Update app.py

app.py

@@ -124,44 +124,24 @@
 # demo.launch()
 
 
-
-import torch
-import numpy as np
+from PIL import Image  # Importing Image from PIL (Pillow)
 import cv2
+import numpy as np
 import json
-import os
-import gradio as gr
-from detectron2.detectron2.engine import DefaultPredictor
-from detectron2.detectron2.config import get_cfg
-from detectron2 import model_zoo
-import torch_utils
-import dnnlib
-import requests
 import base64
+import requests
 from io import BytesIO
 
-# Create output directory if it doesn't exist
-output_dir = "key/"
-os.makedirs(output_dir, exist_ok=True)
-output_file = os.path.join(output_dir, "keypoints.json")
-
-# Load pre-trained Keypoint R-CNN model
-cfg = get_cfg()
-cfg.merge_from_file(model_zoo.get_config_file("COCO-Keypoints/keypoint_rcnn_R_50_FPN_3x.yaml"))
-cfg.MODEL.WEIGHTS = model_zoo.get_checkpoint_url("COCO-Keypoints/keypoint_rcnn_R_50_FPN_3x.yaml")
-cfg.MODEL.DEVICE = "cuda" if torch.cuda.is_available() else "cpu"
-
-# Load the predictor
-predictor = DefaultPredictor(cfg)
-
-# Replace with your actual Ngrok or backend URL
-NGROK_URL = "https://your-ngrok-url.ngrok.io/measurements"  # Change this
-
+# Function to process the image and calculate measurements
 def process_image(image, user_height_cm):
+    # Check if the user height is valid
+    if user_height_cm == 0 or user_height_cm is None:
+        return "User height cannot be zero or None. Please provide a valid height.", None
+
     # Convert Gradio image input to OpenCV format
     image = cv2.cvtColor(np.array(image), cv2.COLOR_RGB2BGR)
 
-    # Run keypoint detection
+    # Run keypoint detection (using your predictor)
     outputs = predictor(image)
 
     # Extract keypoints
@@ -172,6 +152,7 @@ def process_image(image, user_height_cm):
         return "No keypoints detected.", None
 
     # Save keypoints to JSON
+    output_file = "keypoints.json"
     with open(output_file, "w") as f:
         json.dump({"keypoints": keypoints}, f, indent=4)
 
@@ -205,10 +186,18 @@ def process_image(image, user_height_cm):
     ankle_mid = ((keypoints[L_ANKLE] + keypoints[R_ANKLE]) / 2).tolist()
     pixel_height = get_distance(keypoints[NOSE], ankle_mid)
 
+    # Check if pixel height is non-zero to avoid divide by zero error
+    if pixel_height == 0:
+        return "Error in calculating pixel height. Please check the input image.", None
+
     # Estimated full body height (add approx head length)
     estimated_full_pixel_height = pixel_height / 0.87  # Since 87% = nose to ankle
     pixels_per_cm = estimated_full_pixel_height / user_height_cm
 
+    # Ensure pixels_per_cm is valid (non-zero)
+    if pixels_per_cm == 0:
+        return "Error in calculating pixels per cm. Ensure the user height is correct.", None
+
     # Waist and shoulder measurements
     shoulder_width_px = get_distance(keypoints[L_SHOULDER], keypoints[R_SHOULDER])
     waist_width_px = get_distance(keypoints[L_HIP], keypoints[R_HIP])
@@ -253,13 +242,13 @@ def save_to_database(measurements, image, user_height_cm):
 
     # Convert image to base64
     buffered = BytesIO()
-    pil_image = Image.fromarray(image)
+    pil_image = Image.fromarray(image)  # Convert to PIL Image from NumPy array
     pil_image.save(buffered, format="JPEG")
     img_str = base64.b64encode(buffered.getvalue()).decode("utf-8")
 
-    # Send POST request
+    # Send POST request to save measurements and image to the database
     response = requests.post(
-        "https://9dbc-210-212-162-140.ngrok-free.app/upload",  # Change this
+        "https://9dbc-210-212-162-140.ngrok-free.app/upload",  # Replace with your actual URL
         json={
             "imageBase64": img_str,
             "heightCm": user_height_cm,
@@ -273,28 +262,25 @@ def save_to_database(measurements, image, user_height_cm):
 
     return f"Status: {response.status_code}, Message: {response.text}"
 
-# Gradio Interface
-def create_interface():
-    with gr.Blocks() as demo:
-        with gr.Row():
-            img_input = gr.Image(type="pil", label="Upload Image")
-            height_input = gr.Number(label="User Height (cm)")
-        
-        with gr.Row():
-            output_measurements = gr.JSON(label="Measurements")
-            output_image = gr.Image(type="pil", label="Keypoint Overlay")
-
-        save_btn = gr.Button("Save to Database")
-        save_status = gr.Textbox(label="Save Status", interactive=False)
-
-        img_input.change(fn=process_image, inputs=[img_input, height_input], outputs=[output_measurements, output_image])
-        save_btn.click(
-            fn=save_to_database,
-            inputs=[output_measurements, output_image, height_input],
-            outputs=save_status
-        )
-
-    return demo
-
-demo = create_interface()
-demo.launch()
+# Example usage (integrating with your Gradio app)
+import gradio as gr
+
+# Gradio interface setup
+def process_and_save(image, user_height_cm):
+    measurements, processed_image = process_image(image, user_height_cm)
+    if measurements:
+        save_message = save_to_database(measurements, processed_image, user_height_cm)
+        return processed_image, measurements, save_message
+    else:
+        return None, None, "Error in processing image."
+
+# Gradio interface setup
+iface = gr.Interface(
+    fn=process_and_save,
+    inputs=[gr.Image(type="numpy"), gr.Number(label="User Height (cm)")],
+    outputs=[gr.Image(type="numpy"), gr.JSON(), gr.Textbox()],
+    live=True
+)
+
+iface.launch(share=True)
+