Smart Microwave Oven with Image Classification and Temperature Recommendation Algorithm

Abstrak

When food is warmed in a microwave oven, the user guesses the estimated time for the heating. This cognitive process of guessing can be incorrect - resulting the final food temperature to be too hot or still cold. In this research, a novel closed-loop microwave oven is designed which automatically suggests the target temperature of a food by learning from previous experiences and the heating stops automatically when the food temperature reaches the target temperature. The proposed microwave captures and classifies the food image, and recommends the target temperature, thus the user does not need to remember the target food temperature each time the same food is warmed. The algorithm gradually learns the type of foods that are used in that household and becomes smarter in the recommendation. The proposed algorithm can recommend target temperature with an accuracy of 86.31% for solid food and 100% for liquid food. A prototype of the proposed microwave is developed using the embedded system and tested.