Related Experiment Video
Updated: Mar 7, 2026
09:30
Super-resolution Imaging of Neuronal Dense-core Vesicles
Published on: July 2, 2014
10.1K
Estimating babassu palm density using automatic palm tree detection with very high spatial resolution satellite
Alessio Moreira Dos Santos1, Danielle Mitja2, Eric Delaître2
1Universidade Federal Rural da Amazonia (UFRA), CP. 917, Belém, Pará, 66077-530, Brazil; Universidade Federal do Sul e Sudeste do Pará (UNIFESSPA), Folha 31, Quadra 07, Lote Especial, Nova Marabá, 68507-590, Marabá, Brazil.
Journal of Environmental Management
|February 13, 2017
Summary
This study developed an algorithm to automatically detect babassu palm trees using high-resolution satellite imagery in the Brazilian Amazon. The method accurately identified mature trees, aiding in resource assessment for economic and social benefits.
Area of Science:
- Remote Sensing
- Ecology
- Forestry
Background:
- High spatial resolution imagery and object detection algorithms are advancing biodiversity and forest plantation studies.
- The native babassu palm (Attalea speciosa) is crucial to the Brazilian economy and local communities.
Purpose of the Study:
- To analyze the automatic detection of large circular crown (LCC) babassu palm trees using high spatial resolution GeoEye imagery.
- To propose auxiliary methods for estimating babassu palm density based on detection results.
- To validate spatial information using field methods.
Main Methods:
- Utilized the "Compt-palm" algorithm, detecting palm tree shadows in open areas via mathematical morphology.
- Employed high spatial resolution panchromatic GeoEye imagery (0.50 m).
- Validated results using field methods including structural census and georeferencing.
Main Results:
- The algorithm achieved approximately 96% detection for stage 6 babassu individuals and 60% for stage 5.
- Detected individuals in life stages 5 and 6 were evaluated using extraction percentage, branching factor, and quality percentage.
- Principal components analysis indicated structural differences in the studied species compared to others.
Conclusions:
- Automatic detection of LCC babassu palm trees is feasible and effective for monitoring this vital resource.
- The developed methods provide a foundation for large-scale monitoring of babassu palm populations.
- Accurate density estimation supports sustainable management and economic utilization of babassu resources.
