<![CDATA[Object recognition has become an essential area of computer vision, enabling systems to identify and classify visual information for practical, everyday applications. This study aims to develop an object-recognition system capable of detecting common food ingredients and retrieving suitable recipes to assist users in cooking decisions. The research employs a dataset of 1,800 images representing six vegetable categories and implements the Single Shot Multibox Detector within an application that integrates automated recipe retrieval through a conversational artificial intelligence model. The system is evaluated through functional testing, scenario-based testing, and quantitative performance measurement using established object-recognition metrics. The results show that the model performs most effectively with an eighty–twenty training–testing split, achieving high precision and robust detection across varied lighting conditions, backgrounds, and object distances. The study also demonstrates that the system successfully identifies multiple ingredients within a single image and provides relevant recipe suggestions. These findings indicate that the approach can support domestic decision-making by connecting real-time ingredient recognition with intelligent retrieval of cooking information. The outcomes highlight the practical potential of object recognition for daily-use applications and suggest directions for expanding ingredient categories and improving detection accuracy in future work. Keywords: Ingredient detection; object recognition; recipe retrieval]]>
