Spaces:

yassonee
/

knoch

Sleeping

App Files Files

yassonee commited on Jan 27

Commit

d813e66

verified ·

1 Parent(s): d21ffe2

Update app.py

Browse files

Files changed (1) hide show

app.py +301 -28

app.py CHANGED Viewed

@@ -1,68 +1,341 @@
 from fastapi import FastAPI, File, UploadFile
 from fastapi.responses import HTMLResponse
 from transformers import pipeline
-from PIL import Image
 import io
 import uvicorn
 app = FastAPI()
 # Chargement des modèles
 def load_models():
     return {
-        "model": pipeline("image-classification", model="Heem2/bone-fracture-detection-using-xray")
     }
 models = load_models()
-# Page d'accueil simple
 @app.get("/", response_class=HTMLResponse)
 async def main():
     content = """
-    <body style="font-family: Arial; max-width: 800px; margin: 0 auto; padding: 20px;">
-        <h1>Fraktur Detektion</h1>
-        <form action="/upload" enctype="multipart/form-data" method="post">
-            <input type="file" name="file" accept="image/*"><br><br>
-            <input type="submit" value="Analysieren">
-        </form>
     </body>
     """
     return content
-# Route pour l'upload et l'analyse
-@app.post("/upload", response_class=HTMLResponse)
-async def upload_file(file: UploadFile = File(...)):
     try:
         # Lecture de l'image
         contents = await file.read()
         image = Image.open(io.BytesIO(contents))
-        # Analyse
-        predictions = models["model"](image)
-        # Formatage des résultats
-        results = "<h2>Ergebnisse:</h2>"
-        for pred in predictions:
-            confidence = pred['score'] * 100
-            label = "Knochenbruch" if "fracture" in pred['label'].lower() else "Kein Knochenbruch"
-            color = "red" if "fracture" in pred['label'].lower() else "green"
-            results += f'<p style="color: {color};">{label}: {confidence:.1f}%</p>'
-        return f"""
-        <body style="font-family: Arial; max-width: 800px; margin: 0 auto; padding: 20px;">
-            <h1>Analyse Ergebnisse</h1>
-            {results}
-            <br>
-            <a href="/">Zurück</a>
         </body>
         """
     except Exception as e:
         return f"""
         <body style="font-family: Arial; max-width: 800px; margin: 0 auto; padding: 20px;">
             <h1>Fehler</h1>
             <p>Ein Fehler ist aufgetreten: {str(e)}</p>
             <br>
-            <a href="/">Zurück</a>
         </body>
         """

