7_Coastal Models

These models have been made for the Seg2Map package for application of <=1m nadir and orthomosaic imagery

💾 Coast Train

• Coast Train is a dataset for segmentation of images of coastal environments
• Coast Train
• dataset

Example labeled imagery:

2-class / Orthomosaic + NAIP imagery

• Zenodo release: https://zenodo.org/record/7574784#.Y-FkUxzMLRY
• segmentation zoo model name: orthoCT_2class_7574784
• Buscombe, Daniel. (2023). Doodleverse/Segmentation Zoo/Seg2Map Res-UNet models for CoastTrain water/other segmentation of RGB 768x768 orthomosaic images (v1.0) [Data set]. Zenodo. https://doi.org/10.5281/zenodo.7574784

Classes: 0. water

1. other

Training curve for the best model:

Some example validation outputs:

5-class / NAIP imagery

• Zenodo release: https://zenodo.org/record/7566992#.Y-FkURzMLRY
• segmentation zoo model name: orthoCT_5class_7566992
• Buscombe, Daniel. (2023). Doodleverse/Segmentation Zoo/Seg2Map Res-UNet models for CoastTrain/5-class segmentation of RGB 768x768 NAIP images (v1.0) [Data set]. Zenodo. https://doi.org/10.5281/zenodo.7566992

Classes: 0. water

1. whitewater
2. sediment
3. bare terrain
4. other terrain

Training curve for the best model:

Some example validation outputs:

5-class SegFormer model / NAIP imagery

• Zenodo release: https://zenodo.org/record/7641708#.Y_pdB4DMJkg
• segmentation zoo model name: orthoCT_5class_segformer_7641708
• Buscombe, Daniel. (2023). Doodleverse/Segmentation Zoo/Seg2Map SegFormer models for CoastTrain/5-class segmentation of RGB 768x768 NAIP images (v1.0) [Data set]. Zenodo. https://doi.org/10.5281/zenodo.7641708

Classes: 0. water

1. whitewater
2. sediment
3. bare terrain
4. other terrain

Training curve for the best model:

Some example validation outputs:

8-class / NAIP imagery

• Zenodo release: https://zenodo.org/record/7570583#.Y-FkUhzMLRY
• segmentation zoo model name: orthoCT_8class_7570583
• Buscombe, Daniel. (2023). Doodleverse/Segmentation Zoo/Seg2Map Res-UNet models for CoastTrain/8-class segmentation of RGB 768x768 NAIP images (v1.0) [Data set]. Zenodo. https://doi.org/10.5281/zenodo.7570583

Classes: 0. water

1. whitewater
2. sediment
3. other_bare_natural_terrain
4. marsh_vegetation
5. terrestrial_vegetation
6. agricultural
7. development

Training curve for the best model:

Some example validation outputs:

8-class SegFormer model / NAIP imagery

• Zenodo release: https://doi.org/10.5281/zenodo.7641724
• segmentation zoo model name: orthoCT_8class_segformer_7641724
• Buscombe, Daniel. (2023). Doodleverse/Segmentation Zoo/Seg2Map SegFormer models for CoastTrain/8-class segmentation of RGB 768x768 NAIP images (v1.0) [Data set]. Zenodo. https://doi.org/10.5281/zenodo.7641724

Classes: 0. water

1. whitewater
2. sediment
3. other_bare_natural_terrain
4. marsh_vegetation
5. terrestrial_vegetation
6. agricultural
7. development

Training curve for the best model:

Some example validation outputs:

💾 Chesapeake

• Chesapeake is a dataset for generic/coastal landuse/cover mapping on high-resolution orthomosaic imagery
• Chesapeake Landcover (CCLC) / NAIP
• 7 class dataset (water, tree canopy / forest, low vegetation / field, barren land, impervious (other), impervious (road), no data)
• Robinson C, Hou L, Malkin K, Soobitsky R, Czawlytko J, Dilkina B, Jojic N. Large Scale High-Resolution Land Cover Mapping with Multi-Resolution Data. Proceedings of the 2019 Conference on Computer Vision and Pattern Recognition (CVPR 2019)
• Classes:

0. water
1. tree_canopy_forest
2. low_vegetation_field
3. barren land
4. impervious_other
5. impervious_road
6. no_data

Example labeled imagery

Model for 7 classes, 512x512 imagery

• Zenodo release: https://zenodo.org/record/7576904#.Y-FnkRzMLRY
• segmentation zoo model name: chesapeake_7class_7576904
• Buscombe, Daniel. (2023). Doodleverse/Segmentation Zoo/Seg2Map Res-UNet models for Chesapeake/7-class segmentation of RGB 512x512 high-res. images (v1.0) [Data set]. Zenodo. https://doi.org/10.5281/zenodo.7576904

Training curve for the best model:

Some example validation outputs:

Segformer model for 7 classes, 512x512 imagery

• Zenodo release: https://zenodo.org/record/7677506#.Y_pdBIDMJkg
• segmentation zoo model name: chesapeake_7class_segformer_7677506
• Buscombe, Daniel. (2023). Doodleverse/Segmentation Zoo/Seg2Map Segformer models for Chesapeake/7-class segmentation of RGB 512x512 high-res. images (v1.0) [Data set]. Zenodo. https://doi.org/10.5281/zenodo.7677506

Training curve for the best model:

Some example validation outputs:

💾 Barrier Islands

• BarrierIslands is a dataset for segmentation of images of coastal barrier island environments
• dataset
• paper

Example labeled imagery

Model for 768x512 imagery

• Zenodo release: forhcoming
• model citation: forthcoming
• segmentation zoo model name:

Training curve for the best model:

Some example validation outputs: