Plastic Egg Detection Using Machine Learning Approach

Abstract

In Bangladesh egg is almost like a daily meal for many peoples. And egg has so much cholesterol and vitamin in it. As a result, the demand of egg is so much higher in our country. So, there are many types of egg supplies seen in the past. But recently, it is seen that there are the supplies of plastic eggs. Which is pretty tough to recognize by physically whether it is genuine or fake. Nowadays, there are so many report in the newspaper, journal about trading plastic eggs in the market and peoples are unconsciously buy them as well. Those eggs are so much harmful for human health. Therefore, we have thought about to do something for this problem. We have implemented a machine learning approach to identify eggs in the market. For that, we have collected samples of real and plastic eggs and made large datasets by capturing and collecting the images. Using those datasets, we have created some Convolutional Neural Network (CNN) models for check their accuracy in identification of an egg that is real or plastic. Among the models, InceptionV3 was the most accurate which is 99.7%.