 from fastapi import FastAPI, File, UploadFile
 from fastapi.responses import HTMLResponse
 from transformers import pipeline
+from PIL import Image, ImageDraw
+import numpy as np
 import io
 import uvicorn
+import base64
 app = FastAPI()
 # Chargement des modèles
 def load_models():
     return {
+        "KnochenAuge": pipeline("object-detection", model="D3STRON/bone-fracture-detr"),
+        "KnochenWächter": pipeline("image-classification", model="Heem2/bone-fracture-detection-using-xray"),
+        "RöntgenMeister": pipeline("image-classification",
+            model="nandodeomkar/autotrain-fracture-detection-using-google-vit-base-patch-16-54382127388")
     }
 models = load_models()
+def translate_label(label):
+    translations = {
+        "fracture": "Knochenbruch",
+        "no fracture": "Kein Knochenbruch",
+        "normal": "Normal",
+        "abnormal": "Auffällig",
+        "F1": "Knochenbruch",
+        "NF": "Kein Knochenbruch"
+    }
+    return translations.get(label.lower(), label)
+def create_heatmap_overlay(image, box, score):
+    overlay = Image.new('RGBA', image.size, (0, 0, 0, 0))
+    draw = ImageDraw.Draw(overlay)
+    x1, y1 = box['xmin'], box['ymin']
+    x2, y2 = box['xmax'], box['ymax']
+    if score > 0.8:
+        fill_color = (255, 0, 0, 100)
+        border_color = (255, 0, 0, 255)
+    elif score > 0.6:
+        fill_color = (255, 165, 0, 100)
+        border_color = (255, 165, 0, 255)
+    else:
+        fill_color = (255, 255, 0, 100)
+        border_color = (255, 255, 0, 255)
+    draw.rectangle([x1, y1, x2, y2], fill=fill_color)
+    draw.rectangle([x1, y1, x2, y2], outline=border_color, width=2)
+    return overlay
+def draw_boxes(image, predictions):
+    result_image = image.copy().convert('RGBA')
+    for pred in predictions:
+        box = pred['box']
+        score = pred['score']
+        overlay = create_heatmap_overlay(image, box, score)
+        result_image = Image.alpha_composite(result_image, overlay)
+        draw = ImageDraw.Draw(result_image)
+        temp = 36.5 + (score * 2.5)
+        label = f"{translate_label(pred['label'])} ({score:.1%} • {temp:.1f}°C)"
+        text_bbox = draw.textbbox((box['xmin'], box['ymin']-20), label)
+        draw.rectangle(text_bbox, fill=(0, 0, 0, 180))
+        draw.text(
+            (box['xmin'], box['ymin']-20),
+            label,
+            fill=(255, 255, 255, 255)
+        )
+    return result_image
+def image_to_base64(image):
+    buffered = io.BytesIO()
+    image.save(buffered, format="PNG")
+    img_str = base64.b64encode(buffered.getvalue()).decode()
+    return f"data:image/png;base64,{img_str}"
+# Page d'accueil avec interface améliorée
 @app.get("/", response_class=HTMLResponse)
 async def main():
     content = """
+    <!DOCTYPE html>
+    <html>
+    <head>
+        <title>Fraktur Detektion</title>
+        <style>
+            body {
+                font-family: -apple-system, BlinkMacSystemFont, 'Segoe UI', Roboto, sans-serif;
+                background: #f0f2f5;
+                margin: 0;
+                padding: 20px;
+                color: #1a1a1a;
+            }
+            .container {
+                max-width: 1200px;
+                margin: 0 auto;
+                background: white;
+                padding: 20px;
+                border-radius: 10px;
+                box-shadow: 0 2px 4px rgba(0,0,0,0.1);
+            }
+            .upload-section {
+                background: #f8f9fa;
+                padding: 20px;
+                border-radius: 8px;
+                margin: 20px 0;
+                text-align: center;
+            }
+            .result-box {
+                background: #f8f9fa;
+                padding: 15px;
+                border-radius: 8px;
+                margin: 10px 0;
+                border: 1px solid #e9ecef;
+            }
+            .button {
+                background: #0066cc;
+                color: white;
+                border: none;
+                padding: 10px 20px;
+                border-radius: 5px;
+                cursor: pointer;
+                transition: all 0.3s ease;
+                font-size: 16px;
+            }
+            .button:hover {
+                background: #0052a3;
+                transform: translateY(-1px);
+            }
+            .results-grid {
+                display: grid;
+                grid-template-columns: 1fr 1fr;
+                gap: 20px;
+                margin-top: 20px;
+            }
+            .confidence-slider {
+                width: 100%;
+                max-width: 300px;
+                margin: 20px auto;
+            }
+            img {
+                max-width: 100%;
+                border-radius: 8px;
+                box-shadow: 0 2px 4px rgba(0,0,0,0.1);
+            }
+            .loading {
+                display: none;
+                text-align: center;
+                padding: 20px;
+                font-weight: bold;
+            }
+            .score-high { color: #0066cc; }
+            .score-medium { color: #ffa500; }
+            .score-low { color: #dc3545; }
+        </style>
+    </head>
+    <body>
+        <div class="container">
+            <h1>📤 Fraktur Detektion</h1>
+            <div class="upload-section">
+                <form action="/analyze" method="post" enctype="multipart/form-data">
+                    <div>
+                        <input type="file" name="file" accept="image/*" required>
+                    </div>
+                    <div class="confidence-slider">
+                        <label for="threshold">Konfidenzschwelle: <span id="thresholdValue">0.60</span></label>
+                        <input type="range" id="threshold" name="threshold"
