+        )
+        self.modes_combobox_2_MLF = ComboBox(
+            choices=self.data.modes,
+            label="Select the second dataset",
+            value="PL",
+        )
+        self.modes_combobox_3_MLF = ComboBox(
+            choices=self.data.modes,
+            label="Select the third dataset",
+            value="-",
+        )
+
+        MLfusion_perform_btn = PushButton(text="Fuse the chosen datasets")
+        MLfusion_perform_btn.clicked.connect(self.MLfusion_perform_btn_f)
+
+        layout.addWidget(
+            Container(
+                widgets=[
+                    self.PCA_components,
+                    self.modes_combobox_1_MLF,
+                    self.modes_combobox_2_MLF,
+                    self.modes_combobox_3_MLF,
+                    MLfusion_perform_btn,
+                ]
+            ).native
+        )
+        fusion_box.setLayout(layout)
+        return fusion_box
+
+    def fusion_perform_btn_f(self):
+        """ """
+        # Parameters
+        self.data.fusion_norm = self.modes_fusion.value  # Norm used for fusion
+        self.data.fusion_modes = (
+            []
+        )  # Tells which datasets are fused (Refle-PL, PL1-PL2, etc.)
+        self.data.fusion_modes.append(self.modes_combobox_1.value)
+        self.data.fusion_modes.append(self.modes_combobox_2.value)
+        wl1 = self.data.wls[self.modes_combobox_1.value]
+        wl2 = self.data.wls[self.modes_combobox_2.value]
+        #
+        # If a third dataset is selected
+        if self.modes_combobox_3.value != "-":
+            self.data.fusion_modes.append(self.modes_combobox_3.value)
+            wl3 = self.data.wls[self.modes_combobox_1.value]
+        #
+        # If we want to fuse the reduced datasets
+        if self.reduced_dataset_checkbox.value:
+            # If 1 and 2 have the spatial reduction (should be when I have both spatial and spectral reduction)
+            # First are fused the spatial reduced datasets
+            if (
+                self.data.hypercubes_spatial_red.get(
+                    self.modes_combobox_1.value
+                )
+                is not None
+                and self.data.hypercubes_spatial_red.get(
+                    self.modes_combobox_2.value
+                )
+                is not None
+            ):
+                dataset1 = self.data.hypercubes_spatial_red[
+                    self.modes_combobox_1.value
+                ]
+                dataset2 = self.data.hypercubes_spatial_red[
+                    self.modes_combobox_2.value
+                ]
+                (
+                    self.data.hypercubes_spatial_red[
+                        "Fused"
+                    ],  # Fusion of spatial reduced datasets
+                    self.data.wls["Fused"],
+                    self.data.fusion_params,
+                ) = datasets_fusion(
+                    dataset1, dataset2, wl1, wl2, norm=self.modes_fusion.value
+                )
+            # dataset1 and dataset2 for the fusion
+            dataset1 = self.data.hypercubes_red[self.modes_combobox_1.value]
+            dataset2 = self.data.hypercubes_red[self.modes_combobox_2.value]
+
+            self.data.rgb_red["Fused"] = self.data.rgb_red[
+                self.modes_combobox_1.value
+            ]  # The rgb of the fused is the rgb of dataset 1
+            (
+                self.data.hypercubes_red["Fused"],
+                self.data.wls["Fused"],
+                self.data.fusion_params,
+            ) = datasets_fusion(
+                dataset1, dataset2, wl1, wl2, norm=self.modes_fusion.value
+            )
+            # If we have a third dataset
+            if self.modes_combobox_3.value != "-":
+                if (
+                    self.data.hypercubes_spatial_red.get(
+                        self.modes_combobox_3.value
+                    )
+                    is not None
+                ):  # If we apply both spatial and spectral reduction, append 3rd spatial reduced dataset to the fused
