Update app.py

app.py

@@ -32,7 +32,7 @@ outputs = ["detection_boxes:0", "detection_scores:0", "detection_classes:0", "nu
 detection_fn = wrap_frozen_graph(graph_def, inputs, outputs)
 
 # TensorFlow function for detection
-@tf.function(input_signature=[tf.TensorSpec(shape=[None, None, None, 3], dtype=tf.uint8)])
+@tf.function(input_signature=[tf.TensorSpec(shape=[1, None, None, 3], dtype=tf.uint8)])
 def detect_objects(image):
     return detection_fn(image)
 
@@ -139,8 +139,20 @@ def match_and_identify(features, bbox):
 
         return identity, color
 
+import numpy as np
+import cv2
+import tensorflow as tf
+from tensorflow.keras.applications import ResNet50
+from tensorflow.keras.applications.resnet50 import preprocess_input
+from sklearn.metrics.pairwise import cosine_similarity
+from filterpy.kalman import KalmanFilter
+import gradio as gr
+
+# ... (previous code remains the same) ...
+
 def process_image(image):
     if image is None:
+        print("Input image is None")
         return None
 
     # Convert image to RGB if it's not
@@ -167,6 +179,8 @@ def process_image(image):
         classes = detections[2].numpy()[0]
         num_detections = int(detections[3].numpy()[0])
 
+        print(f"Number of detections: {num_detections}")
+
         # Filter detections for 'person' class
         threshold = 0.3  # Adjust this threshold as needed
         for i in range(num_detections):
@@ -196,13 +210,17 @@ def process_image(image):
                 cv2.rectangle(image, (left, top), (right, bottom), color, 2)
                 cv2.putText(image, f'Person {identity}', (left, top - 10), cv2.FONT_HERSHEY_SIMPLEX, 0.5, color, 2)
 
+                print(f"Detected person {identity} at ({left}, {top}, {right}, {bottom})")
+
     except Exception as e:
         print(f"Error during processing: {str(e)}")
         return image  # Return original image if there's an error
 
     return image
+
 def gradio_interface(input_image):
     if input_image is None:
+        print("Input image is None")
         return None
     
     # Convert PIL Image to numpy array if necessary
@@ -211,6 +229,18 @@ def gradio_interface(input_image):
     
     # Process the input image
     output_image = process_image(input_image)
+    
+    if output_image is None:
+        print("Output image is None")
+        return None
+    
+    print(f"Output image shape: {output_image.shape}")
+    print(f"Output image dtype: {output_image.dtype}")
+    
+    # Ensure the output is in the correct format for Gradio
+    if output_image.dtype != np.uint8:
+        output_image = (output_image * 255).astype(np.uint8)
+    
     return output_image
 
 # Create Gradio interface
@@ -219,7 +249,7 @@ iface = gr.Interface(
     inputs=gr.Image(),
     outputs=gr.Image(),
     title="Person Detection and Tracking",
-    description="Upload an image to detect and track persons."
+    description="Upload an image to detect and track persons.",
 )
 
 # Launch the interface