IoT based crop monitoring scheme using smart device with machine learning methodology
IoT based crop monitoring scheme using smart device with machine learning methodology
复制标题
使用智能设备和机器学习方法的基于物联网的农作物监测方案
DOI：
10.1088/1742-6596/2027/1/012019
复制
发表时间：
2021
期刊：
影响因子：
--
通讯作者：
M.Ramkumar Prabhu
中科院分区：
文献类型：
--
作者：
S. Shylaja;S. Fairooz;J. Venkatesh;D. Sunitha;R. P. Rao;M.Ramkumar Prabhu
物联网（IoT）是所有智能应用中最重要的媒介，它以精细的方式为农业产业提供了巨大的支持。在文献中，有很多智能设备可用于监测作物和农田，但都受到一定的限制，如电力问题，成本昂贵等。本文旨在设计一种新的支持机器学习的智能物联网媒介，以适当的方式支持农业领域。本文介绍了一种智能作物监测装置（ICMD），用于对农田上的作物进行24x7全天候监测。这种监测装置提高了农业及相关产品的生产和服务质量。本文将一种名为机器学习的创新技术与智能设备联系起来，但这种方法没有使用经典的学习方案，而是引入了一种名为基于改进学习的领域分析策略（MLFAS）的新方案。这种方法的灵感来自经典的机器学习方案卷积神经网络（CNN），其中提出的智能设备ICMD积累实时农田细节，并将其传递给监控单元进行操作。操作端将数据维护到服务器单元，在服务器单元中，称为MLFAS的机器学习模型获取接收到的现场数据并根据训练样本进行处理。训练样本只是从农业领域收集的数据，收集到的数据保存到服务器端进行处理，提出的MLFAS模型对数据进行操作并创建为进一步测试的模型。新到达的现场数据被视为测试数据，并将该数据交叉验证到训练模型中。从农田获取的数据是温度、湿度和土壤湿度水平，其中这些记录通过与ICMD相关的物联网模块传递到服务器单元。农民可以随时随地轻松地监控服务器中可用的数据。学习模型通过分析从实时测试输入中获得的输入来预测田间作物的状态，并将其报告给相应的农民，以便采取适当的行动。该系统对农业生产具有实用价值，为农民远程监控农田作物提供了良好的支持。通过使用该方案，农民可以利用ICMD智能设备获得的结果做出准确有效的作物管理决策。
Internet of Things (IoT) is the most considerable medium for all smart applications, in which it provides a huge support to agricultural industry in fine manner. In literature, there are lots of smart devices are available for monitoring the crops and agricultural field, but all are strucked under certain limitations such as power problem, cost expensiveness and so on. This paper is intended to design a new machine learning enabled Smart Internet of Things medium to support agricultural field in proper way. In this paper, a Intelligent Crop Monitoring Device (ICMD) is introduced to monitor the crops over the agricultural field in 24x7 manner. This kind of monitoring devices enhances the production and quality-of-service of the agriculture as well as related products. This paper associates an innovative technology to the Smart Device called Machine Learning, but instead of using the classical learning schemes, this approach introduced a new scheme called Modified Learning based Field Analysis Strategy (MLFAS). This approach is inspired from the classical machine learning scheme called Convolutional Neural Network (CNN), in which the proposed Smart Device called ICMD accumulates the real-time agricultural field details and pass it to the monitoring unit for manipulation. The manipulation end maintains the data into the server unit, in which the machine learning model called MLFAS acquires the received field data and process it based on the training samples. The training samples are nothing but data collected from the agriculture field, the collection of received data are maintained into the server end for processing, the proposed MLFAS model manipulates the data and created as a model for further testing. The newly arrived data from the field is considered as a testing data and cross-validate that data into the trained model. The data acquired from the agriculture fields are temperature, humidity and soil moisture level, in which these records are passed to the server unit by using IoT module associated with the ICMD. The data available into the server can easily be monitored by the farmer from anywhere at any time. The learning model predicts the status of the crop in the field by means of analyzing the input acquired from the real-time testing input and report that to the respective farmer for taking an appropriate action. For all this system is useful to the agricultural field and provides good support to farmers to monitor the crops over the agricultural field from the remote place even. By using this proposed scheme, the farmers can make accurate and efficient crop management decisions with the use of results obtained by using the Smart Device called ICMD.