IoT Enabled Crop Detection System using Soil Analysis

Kumar, Khushi; Sharma, Sachin; Pandey, Piyush; Goyal, Himanshu Rai

doi:10.1109/icces54183.2022.9835959

Cited by 8 publications

(1 citation statement)

References 0 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…With forecast effectiveness, the Direct Discriminant investigation calculation yields the highest productivity of 91.25 percent. Khushi et al [24] expressed that Dirt observing empowered by the Web of Things permits ranchers to anticipate which harvests will develop in view of soil factors. A yield forecasting tool for a particular soil type is what we are after.…”

Section: Literature Surveymentioning

confidence: 99%

Multivariate Soil Monitoring and Crop Prediction Model Based on AAD-ARIMA and LCV-OXGBOOST Techniques

Shenbagavadivu,

2023

Preprint

View full text Add to dashboard Cite

Farmers must adjust to the rising environment while producing more food with better nutrition. To boost crop production and growth, the farm worker must be knowledgeable of the soil conditions, which will aid in selecting the best crop to sow in the given conditions. By continuously monitoring the land, IoT-based smart farming enhances the agricultural industry as a whole. It maintains numerous variables, including sediment, temperature, and moisture. According to them, the project intends to assist farmers in making wise decisions by forecasting the crops and simultaneously monitoring the soil. Based on AAD-ARIMA and LCV-OXGBOOST, a multivariate soil monitoring and crop prediction model has been created. First, the data has been normalized, which helps to determine the likelihood of inaccuracy for the data. Missing values are handled based on the results of the preprocessing, which includes categorization the missing value using SD-CCC. After that, +-shift-ROS is used to manage the data's unequal distribution before LE-PT scaling. After that, this research has created an MLE-CFO strategy that offers the correlation between the materials by thinking about the causality and maintains an ideal working length as well as correctness in order to acquire data knowledge. Following that, the characteristics are divided using MIC-DBSCAN for crop prediction and soil monitoring. The selected characteristic was then tested against by the LCV-OXGBOOST for crop prediction and the AAD-ARIMA for monitoring. The suggested method works more effectively and dependably while reducing false alarm rates (FARs) and inaccuracy rates based on the dataset collected from Soil of Chengalpattu. Additionally, the work controls the stochastic and unpredictable behavior of uncertain data and yields a suitable outcome. When compared to the current top-notch system, empirical testing shows that the work delivers superior accuracy, reaction rate, and is significantly more expandable and safe.

show abstract

Section: Literature Surveymentioning

confidence: 99%