+                    # spatial reduced datasets
+                    dataset3 = self.data.hypercubes_spatial_red[
+                        self.modes_combobox_3.value
+                    ]
+                    (
+                        self.data.hypercubes_spatial_red["Fused"],
+                        self.data.wls["Fused"],
+                        self.data.fusion_params,
+                    ) = datasets_fusion(
+                        self.data.hypercubes_spatial_red["Fused"],
+                        dataset3,
+                        self.data.wls["Fused"],
+                        wl3,
+                        norm=self.modes_fusion.value,
+                    )
+                dataset3 = self.data.hypercubes_red[
+                    self.modes_combobox_3.value
+                ]
+                (
+                    self.data.hypercubes_red["Fused"],
+                    self.data.wls["Fused"],
+                    self.data.fusion_params,
+                ) = datasets_fusion(
+                    self.data.hypercubes_red["Fused"],
+                    dataset3,
+                    self.data.wls["Fused"],
+                    wl3,
+                    norm=self.modes_fusion.value,
+                )
+            self.viewer.add_image(
+                self.data.hypercubes_red["Fused"].transpose(2, 0, 1),
+                name="Fused",
+                metadata={"type": "hyperspectral_cube"},
+            )
+        #
+        # If the dataset are not reduced
+        else:
+            dataset1 = self.data.hypercubes[self.modes_combobox_1.value]
+            dataset2 = self.data.hypercubes[self.modes_combobox_2.value]
+            if self.data.rgb.get(self.modes_combobox_1.value) is not None:
+                self.data.rgb["Fused"] = self.data.rgb[
+                    self.modes_combobox_1.value
+                ]
+            else:  # If we did not create the rgb of dataset 1, we compute it here
+                self.data.rgb[self.modes_combobox_1.value] = HSI2RGB(
+                    self.data.wls[self.modes_combobox_1.value],
+                    dataset1,
+                    dataset1.shape[0],
+                    dataset1.shape[1],
+                    65,
+                    False,
+                )
+
+                self.viewer.add_image(
+                    self.data.rgb[self.modes_combobox_1.value],
+                    name=str(self.modes_combobox_1.value) + " RGB",
+                    metadata={"type": "rgb"},
+                )
+            # Compute the fusion
+            (
+                self.data.hypercubes["Fused"],
+                self.data.wls["Fused"],
+                self.data.fusion_params,
+            ) = datasets_fusion(
+                dataset1, dataset2, wl1, wl2, norm=self.modes_fusion.value
+            )
+            # If we have a third dataset
+            if self.modes_combobox_3.value != "-":
+                dataset3 = self.data.hypercubes[self.modes_combobox_3.value]
+                (
+                    self.data.hypercubes["Fused"],
+                    self.data.wls["Fused"],
+                    self.data.fusion_params,
+                ) = datasets_fusion(
+                    self.data.hypercubes["Fused"],
+                    dataset3,
+                    self.data.wls["Fused"],
+                    wl3,
+                    norm=self.modes_fusion.value,
+                )
+            self.viewer.add_image(
+                self.data.hypercubes["Fused"].transpose(2, 0, 1),
+                name="Fused",
+                metadata={"type": "hyperspectral_cube"},
+            )
+        QTimer.singleShot(0, lambda: show_info("Fusion completed!"))
+
+    # xxx sistemarte
+    def MLfusion_perform_btn_f(self):
+        """ """
+        # Parameters
+        self.data.fusion_norm = self.modes_fusion.value  # Norm used for fusion
+        self.data.fusion_modes = (
+            []
+        )  # Tells which datasets are fused (Refle-PL, PL1-PL2, etc.)
