Update app.py

app.py

@@ -0,0 +1,72 @@
+import io
+from flask import Flask, request, jsonify
+from flask_cors import CORS
+from ultralytics import YOLO
+from PIL import Image
+
+
+app = Flask(__name__)
+CORS(app) 
+
+
+MODEL_PATH = 'best.pt'
+
+print("Mencoba memuat model YOLO...")
+try:
+    model_yolo = YOLO(MODEL_PATH)
+    print(f"Model YOLO berhasil dimuat. Kelas: {model_yolo.names}")
+except Exception as e:
+    print(f"Error saat memuat model YOLO: {e}")
+    model_yolo = None
+
+
+@app.route('/', methods=['GET'])
+def health_check():
+    return jsonify({
+        "status": "ok",
+        "message": "Selamat! Server Flask v2 sedang berjalan!",
+        "version": "2.0" 
+    })
+
+
+@app.route('/predict', methods=['POST']) 
+def predict():
+    if model_yolo is None:
+        return jsonify({'status': 'error', 'message': 'Model tidak tersedia atau gagal dimuat.'}), 500
+
+    if 'image' not in request.files:
+        return jsonify({'status': 'error', 'message': 'File gambar tidak ditemukan dalam request.'}), 400
+
+    file = request.files['image']
+
+    try:
+        img = Image.open(file.stream).convert("RGB")
+        results = model_yolo.predict(source=img, conf=0.25, verbose=False)
+
+        detected_objects_list = []
+        if results and results[0].boxes.shape[0] > 0:
+            print(f"Objek terdeteksi: {len(results[0].boxes)}")
+            for box in results[0].boxes:
+                cls_id = int(box.cls[0])
+                confidence = float(box.conf[0])
+                class_name = model_yolo.names.get(cls_id, f"ID_Kelas:{cls_id}")
+                
+                detected_objects_list.append({
+                    "jenis_sampah": class_name,
+                    "confidence": round(confidence, 2),
+                    "bounding_box (xyxy)": [round(coord, 2) for coord in box.xyxy[0].tolist()]
+                })
+        else:
+            print("Tidak ada objek yang terdeteksi.")
+
+        return jsonify({
+            'status': 'success',
+            'detections': detected_objects_list
+        })
+
+    except Exception as e:
+        print(f"Error saat prediksi: {e}")
+        return jsonify({'status': 'error', 'message': f'Terjadi kesalahan saat pemrosesan: {e}'}), 500
+
+if __name__ == '__main__':
+    app.run(debug=True, port=5000)