oklu Girdi Verilerinin CNN-LSTM Modellerinde Entegrasyonu: Akilli Firinlarda Yapay Zek Destekli Pi sirme Sonlandirma

In this study, an AI-powered smart oven system is proposed, which can automatically terminate the cooking process based on user preferences (rare, regular, or well-done). A camera-integrated oven prototype was developed for five different food types (fresh pizza, frozen pizza, tray pastry, mini pastry, and salmon), and a comprehensive dataset was created using images captured throughout the cooking process. Visual data were subjected to feature extraction using image processing techniques, and the resulting features were combined with numerical inputs such as temperature and cooking time to form the input of an LSTM-based deep learning model. The model was evaluated using both the collected dataset and real-world user scenarios, and was shown to perform with high accuracy. The developed system adapts to individual cooking preferences, prevents overcooking, and contributes to energy efficiency and reduction of food waste.