Metadata Catalogue
Automatic identification of high-mountain shrubs in very-high resolution satellite images using deep learning: Juniperus case study
Accurately mapping vegetation is crucial for environmental monitoring. Traditional methods for identifying shrubs are labor-intensive and impractical for large areas. This workflow uses remote sensing and deep learning to detect Juniperus shrubs from high-resolution RGB satellite images, making shrub identification more efficient and accessible to non-experts in machine learning.<div><br></div><div>Background</div><div>In a dynamic climate, accurately mapping vegetation distribution is essential for environmental monitoring, biodiversity conservation, forestry, and urban planning. One important application of vegetation mapping is the identification of shrub individuals. We term by shrub identification, detection of shrub location and segmentation of shrub morphology.</div><div><br></div><div>Introduction</div><div>Yet, shrub species monitoring is a challenging task. Ecologists used to identify shrubs using classical field surveying methods, however, this process poses a significant challenge since the shrubs are often distributed in large areas that are most of the time inaccessible. Thus, these methods are considered labor-intensive, costly, time-consuming, unsustainable, limited to a small spatial and temporal scale, and their data are often not publicly available. Combining remote sensing and deep learning, however, can play a significant role in tackling these challenges providing a great opportunity to improve plant surveying. First, remote sensing can offer highly detailed spatial resolution granting exceptional flexibility in data acquisition. Then, these data can be afterward processed by deep learning models for automatic identification of shrubs.</div><div><br></div><div>Aims</div><div>The objective of this workflow is to help scientists, non-expert in machine learning, detect Juniperus shrubs from RGB very-high resolution satellite images using deep learning and remote sensing tools.<br></div><div><br></div><div>Scientific Question: can we accurately detect high-mountain Juniperus shrubs from RGB very-high resolution satellite images using deep learning?</div>
Default
- Date ( Publication)
- 2023-12-31T00:00:00
- Status
- Under development / Pre operational
- Keywords
-
Juniperus
- Keywords
-
High-Mountain Shrubs
- Keywords
-
Satellite Images
- Keywords
-
Deep Learning
- Keywords
-
Remote Sensing
- Keywords
-
Vegetation Mapping
- Keywords
-
Shrub Detection
- Keywords
-
Environmental Monitoring
- Keywords
-
Biodiversity Conservation
- Keywords
-
Image Segmentation
- Keywords
-
Machine Learning
- Keywords
-
Data Analysis
- Access constraints
- Copyright
- Other constraints
-
To be confirmed
- Protocol
-
DOI
- Service Name
-
Tiles Generator process
- Service Description
-
The main objective of this process is to crop the input tiff image into small tiff tiles to facilitate the
coming analysis. The source could be an online catalogue or the esearcher’s local machine. The output file are small tiles of the RGB very-high resolution satellite image.
- Service Name
-
Shrub Finder process
- Service Description
-
Apply a deep learning (DL) model to identify shrubs in each tiff tile, then generate for each tile a shapefile. The input are Satellite image tiles and the output is the output of Shrub Finder from a set of shapefiles, mapping each tiff tile, containing the DL model detections.
- Service Name
-
Shrub Descriptor process
- Service Description
-
Analyze each model’s detection from all shapefiles. It aims to compute several metrics describing the color, shape, and texture of each detection, then generate a csv file. The output is is the output of Shrub Descriptor, a file describing all the model detections using color, shape, and texture metrics.
- Service Name
-
Shrub Filter process
- Service Description
-
Filter out the false detections through applying a machine learning classifier that takes as inputs the values of the abovementioned metrics, then generates a csv file with the true shrubs geometries and their descriptive metrics. The output is is the output of Shrub Filter, a file containing the detected shrubs after filtering out the false detections using a machine learning (ML) model applied on Shrub Descriptor output.
- Service Name
-
Converter process
- Service Description
-
It takes the output of the previous process as csv file and convert it to a shp file, a map containing all the detected Juniperus.
- Workflow Helpdesk
Metadata
- File identifier
- 8f0e8983-ac0f-48b9-b5c0-dcea3b9b8269 XML
- Metadata language
- en
- Hierarchy level
- Workflow
- Metadata Schema Version
-
1.0