+        self.data.fusion_modes.append(self.modes_combobox_1.value)
+        self.data.fusion_modes.append(self.modes_combobox_2.value)
+        wl1 = self.data.wls[self.modes_combobox_1.value]
+        wl2 = self.data.wls[self.modes_combobox_2.value]
+        #
+        # If a third dataset is selected
+        if self.modes_combobox_3.value != "-":
+            self.data.fusion_modes.append(self.modes_combobox_3.value)
+            wl3 = self.data.wls[self.modes_combobox_1.value]
+        #
+        # If we want to fuse the reduced datasets
+        if self.reduced_dataset_checkbox.value:
+            # If 1 and 2 have the spatial reduction (should be when I have both spatial and spectral reduction)
+            # First are fused the spatial reduced datasets
+            if (
+                self.data.hypercubes_spatial_red.get(
+                    self.modes_combobox_1.value
+                )
+                is not None
+                and self.data.hypercubes_spatial_red.get(
+                    self.modes_combobox_2.value
+                )
+                is not None
+            ):
+                dataset1 = self.data.hypercubes_spatial_red[
+                    self.modes_combobox_1.value
+                ]
+                dataset2 = self.data.hypercubes_spatial_red[
+                    self.modes_combobox_2.value
+                ]
+                (
+                    self.data.hypercubes_spatial_red[
+                        "Fused"
+                    ],  # Fusion of spatial reduced datasets
+                    self.data.wls["Fused"],
+                    self.data.fusion_params,
+                ) = datasets_fusion(
+                    dataset1, dataset2, wl1, wl2, norm=self.modes_fusion.value
+                )
+            # dataset1 and dataset2 for the fusion
+            dataset1 = self.data.hypercubes_red[self.modes_combobox_1.value]
+            dataset2 = self.data.hypercubes_red[self.modes_combobox_2.value]
+
+            self.data.rgb_red["Fused"] = self.data.rgb_red[
+                self.modes_combobox_1.value
+            ]  # The rgb of the fused is the rgb of dataset 1
+            (
+                self.data.hypercubes_red["Fused"],
+                self.data.wls["Fused"],
+                self.data.fusion_params,
+            ) = datasets_fusion(
+                dataset1, dataset2, wl1, wl2, norm=self.modes_fusion.value
+            )
+            # If we have a third dataset
+            if self.modes_combobox_3.value != "-":
+                if (
+                    self.data.hypercubes_spatial_red.get(
+                        self.modes_combobox_3.value
+                    )
+                    is not None
+                ):  # If we apply both spatial and spectral reduction, append 3rd spatial reduced dataset to the fused
+                    # spatial reduced datasets
+                    dataset3 = self.data.hypercubes_spatial_red[
+                        self.modes_combobox_3.value
+                    ]
+                    (
+                        self.data.hypercubes_spatial_red["Fused"],
+                        self.data.wls["Fused"],
+                        self.data.fusion_params,
+                    ) = datasets_fusion(
+                        self.data.hypercubes_spatial_red["Fused"],
+                        dataset3,
+                        self.data.wls["Fused"],
+                        wl3,
+                        norm=self.modes_fusion.value,
+                    )
+                dataset3 = self.data.hypercubes_red[
+                    self.modes_combobox_3.value
+                ]
+                (
+                    self.data.hypercubes_red["Fused"],
+                    self.data.wls["Fused"],
+                    self.data.fusion_params,
+                ) = datasets_fusion(
+                    self.data.hypercubes_red["Fused"],
+                    dataset3,
+                    self.data.wls["Fused"],
+                    wl3,
+                    norm=self.modes_fusion.value,
+                )
+        #
+        # If the dataset are not reduced
+        else:
+            dataset1 = self.data.hypercubes[self.modes_combobox_1.value]
+            dataset2 = self.data.hypercubes[self.modes_combobox_2.value]
+            if self.data.rgb.get(self.modes_combobox_1.value) is not None:
+                self.data.rgb["Fused"] = self.data.rgb[
+                    self.modes_combobox_1.value
+                ]
+            else:  # If we did not create the rgb of dataset 1, we compute it here
+                self.data.rgb[self.modes_combobox_1.value] = HSI2RGB(
+                    self.data.wls[self.modes_combobox_1.value],
+                    dataset1,
+                    dataset1.shape[0],
+                    dataset1.shape[1],
+                    65,
+                    False,
+                )
+
+                self.viewer.add_image(
+                    self.data.rgb[self.modes_combobox_1.value],
+                    name=str(self.modes_combobox_1.value) + " RGB",
+                    metadata={"type": "rgb"},
+                )
+            # Compute the fusion
+            (
+                self.data.hypercubes["Fused"],
+                self.data.wls["Fused"],
+                self.data.fusion_params,
+            ) = datasets_fusion(
+                dataset1, dataset2, wl1, wl2, norm=self.modes_fusion.value
+            )
+            # If we have a third dataset
+            if self.modes_combobox_3.value != "-":
+                dataset3 = self.data.hypercubes[self.modes_combobox_3.value]
+                (
+                    self.data.hypercubes["Fused"],
+                    self.data.wls["Fused"],
+                    self.data.fusion_params,
+                ) = datasets_fusion(
+                    self.data.hypercubes["Fused"],
+                    dataset3,
+                    self.data.wls["Fused"],
+                    wl3,
+                    norm=self.modes_fusion.value,
+                )
+        self.viewer.add_image(
+            self.data.hypercubes["Fused"].transpose(2, 0, 1),
+            name="Fused",
+            metadata={"type": "hyperspectral_cube"},
+        )
+        QTimer.singleShot(0, lambda: show_info("Fusion completed!"))