| Algorithm 1 Spatial Augmentation Procedure for Data Up-Scaling | |
| 1 | function Spatial Augmentation (S) |
| Input: Number of sensor elements S in the sensing unit OR Length of raw data vectors | |
| Output: The spatially augmented data vectors compatible with the input to the 2D-CNN | |
| 2 | BEGIN |
| 3 | IF S is a perfect square number: |
| 4 | Represent the data vector in the squared array of size: [√(S) × √(S)] |
| 5 | Following stage 2 as discussed, data vector spatially augmented to the size: [(3 × √(S)) × (3 × √(S))] |
| 6 | ELSE |
| 7 | Find the number of required virtual sensors to make the total number of sensor elements equal to the nearest perfect square number, say α |
| 8 | Following stage 1 virtual sensors were included to make the total number of sensor elements in the sensing unit, [(S + α)] |
| 9 | Represent the obtained data vectors in the squared array of size: [√(S + α) × √(S + α)] |
| 10 | Following stage 2 as discussed, data vector spatially augmented to the size [(3×√(S + α)) × (3×√(S + α))] |
| 11 | STOP |