Rapid detection of sea bass quality level with machine learning and electronic nose

Abstract

SummaryIn the present study, the odour changes of sea bass was measured and recorded with Arduino‐based six gas (MQ) sensors during 7 days of storage. In addition, the changes in biogenic amines (BAs) and total viable count (TVC) occurring in fish meat during storage were determined daily and compared with electronic nose data. For image processing, 500 fish data were taught to Teachable Machine (TM) daily, a web‐based machine learning platform. In the study, it was determined that if the level of putrescine (PUT) exceeds 1 mg kg−1 and spermidine (SPMD) exceeds the level of 2 mg kg−1, the MQ4, MQ5, MQ8 and MQ9 sensors exceed the value of 500. In the study, it was determined that combined machine learning and electronic nose could be used as a rapid quality determination tool in a perishable food such as fish.