Neural network based agriculture activity detection using mobile accelerometer sensors

The information about the agricultural activities being performed on the farm is useful for providing agriculture advisory to the farmers. In this paper, we present the Neural Network based approach for the classification of agriculture activities like harvesting, bed-making, transplantation, walking and standstill from the acceleration data obtained from mobile phone carried by the farmer. The performance of the neural network based classifier has been compared with the Linear Discriminant Analysis, k-Nearest Neighbors and Naive Bayes classifiers. The trained neural network based classifiers are found be more accurate as compared to the other classifiers.