Update app.py

app.py

@@ -1,23 +1,58 @@
 import gradio as gr
 import cv2
 import numpy as np
-
+from Utils.Image_stitching import *
 # Define a function that performs the image operation
-def perform_image_operation(image1, image2):
-    # Perform the image operation (e.g., blending)
-    alpha = 0.5  # Adjust alpha for blending
-    blended_image = cv2.addWeighted(image1, alpha, image2, 1 - alpha, 0)
+def perform_image_operation(image1, image2,image3,image4,image5,image6,image7,image8,image9):
+    for dataset_name, img_files_dataset in dataset_img_files.items():
+        # Retrieve the batch size for the current dataset
+        batch_size = 9
+
+        subfolder_paths = dataset_subfolder_paths[dataset_name]
+        visual_subfolder, analysis_subfolder, preprocess_subfolder, segmented_subfolder, stitched_subfolder, dataframe = subfolder_paths
+
+        # Your existing code here for reading, processing, and saving images
+        
+        # Loop through the images in analysis_subfolder and perform image stitching
+        analysis_images = []  # Create an empty list to store image paths from the analysis_folder
+        for root, _, filenames in os.walk(analysis_subfolder):
+            for filename in filenames:
+                if filename.endswith(".png"):  # You can adjust the file extension if needed
+                    image_path = os.path.join(root, filename)
+                    analysis_images.append(image_path)
+
+        num_batches = len(analysis_images) // batch_size
+
+        for batch_idx in range(num_batches):
+            start_idx = batch_idx * batch_size
+            end_idx = start_idx + batch_size
+            batch_img_indices = analysis_images[start_idx:end_idx]  # Use images from analysis_folder
+
+            # Call the image_stitching function with the batch of image paths
+            stitched_batch = image_stitching(batch_img_indices)
+
+            # Save the stitched batch image
+            batch_filename = '{}_batch_{}.png'.format(dataset_name, batch_idx)  # Include a unique identifier for the batch
+            stitched_batch_img = Image.fromarray(np.uint8(stitched_batch))
+            
 
     # Convert the result image to PIL format for Gradio display
-    result_pil_image = cv2.cvtColor(blended_image, cv2.COLOR_BGR2RGB)
-    return result_pil_image
+            result_pil_image = cv2.cvtColor(stitched_batch_img, cv2.COLOR_BGR2RGB)
+            return result_pil_image
 
 # Define the Gradio interface
 iface = gr.Interface(
     fn=perform_image_operation,
     inputs=[
         gr.inputs.Image(label="Input Image 1"),
-        gr.inputs.Image(label="Input Image 2")
+        gr.inputs.Image(label="Input Image 2"),
+        gr.inputs.Image(label="Input Image 3"),
+        gr.inputs.Image(label="Input Image 4"),
+        gr.inputs.Image(label="Input Image 5"),
+        gr.inputs.Image(label="Input Image 6"),
+        gr.inputs.Image(label="Input Image 7"),
+        gr.inputs.Image(label="Input Image 8"),
+        gr.inputs.Image(label="Input Image 9")
     ],
     outputs=gr.outputs.Image(type="pil", label="Blended Image")
 )