Determining the Ripeness of Fruit Juices based on Image Processing Technology and Neural Network Classification

Abstract

In this paper, the ripeness of tomato is divided into three levels as reached, ripe and overripe by image processing and neural network classification. The physical properties including color and size of tomato and some of its chemical properties to develop a rating system were used as auto reviews. The automatic, three-stage maturity stage of the proceedings, the proceedings were also considered to handle. Finally, with the help of neural networks designed and trained, the classification was done for the fruits, chemical, mechanical and physical methods were checked and the most effective features for classification fruit were selected. The results showed that the neural network classification methods when used and trained, can be successful in 92%. The best structure of the neural network has been found by the search algorithm PSO.