Classification of Local Fish Meat Types in the Philippines Using Electronic Nose

The study develops an AI-based Electronic Nose (E-nose) system to classify four common Philippine fish species—Bangus (Milkfish), Galunggong, Lapu-Lapu (Grouper), and Tilapia—based on their volatile organic compound (VOC) signatures. The system utilizes an array of MQ gas sensors for data collection, processed by a Raspberry Pi 5 microcontroller. Principal Component Analysis (PCA) is applied for dimensionality reduction, and a Support Vector Machine (SVM) is used for classification. A dataset of 300 fish meat samples (75 per species) was split 80–20 for training and testing. Performance evaluation through a confusion matrix showed a high classification accuracy of 98.66%. These results highlight the potential of E-nose technology for fish meat authentication, ensuring food quality and integrity in the Philippines, where seafood is a vital part of the diet and economy. The system's accuracy demonstrates the effectiveness of AI-driven odor classification for seafood products. Future research can expand this classification system by including more fish species and investigating correlations between fish type and freshness decay, further enhancing food safety and quality control measures in the seafood industry.