Name
Abstract | ||
|---|---|---|
With the advancements in precision farming, crop sensing is gaining importance for timely crop health management. Leaf water content (LWC) is key component to determine vegetation health and nourishment. Timely estimation of LWC could save us from hazardous damage by pre-planning: drought stress on plants, irrigation and prediction of woodland fire. The retrieval of LWC from visible to shortwave infrared (VSWIR: 0.39 to 2.5 μm) mid- and thermal-infrared (MIR and TIR: 2.50 to 14.0 μm) windows of electromagnetic spectrum has been investigated using different statistical algorithms. Deep learning is modernizing the fast growing field of machine learning and image processing. The convolutional neural network (CNN) is ultramodern technique of deep learning that learns and extracts features directly from data. This research is focused on the extraction of different features of different plant species by using CNN for Regression. The modeled CNN architecture automatically detects prominent features to estimate LWC in plant species from its reflectance spectra, recorded for varying amount of LWC. Previous methods applied on same dataset yielded accuracy of 93% and Root Mean Square Error (RMSE) of 7.1, however, CNN resulted in better and swift results with an accuracy of 98.4% and RMSE of 4.183. This study helps in identifying the important spectral regions for quantifying water stresses in vegetation. The outcomes of this study can enable the future space missions to foresee water content of different plant species on the basis of their spectral signatures for illustrating vegetation stresses. |
Year | DOI | Venue |
|---|---|---|
2019 | 10.1109/INTELLECT47034.2019.8954985 | 2019 Second International Conference on Latest trends in Electrical Engineering and Computing Technologies (INTELLECT) |
Keywords | DocType | ISBN |
Deep learning,convolutional neural network,leaf water content,regression,hyperspectral imaging,remote sensing | Conference | 978-1-7281-2436-0 |
Citations | PageRank | References |
0 | 0.34 | 6 |
Authors | ||
4 | ||
Name | Order | Citations | PageRank |
|---|---|---|---|
| Rida Nasir | 1 | 0 | 0.34 |
| Muhammad Jaleed Khan | 2 | 0 | 1.35 |
| Muhammad Arshad | 3 | 0 | 0.34 |
| Khurram Khurshid | 4 | 129 | 15.94 |