Use of Deep Learning to Determine the Freshness of Egg

Abstract

The freshness of the egg is important for both hatching and human consumption. It is quite difficult to determine the freshness of the egg without damaging it with classical methods. Deep learning is a powerful method used to classify data without processing or with much less processing. In this study, 30 eggs were photographed as experimental material for 29 days and the images obtained were used as data. It is aimed to determine how many days old the eggs are, which are foldered according to the days of the photos obtained. As a result of the study, 91.78% valuation accuracy value was obtained. Obtaining inputs without preprocessing shows that the Deep learning method can be used when a fast decision is required and the machine needs to make its own decision.