A Deep Learning-Based Orchid Classification Methodology for Automated Packaging in Production Lines

Today, the orchid production chain has the problem of tedious and time-consuming production, and automatic control technology can solve this problem. This paper proposes a deep learning-based classification methodology for the export packaging part of the orchid production line. The proposed orchid classification methodology uses CNNs for transfer learning. The experimental results show that the proposed orchid classification methodology can classify 58 orchid species, and the orchid classification model provides an accuracy rate of 91.4%.