Upload 3 files

app.py

@@ -0,0 +1,64 @@
+import gradio as gr
+from ultralytics import YOLO
+import json
+
+model = YOLO('best-seg.pt')
+
+CLASS_NAMES = {0: 'Capped Cell', 1: 'Failed Cell', 2: 'Matured Cell', 3: 'Open Cell', 4: 'Semi-Matured Cell'}
+MATURITY_MAP = {
+    'Open Cell': {'days': 10, 'percentage': 10},
+    'Capped Cell': {'days': 7, 'percentage': 40},
+    'Semi-Matured Cell': {'days': 5, 'percentage': 70},
+    'Matured Cell': {'days': 2, 'percentage': 95},
+    'Failed Cell': {'days': 0, 'percentage': 0}
+}
+
+def analyze(image):
+    results = model(image)
+    detections = []
+    distribution = {'open': 0, 'capped': 0, 'mature': 0, 'semiMature': 0, 'failed': 0}
+    
+    for r in results:
+        if r.boxes is not None:
+            for i, box in enumerate(r.boxes):
+                cls = int(box.cls[0])
+                conf = float(box.conf[0])
+                x1, y1, x2, y2 = box.xyxy[0].tolist()
+                
+                class_name = CLASS_NAMES.get(cls, 'Unknown')
+                maturity_info = MATURITY_MAP.get(class_name, {'days': 0, 'percentage': 0})
+                
+                detections.append({
+                    'id': i + 1,
+                    'type': class_name,
+                    'confidence': round(conf * 100),
+                    'bbox': [int(x1), int(y1), int(x2-x1), int(y2-y1)],
+                    'percentage': maturity_info['percentage'],
+                    'estimatedHatchingDays': maturity_info['days']
+                })
+                
+                if 'Semi-Matured' in class_name:
+                    distribution['semiMature'] += 1
+                elif 'Matured' in class_name:
+                    distribution['mature'] += 1
+                elif 'Capped' in class_name:
+                    distribution['capped'] += 1
+                elif 'Open' in class_name:
+                    distribution['open'] += 1
+                elif 'Failed' in class_name:
+                    distribution['failed'] += 1
+    
+    return json.dumps({
+        'totalQueenCells': len(detections),
+        'cells': detections,
+        'maturityDistribution': distribution
+    }, indent=2)
+
+demo = gr.Interface(
+    fn=analyze,
+    inputs=gr.Image(type="numpy"),
+    outputs=gr.JSON(),
+    title="iBrood Queen Cell Analyzer"
+)
+
+demo.launch()

best-seg.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:dd29def2ee1f0db481962a8080186f9ed8ad77acbfba15281a38f0532683ff99
+size 45162038

requirements.txt

@@ -0,0 +1,6 @@
+gradio
+ultralytics
+opencv-python-headless
+pillow
+torch
+torchvision