Classification of chicken products by deep learning

Information regarding the storage, monitoring and consumption of food products is of critical importance in human life. Chicken products, which people consume most in daily life, are considered the most important food product among these products. These products must be stored and delivered in good conditions until they reach the end user. Although there are automatic systems, these systems are still controlled by humans. While the internet of things makes this easier, the control and monitoring of chicken products by real-time systems is very important. Deep learning is frequently used in tracking systems where images are used. In this study, six types of chicken products sold in markets and frequently consumed were classified using the deep learning method. Each product was stored under appropriate storage conditions and sufficient images were recorded for a week. These recorded images were classified in-class on a daily basis with ResNet-50, one of the deep learning models. It is also classified on a class-by-class basis. Chicken products were classified with a deep learning algorithm with over 94% success in inter-class and intra-class classification.