Fish Freshness Detection Through Artificial Intelligence Approaches: A Comprehensive Study

Creative Commons License

Kılıçarslan S., ÇİÇEKLİYURT M. M., KILIÇARSLAN S.

TURKISH JOURNAL OF AGRICULTURE: FOOD SCIENCE AND TECHNOLOGY, cilt.12, sa.2, ss.290-295, 2024 (Hakemli Dergi)

Özet

Fishisregardedasanimportantproteinsourceinhumannutritionduetoitshighconcentrationofomega-3fattyacidsIntraditionalglobalcuisine,fishholdsaprominentposition,withseafoodrestaurants,fishmarkets,andeateriesservingaspopularvenuesforfishconsumption.However,itisimperativetopreservefishfreshnessasimproperstoragecanleadtorapidspoilage,posingrisksofpotentialfoodborneillnesses.Toaddressthisconcern,artificialintelligencetechniqueshavebeenutilizedtoevaluatefishfreshness,introducingadeeplearningandmachinelearningapproach.Leveragingadatasetof4476fishimages,thisstudyconductedfeatureextractionusingthreetransferlearningmodels(MobileNetV2,Xception,VGG16)andappliedfourmachinelearningalgorithms(SVM,LR,ANN,RF)forclassification.ThesynergyofXceptionandMobileNetV2withSVMandLRalgorithmsachieveda100%successrate,highlightingtheeffectivenessofmachinelearninginpreventingfoodborneillnessandpreservingthetasteandqualityoffishproducts,especiallyinmassproductionfacilities

Fishisregardedasanimportantproteinsourceinhumannutritionduetoitshighconcentrationofomega-3fattyacidsIntraditionalglobalcuisine,fishholdsaprominentposition,withseafoodrestaurants,fishmarkets,andeateriesservingaspopularvenuesforfishconsumption.However,itisimperativetopreservefishfreshnessasimproperstoragecanleadtorapidspoilage,posingrisksofpotentialfoodborneillnesses.Toaddressthisconcern,artificialintelligencetechniqueshavebeenutilizedtoevaluatefishfreshness,introducingadeeplearningandmachinelearningapproach.Leveragingadatasetof4476fishimages,thisstudyconductedfeatureextractionusingthreetransferlearningmodels(MobileNetV2,Xception,VGG16)andappliedfourmachinelearningalgorithms(SVM,LR,ANN,RF)forclassification.ThesynergyofXceptionandMobileNetV2withSVMandLRalgorithmsachieveda100%successrate,highlightingtheeffectivenessofmachinelearninginpreventingfoodborneillnessandpreservingthetasteandqualityoffishproducts,especiallyinmassproductionfacilitiesFishisregardedasanimportantproteinsourceinhumannutritionduetoitshighconcentrationofomega-3fattyacidsIntraditionalglobalcuisine,fishholdsaprominentposition,withseafoodrestaurants,fishmarkets,andeateriesservingaspopularvenuesforfishconsumption.However,itisimperativetopreservefishfreshnessasimproperstoragecanleadtorapidspoilage,posingrisksofpotentialfoodborneillnesses.Toaddressthisconcern,artificialintelligencetechniqueshavebeenutilizedtoevaluatefishfreshness,introducingadeeplearningandmachinelearningapproach.Leveragingadatasetof4476fishimages,thisstudyconductedfeatureextractionusingthreetransferlearningmodels(MobileNetV2,Xception,VGG16)andappliedfourmachinelearningalgorithms(SVM,LR,ANN,RF)forclassification.ThesynergyofXceptionandMobileNetV2withSVMandLRalgorithmsachieveda100%successrate,highlightingtheeffectivenessofmachinelearninginpreventingfoodborneillnessandpreservingthetasteandqualityoffishproducts,especiallyinmassproductionfacilities