Deep learning-based prediction for stand age and land utilization of rubber plantation
share
Smart agriculture has been attracting greater attention from the agricultural research community to enhance current practices through the incorporation of data engineering techniques. This chapter presents an approach to classify the stand age and land utilization of rubber plantation using deep learning techniques in conjunction with remote sensing imagery of profiles of vegetation indices. We apply fully convolutional network (FCN) and U-net-based approaches to classify the satellite image datasets collected over several periods to construct the crop monitoring and land-use predictive model. The FCN with VGG and U-net as deep learning models provide an accuracy of 87.25% and 94.13%, respectively, when predicting the stand age of rubber plantations. The designed deep learning-based monitoring will help to improve the accuracy of input and output data such as fertilizer use, yield, and production levels and help fieldwork management technologies.
READ FULL TEXT
