Open AccessA Tomato Quality Identification Method Based on Raman Spectroscopy and Convolutional Neural Network
Author(s) -
Yin Wu,
Chenying He,
Qingxiang Wu
Publication year - 2020
Publication title -
journal of physics. conference series
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.21
H-Index - 85
eISSN - 1742-6596
pISSN - 1742-6588
DOI - 10.1088/1742-6596/1438/1/012029
Subject(s) - raman spectroscopy , computer science , convolutional neural network , identification (biology) , artificial intelligence , artificial neural network , layer (electronics) , quality (philosophy) , deep learning , network layer , construct (python library) , pattern recognition (psychology) , materials science , computer network , nanotechnology , optics , philosophy , botany , physics , epistemology , biology
in recent years, more and more technologies have been applied in monitoring growth and production efficiency of plants, i.e. agricultural Internet of Things (IOT) and new information-aware technologies. The architecture of the IOT is divided into four layers, i.e., the sensing layer, network layer, processing layer and application layer[1-4]. Among them, the perception layer is the facial features and the skin of the IOT, which is the basis of the IOT [5]. Raman spectroscopy technology has the advantages of fast, simplicity, accuracy, non-destructive and automatic identification, which has become a powerful analytical verifying method. The method of tomato quality identification that based on the Raman spectroscopy combined with convolutional neural network (CNN)[6]was explored. The Raman spectrum of tomato was collected by Raman sensor to construct a neural network with deep network structure. Through repeatedly learning and training in Raman map, we can determine the map recognition model of high quality tomatoes and use matplotlib to realize the identification simulation.