Cooking Method Detection Based on Electronic Nose in Smart Kitchen

Dietary health study in home environment has been a long-standing research direction, which is mainly focused on dietary habits and nutritional balance, but relatively little research has been conducted on cooking methods. In this paper, an electronic nose-based cooking method was developed in a smart kitchen scenario. The gas sensor array constructed by MOS gas sensors was used to continuously collect the gas generated during the cooking process of 84 dishes, and the classification models based on decision tree and random forest are compared, and the results showed that the average classification accuracy of the latter is 95% for three cooking methods: boiling, frying and deep-frying. In addition, 6 users participated in on-site experience and interviews to evaluate the usability of the method. The potential of the method to help users record their diets and manage their health was demonstrated, and its interactive applications were explored to summarize its future design patterns and interactions in smart kitchens, providing design suggestions and directions for subsequent research.