Determination of Tomato Fruit Stages Using Principal Component Analysis and Fuzzy Logic Algorithm

E-Nose Technology is an excellent non-destructive way to identify different gases and perform diverse working environments. It also provides accurate data and is less expensive. Philippine Statistic Authority data shows that the country's tomato production is still rising. The goal is to develop a portable e-nose device that uses fuzzy logic and principal component analysis to identify tomato ripeness. The researchers used MQ3, MQ4, MQ6, MQ7, and MQ135 sensors for the 60-day data acquisition of unripe tomatoes, where the researchers used principal component analysis. After applying the algorithm, MQ3 and MQ135 show low sensor responses. An Arduino Uno controlled the prototype and was connected to the Raspberry Pi 4 for its portability. The three essential sensors demonstrate a rise with each ripeness stage, while MQ6 and MQ7 show little differences in values. The result of the model's overall accuracy is 88%, while the weighted average precision for each classification is 88.7%, indicating that the method is reasonably accurate.