Fish Freshness Detection Through Artificial Intelligence Approaches: A Comprehensive Study

Creative Commons License

Kılıçaslan S., Çiçekliyurt M. M., Kılıçaslan S.

Turkish Journal of Agriculture - Food Science and Technology (TURJAF) , vol.12, no.2, pp.290-295, 2024 (Peer-Reviewed Journal)

Abstract

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

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