Crab Quality Detection with Gas Sensors Using a Machine Learning

Crab is a widely recognized and favored seafood product globally. Crab’s delicious taste and high nutritional value, particularly its protein content, make it a desirable food choice. Given the global popularity of seafood, including crabs, maintaining its quality is essential for both economic and consumption purposes. However, seafood products are prone to rapid spoilage due to their high-water content, with spoilage rates varying among different types of seafood. It is crucial for industries to monitor and ensure the quality of their products before they reach the market. Given the high demand for crabs, there is a pressing need for an effective method to assess their quality. This research seeks to establish a method for assessing the freshness and quality of crabs using an electronic nose (e-nose) system, employing machine learning algorithms for classification analyses. Three algorithms will be utilized, along with hyperparameter optimization, to achieve optimal accuracy in evaluating crab quality. These algorithms are K-Nearest Neighbor (K-NN), Support Vector Machine (SVM), and Naïve Bayes. The highest result is achieved by K-NN methods with 98% accuracy percentage. The proposed method of this research has acquired targets that can contribute to advancing seafoods production for industries.