Research Article
HE-DFNETS: A Novel Hybrid Deep Learning Architecture for the Prediction of Potential Fishing Zone Areas in Indian Ocean Using Remote Sensing Images
Table 2
Parameters of CNN (Tier-1) used for segmentation and feature extraction OF SST maps and SSC maps.
| Sl. no. | Layers used | Output size | Parameters |
| 1 | Convolution layer (CL-1) | 256 × 256 × 32 | 564 | 2 | Pooling layer (PL-1) | 128 × 128 × 32 | 2 | 3 | Convolution layer (CL-2) | 128 × 128 × 64 | 2784 | 4 | Pooling layer (PL-2) | 64 × 64 × 128 | 0 | 5 | Convolution layer (CL-3) | 64 × 64 × 128 | 9673 | 6 | Pooling layer (PL-3) | 32 × 32 × 256 | 0 | 7 | Convolution layer (CL-4) | 32 × 32 × 128 | 167653 | 8 | Pooling layer (PL-5) | 16 × 16 × 256 | 0 | 9 | Convolution layer (CL-4) | 16 × 16 × 256 | 208960 | 10 | Pooling layer (PL-5) | 4 × 4 × 512 | 0 |
|
|