Deep learning based individual tree crown delineation from panchromatic aerial imagery
Accurate delineation of individual tree crowns (ITC) enables a better understanding of tree-level growth dynamics and evaluating tree vitality. In recent year, researches have introduced deep learning techniques in this field. However, the precise segmentation relies on high quality annotated datase...
Saved in:
| Main Authors: | , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
Copernicus Publications
2025-07-01
|
| Series: | ISPRS Annals of the Photogrammetry, Remote Sensing and Spatial Information Sciences |
| Online Access: | https://isprs-annals.copernicus.org/articles/X-G-2025/885/2025/isprs-annals-X-G-2025-885-2025.pdf |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| Summary: | Accurate delineation of individual tree crowns (ITC) enables a better understanding of tree-level growth dynamics and evaluating tree vitality. In recent year, researches have introduced deep learning techniques in this field. However, the precise segmentation relies on high quality annotated dataset and test images with limited domain gaps between the training data. Under the framework of the Helmholtz project, panchromatic airborne images are captured over a mixed European forest. In this research, we adopt a UAV benchmark dataset as training data. To close the domain gaps, a deep learning based colorization step is added, for which two deep learning frameworks are compared to achieve an improved ITC delineation result in a dense forest area. |
|---|---|
| ISSN: | 2194-9042 2194-9050 |