Automated Sorting of Rotten or Defective Fruits and Vegetables Using Convolutional Neural Network

Abstract

Listen

Translate article

Modern peoples are very concerned about food safety. It is known to all that the rotten or defective fruits and vegetables are harmful to health. The rotten and defective fruits and vegetables may degrade the quality of food in the food processing industries. It may damage the fresh fruits and vegetables, which are in the surface contact with the rotten fruits or vegetables in the inventory of food processing industries as well as in the store of supermarkets. The rotten or defective fruits and vegetables should be detected and sorted as early as possible. The automated detection and sorting with the help of image analysis and deep learning can be very effective in industries. This paper proposes a convolutional neural network architecture for sorting the rotten or defective fruits and vegetables from a lot of fruits and vegetables. The proposed technique will classify the fresh and non-fresh (rotten or defective) fruits and vegetables using the trained convolutional neural network. The proposed convolutional neural network model has been tested on two datasets. The classification accuracy on the first and second datasets is 98.53% and 98.46%, respectively. The F1 scores are 99% and 99.18% on dataset 1 and dataset 2, respectively.