Design of a Moisture Content Detection System for Yinghong No. 9 Tea Leaves Based on Machine Vision

The moisture content of Yinghong No. 9 tea leaves is an important indicator for their processing. traditional method used to detect the not suitable large-scale production. To improve efficiency processing, a detection system based on machine vision was developed, and relationship between fresh researched. Firstly, nine color features five texture images were extracted, two different databases constructed linear discriminant analysis (LDA) principal component (PCA). Secondly, models prediction built using backpropagation (BP) neural network, which then optimized by particle swarm optimization (PSO) genetic algorithm (GA), respectively. After, preprocessing methods algorithms cross-combined optimize prediction. Finally, above filtered segmental It verified experiments that coefficient determination (R2) combined model PCA-GA-BP PCA-PSO-BP 94.1073%, RMSE 1.1490%, MAE 0.9982%. results this paper can help in instantaneous during improving production tea.