Seagrass LOST Pipeline

This Repo is a setting up example for semantic Segmentation in LOST. This Repo is based on this Repo (Link)

Instruction

1. Install LOST (Link)
2. Clone this Repository into {LOST WORKSPACE}/my_data/{USER}
3. Clone the branch "LostPipeline" from this Repository: (Link) into the project Folder
4. Download the Dataset (Link)
5. Unzip the Dataset
6. cd dataset
7. mkdir Seagrass_Flat
8. Flat the images with cp images/*/* ./Seagrass_Flat/
9. mkdir gt_flat
10. Flat the images with cp ground-truth/*/* ./gt_flat/
11. Copy the Folder Seagrass_Flat/ into {LOST WORKSPACE}/data/media
12. Copy the Folder gt_flat/ into {LOST WORKSPACE}/my_data/{USER}
13. import the pipeline (Link)
14. Add a Label Tree "Seagrass" with the Labels "Seagrass" and "Background"
15. Start the Pipeline