摘要
作为世界上最大的粮食生产及消费大国,我国粮食在产后的储存安全方面显得尤为重要。针对多种储粮信息的安全状况分析,采用数据融合技术是获得全面、准确结果的有效方式。数据融合通过对粮食多源信息的综合分析,以获取安全评估的一致性解释和描述。深度学习则是提取数据深层特征的有力工具,因此基于深度学习的多源数据融合方法能够充分发挥数据关联与潜在价值,更加稳定的做出准确判断和决策。本文首先根据相关文献综述数据融合的产生与发展;并对传统数据融合方法应用以及现代数据融合研究成果进行分类对比,分析深度学习在数据融合中的优势,进而提出深度学习在粮食信息融合处理中的可行性;最后结合粮情处理现状,提出一种基于深度学习的粮情二级融合框架,采用温度、湿度、水分、虫害四种储粮信息的融合,对其在储粮安全评估中的应用做出研究和展望。
China is the largest producer and consumer of grain in the world,the post-production storage safety of grain in China is of particular importance.Data fusion technology is an effective way to obtain comprehensive and accurate results for the analysis of the safety status of multiple grain storage information.Data fusion is used to obtain a consistent interpretation and description of the safety assessment through the integrated analysis of multiple sources of grain information.Deep learning,on the other hand,is a powerful tool for extracting deep features of data,so the multi-source data fusion method based on deep learning can give full play to data association and potential value,and make accurate judgments and decisions more consistently.In this paper,the emergence and development of data fusion was firstly reviewed based on relevant literature;and the traditional data fusion method applications and modern data fusion research results were classified and compared,the advantages of deep learning in data fusion were analyzed,and then,the feasibility of deep learning in grain information fusion processing was proposed;finally,the current situation of grain processing was combined,and a deep learning-based secondary fusion framework for grain information was proposed,in which uses the four methods of temperature,finally,a deep learning-based fusion framework for grain information is proposed,using the fusion of four types of grain storage information:temperature,humidity,moisture and insect pests,to make a study and outlook on its application in grain storage safety assessment.
作者
祝玉华
王百皓
李智慧
Zhu Yuhua;Wang Baihao;Li Zhihui(Henan University of Technology,School of Information Science and Engineering,Zhengzhou 450001)
出处
《中国粮油学报》
CAS
CSCD
北大核心
2023年第3期1-9,共9页
Journal of the Chinese Cereals and Oils Association
基金
国家重点研发计划项目(2018YFD0401404)。
关键词
粮食多源信息
数据融合
深度学习
粮情处理
二级融合
multi-source food information
data fusion
deep learning
grain handling
secondary fusion
