摘要
利用深度学习的方法建立一个自动判别模型,以识别白茶生长的物候期。通过输入白茶的图像,使用卷积神经网络CNN来提取图像特征,在此基础上进行分类从而实现对图像内白茶生长物候期的识别,再融合气象特征对识别效果进行优化,从而得出了准确率较高的识别模型。
By using the deep learning method,we establish an automatic discriminant model to identify the growing phenological phase the of white tea in this study.In this experiment,we input image by the deep learning,extract the image feature by using convolution neural network (CNN),and then on the basis of the classification so as to realize the recognition of image white tea grows phenological period,fusion the meteorological characteristics to optimize the recognition effect,obtained the recognition model with higher accuracy.Through experiments,the model can accurately identify the phenological period of the white tea in the image.
作者
王志毅
王嘉佩
杜爱军
刘丽霞
喻宝龙
王旭
Wang Zhiyi;Wang Japei;Du Aijun;Liu Lixia;Yu Baolong;Wang Xu(Chongqing Meteorological Observatory,Chongqing 401147)
出处
《气象科技进展》
2021年第2期119-120,137,共3页
Advances in Meteorological Science and Technology
关键词
白茶生长物候期
深度学习
自动判别
模型
growth phenological stage of the white tea
deep learning
automatic discriminant model