基于深度学习的经典红玫瑰图像分类研究

Research on image classification of classic red roses based on deep learning

经典红玫瑰的分级一直由人工分拣完成,分拣的效率较低。近年来,机器视觉以及深度学习越来越多地被应用于农产品的分拣及工业制造领域的分类与缺陷检测。通过引入深度学习与机器视觉,开展经典红玫瑰分类研究。深度学习网络模型能够自动提取图像中的特征,大幅度减少工作量。通过实验得出,分类准确率为87.22%,基本完成既定的实验目标。

The grading of classic red roses has always been done by labours,and the efficiency of sorting is low.In recent years,machine vision and deep learning have been increasingly used in classification and defect detection in the fields of agricultural products and industrial manufacturing.This paper introduces deep learning and machine vision to carry out the classic red rose grading research.Deep learning network models can automatically extract features in images,which can significantly red uce workload.The experiment show that the classification accuracy of the experiment is 87.22%,which finally completes the goal.