International Research Journal of Engineering and Technology (IRJET)
e-ISSN: 2395-0056
Volume: 09 Issue: 06 | Jun 2022
p-ISSN: 2395-0072
www.irjet.net
Development of Effective Crop Monitoring and Management System with Weather Reporting Gopinath B1, Naveen R S2, Bala Murugan M3, Neranjan B V4 1Associate
Professor, Department of Electronics and Communication Engineering, Kumaraguru College of Technology[autonomous], Coimbatore, Tamil Nadu, India 2,3,4 Department of Electronics and Communication Engineering, Kumaraguru College of Technology[autonomous], Coimbatore, Tamil Nadu, India ---------------------------------------------------------------------***--------------------------------------------------------------------2. METHODS AND MATERIALS Abstract – Hill station agriculture is a place where sudden climatic changes and crop related issues happen. Since farmers owning these hill-station farmlands are often located in down-hills which makes them more inconvenient to reach the farmland on time during a natural calamity resulting in poor crop management and thus affecting production. This paper focuses on developing a virtual environment integrated with real-time components those serve as an input/output devices and as processing device. This proposed method uses machine learning algorithms to solve the inconveniences faced mainly by the hill station farmers. This model also has a weather prediction algorithm added to the crop recommendation algorithm which is used for the precise calculation of crop recommendation that is for the unskilled new person to cultivation field. Agri rating is an additional feature for better understanding of the land.
2.1 Algorithms A. K – Nearest Neighbor A machine learning algorithm, that is the simplest and supervised learning technique. It is the most efficient algorithm that can process a very large dataset with numerous classes (that are defined to output) and provides the new data point or provide most similar class possible from the various classes of the stored data. As KNN approach is a non-parametric algorithm, it accepts no assumptions regarding data. The dataset that is imported has various classes of crops which are based on the respective parameters of the land and environment. If when a crop is seeded, this system gathers the daily data of the land and keep the farmer advised by the final report. This approach facilitates both crop maintenance and crop prediction inside a single model.
Key Words: Crop Recommendation, Weather Prediction, Agri Rating, KNN algorithm, LSTM algorithm.
1. INTRODUCTION
B. Random Forest
In India, agriculture plays a crucial part in the economy. Of about 330 million hectares of geographical land in India, almost 33 percent of the lands are practicing agriculture. 40+ percent of these agriculture lands are in hill stations. One of the major issues facing by farmers of these hill station farming is the lack of proper maintenance of crops due to their farmlands and homes are separated by a longer distance. This separation between farmlands and homes makes farmers not easier to reach the farmland during a heavy rain or other natural calamities that are constantly changing over time in hill stations. The main motive of this project is to eliminate the communication delay and thus providing an efficient management system.
This is a decision tree algorithm which is non-linear and also non-parametric supervised classification type. It can be used for comparing values between a larger dataset except it can only provide the output from only two classes. As for this project, to predict the weather, the output is set to either “Possibility of Rain” or “No possibility of Rain”, hence Random Forest Algorithm is more effective as it performs well for smaller classes of output. C. Long Short-Term Memory Long Short-Term Memory, a model of RNN (Recurrent Neural Network) series that has the capacity of learning long-term dependencies. RNN is useful for short-term dependencies where as LSTM is more effective when comes to long-term. Whatsoever, LSTM uses similar structure of chains of neural network as RNN, it has its each part of the chain that are multiple layer which can interact with the other. This method is used for Agri-rating where the system can compare the land parameters and provide how good the land is, thru a rating scale.
In this project, we proposed a machine learning model using data mining that can help the farmer to know what happens in the farmland periodically and keep them alert of the changes happening up the hill. K-Nearest Neighbor (KNN), Random Forest and Long Short-Term Memory (LSTM) are the Machine Learning Algorithms used for different purposes of the proposed model. These algorithms are discussed in brief in the upcoming sections.
© 2022, IRJET
|
Impact Factor value: 7.529
|
ISO 9001:2008 Certified Journal
|
Page 892