Efficient tree mapping through deep distance transform (DDT) learning
Trees provide essential ecosystem services in urban areas, rural landscapes and forests. Individual tree information can inform forest and risk modelling, health studies and decision-making in public and non-governmental sectors. The increase in available remote sensing data and advances in automate...
Saved in:
| Main Authors: | , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
Elsevier
2025-08-01
|
| Series: | ISPRS Open Journal of Photogrammetry and Remote Sensing |
| Subjects: | |
| Online Access: | http://www.sciencedirect.com/science/article/pii/S2667393225000146 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1850116166775734272 |
|---|---|
| author | Jan Schindler Ziyi Sun Bing Xue Mengjie Zhang |
| author_facet | Jan Schindler Ziyi Sun Bing Xue Mengjie Zhang |
| author_sort | Jan Schindler |
| collection | DOAJ |
| description | Trees provide essential ecosystem services in urban areas, rural landscapes and forests. Individual tree information can inform forest and risk modelling, health studies and decision-making in public and non-governmental sectors. The increase in available remote sensing data and advances in automated object detection makes it feasible to map trees over large areas in unprecedented detail. Deep learning-based instance segmentation methods have thereby become the state-of-the-art in tree crown delineations tasks from aerial ortho-photography. Many of these methods are based on one- and two-stage detector frameworks such as Mask-RCNN and YOLO, which were developed focussing on speed and accuracy against common benchmark datasets. Another class of object detectors is based on encoder-decoder networks such as UNet which offer easy integration into existing workflows and high accuracy even in complex forest scenes in regional and national tree studies. While previous methods had to combine multi-model and multi-task outputs to create decision surfaces, we developed an efficient UNet-based modelling approach which focusses solely on learning the distance transforms of tree objects as cost surface for watershed segmentation. Our algorithm achieves superior instance segmentation across native forest, rural and urban environments in Aotearoa New Zealand, with an overall F1 score of 0.53 — 0.18 for small, 0.45 for medium and 0.67 for large crowns — surpassing previous approaches while decreasing modelling complexity, enabling fast and large-scale tree mapping. |
| format | Article |
| id | doaj-art-7160b44d36744113afacad53d6591c91 |
| institution | OA Journals |
| issn | 2667-3932 |
| language | English |
| publishDate | 2025-08-01 |
| publisher | Elsevier |
| record_format | Article |
| series | ISPRS Open Journal of Photogrammetry and Remote Sensing |
| spelling | doaj-art-7160b44d36744113afacad53d6591c912025-08-20T02:36:23ZengElsevierISPRS Open Journal of Photogrammetry and Remote Sensing2667-39322025-08-011710009510.1016/j.ophoto.2025.100095Efficient tree mapping through deep distance transform (DDT) learningJan Schindler0Ziyi Sun1Bing Xue2Mengjie Zhang3Manaaki Whenua – Landcare Research, Informatics, Wellington, 6011, New Zealand; Corresponding author.Centre for Data Science and Artificial Intelligence & School of Engineering and Computer Science, Victoria University of Wellington, Wellington, 6140, New ZealandCentre for Data Science and Artificial Intelligence & School of Engineering and Computer Science, Victoria University of Wellington, Wellington, 6140, New ZealandCentre for Data Science and Artificial Intelligence & School of Engineering and Computer Science, Victoria University of Wellington, Wellington, 6140, New ZealandTrees provide essential ecosystem services in urban areas, rural landscapes and forests. Individual tree information can inform forest and risk modelling, health studies and decision-making in public and non-governmental sectors. The increase in available remote sensing data and advances in automated object detection makes it feasible to map trees over large areas in unprecedented detail. Deep learning-based instance segmentation methods have thereby become the state-of-the-art in tree crown delineations tasks from aerial ortho-photography. Many of these methods are based on one- and two-stage detector frameworks such as Mask-RCNN and YOLO, which were developed focussing on speed and accuracy against common benchmark datasets. Another class of object detectors is based on encoder-decoder networks such as UNet which offer easy integration into existing workflows and high accuracy even in complex forest scenes in regional and national tree studies. While previous methods had to combine multi-model and multi-task outputs to create decision surfaces, we developed an efficient UNet-based modelling approach which focusses solely on learning the distance transforms of tree objects as cost surface for watershed segmentation. Our algorithm achieves superior instance segmentation across native forest, rural and urban environments in Aotearoa New Zealand, with an overall F1 score of 0.53 — 0.18 for small, 0.45 for medium and 0.67 for large crowns — surpassing previous approaches while decreasing modelling complexity, enabling fast and large-scale tree mapping.http://www.sciencedirect.com/science/article/pii/S2667393225000146Remote sensingImage segmentationTree crown delineationDeep learningObject detection |
| spellingShingle | Jan Schindler Ziyi Sun Bing Xue Mengjie Zhang Efficient tree mapping through deep distance transform (DDT) learning ISPRS Open Journal of Photogrammetry and Remote Sensing Remote sensing Image segmentation Tree crown delineation Deep learning Object detection |
| title | Efficient tree mapping through deep distance transform (DDT) learning |
| title_full | Efficient tree mapping through deep distance transform (DDT) learning |
| title_fullStr | Efficient tree mapping through deep distance transform (DDT) learning |
| title_full_unstemmed | Efficient tree mapping through deep distance transform (DDT) learning |
| title_short | Efficient tree mapping through deep distance transform (DDT) learning |
| title_sort | efficient tree mapping through deep distance transform ddt learning |
| topic | Remote sensing Image segmentation Tree crown delineation Deep learning Object detection |
| url | http://www.sciencedirect.com/science/article/pii/S2667393225000146 |
| work_keys_str_mv | AT janschindler efficienttreemappingthroughdeepdistancetransformddtlearning AT ziyisun efficienttreemappingthroughdeepdistancetransformddtlearning AT bingxue efficienttreemappingthroughdeepdistancetransformddtlearning AT mengjiezhang efficienttreemappingthroughdeepdistancetransformddtlearning |