Classification of tobacco using remote sensing and deep learning techniques.

Tobacco is an important crop in many countries, and its management could be improved by accurate yield predictions. Traditional yield estimation methods like human‐based surveys are inaccurate, time consuming, and expensive. In this work, we consider the problem of tobacco identification and classification from satellite imagery and propose a Conv1D and long short‐term memory (LSTM) based deep learning model. We compare the performance of our proposed Conv1D and LSTM deep learning model with benchmark machine learning models, namely support vector machine, random forest, and LSTM. Our model had an accuracy of 98.4%. The use of accurate models can improve the decision process. Core Ideas: A deep neural network can be used to convert satellite imagery into tobacco yield predictions.A mobile application can be used to collect ground truth data.Combining convolutional and LSTM models can improve accuracy. [ABSTRACT FROM AUTHOR]