+                               min="0" max="1" step="0.05" value="0.60"
+                               oninput="updateThreshold(this.value)">
+                    </div>
+                    <button type="submit" class="button">Analysieren</button>
+                </form>
+            </div>
+            <div id="loading" class="loading">
+                Bild wird analysiert... ⏳
+            </div>
+            <script>
+                function updateThreshold(value) {
+                    document.getElementById('thresholdValue').textContent = parseFloat(value).toFixed(2);
+                }
+                document.querySelector('form').onsubmit = function() {
+                    document.getElementById('loading').style.display = 'block';
+                }
+            </script>
+        </div>
     </body>
+    </html>
     """
     return content
+@app.post("/analyze", response_class=HTMLResponse)
+async def analyze_file(file: UploadFile = File(...)):
     try:
         # Lecture de l'image
         contents = await file.read()
         image = Image.open(io.BytesIO(contents))
+        # Analyse avec tous les modèles
+        predictions_watcher = models["KnochenWächter"](image)
+        predictions_master = models["RöntgenMeister"](image)
+        predictions_locator = models["KnochenAuge"](image)
+        # Création de l'image annotée
+        filtered_preds = [p for p in predictions_locator if p['score'] >= 0.6]
+        if filtered_preds:
+            result_image = draw_boxes(image, filtered_preds)
+        else:
+            result_image = image
+        # Conversion des images en base64
+        result_image_b64 = image_to_base64(result_image)
+        # Construction du HTML pour les résultats
+        results_html = """
+        <!DOCTYPE html>
+        <html>
+        <head>
+            <title>Analyse Ergebnisse</title>
+            <style>
+                body {
+                    font-family: -apple-system, BlinkMacSystemFont, 'Segoe UI', Roboto, sans-serif;
+                    background: #f0f2f5;
+                    margin: 0;
+                    padding: 20px;
+                    color: #1a1a1a;
+                }
+                .container {
+                    max-width: 1200px;
+                    margin: 0 auto;
+                    background: white;
+                    padding: 20px;
+                    border-radius: 10px;
+                    box-shadow: 0 2px 4px rgba(0,0,0,0.1);
+                }
+                .results-grid {
+                    display: grid;
+                    grid-template-columns: 1fr 1fr;
+                    gap: 20px;
+                    margin-top: 20px;
+                }
+                .result-box {
+                    background: #f8f9fa;
+                    padding: 15px;
+                    border-radius: 8px;
+                    margin: 10px 0;
+                    border: 1px solid #e9ecef;
+                }
+                .score-high { color: #0066cc; font-weight: bold; }
+                .score-medium { color: #ffa500; font-weight: bold; }
+                .back-button {
+                    display: inline-block;
+                    background: #0066cc;
+                    color: white;
+                    padding: 10px 20px;
+                    border-radius: 5px;
+                    text-decoration: none;
+                    margin-top: 20px;
+                    transition: all 0.3s ease;
+                }
+                .back-button:hover {
+                    background: #0052a3;
+                    transform: translateY(-1px);
+                }
+                img {
+                    max-width: 100%;
+                    border-radius: 8px;
+                    box-shadow: 0 2px 4px rgba(0,0,0,0.1);
+                }
+            </style>
+        </head>
+        <body>
+            <div class="container">
+                <h1>🔍 Analyse Ergebnisse</h1>
+                <div class="results-grid">
+                    <div>
+                        <h2>🤖 KI-Diagnose</h2>
+        """
+        # KnochenWächter results
+        results_html += "<h3>🛡️ KnochenWächter</h3>"
+        for pred in predictions_watcher:
+            confidence_class = "score-high" if pred['score'] > 0.7 else "score-medium"
+            results_html += f"""
+                <div class="result-box">
+                    <span class="{confidence_class}">{pred['score']:.1%}</span> -
+                    {translate_label(pred['label'])}
+                </div>
+            """
+        # RöntgenMeister results
+        results_html += "<h3>🎓 RöntgenMeister</h3>"
+        for pred in predictions_master:
+            confidence_class = "score-high" if pred['score'] > 0.7 else "score-medium"
+            results_html += f"""
+                <div class="result-box">
+                    <span class="{confidence_class}">{pred['score']:.1%}</span> -
+                    {translate_label(pred['label'])}
+                </div>
+            """
+        # Add image and close HTML
+        results_html += f"""
+                    </div>
+                    <div>
+                        <h2>🎯 Fraktur Lokalisation</h2>
+                        <img src="{result_image_b64}" alt="Analyzed image">
+                    </div>
+                </div>
+                <a href="/" class="back-button">← Zurück</a>
+            </div>
         </body>
+        </html>
         """
+        return results_html
     except Exception as e:
         return f"""
         <body style="font-family: Arial; max-width: 800px; margin: 0 auto; padding: 20px;">
             <h1>Fehler</h1>
             <p>Ein Fehler ist aufgetreten: {str(e)}</p>
             <br>
+            <a href="/" class="back-button">← Zurück</a>
         </body>
         """