Power Line Inspection
for Vegetation Management
GeoXpert provides aerial survey service by drone, then processes the data to generate high definition point clouds and 3D models.
Reconstruct the power pole and insulator using ArcGIS and ArcScene tool. Create an AI model to identify the tree and tree group then formulate the relationship between the vegetation and the power asset.
GeoXpert also provides a cloudbase GIS platform for clients to review the aerial survey result and interact with the data.
In this project, The tree counting algorithm is based on YOLO algorithm.
Slope Inspection Project
Provide training on drone inspection techniques for slope inspection, provide point clouds and 3D model processing service for slope inspection.
Provide a GIS platform to review the slope by drone image and cut section report.
Building progress survey
In this project, GeoXpert platform is used to track construction progress and identify discrepancies between architecture/structural models and the as-built construction.
A 3D model of the actual building was built in collaboration with Insight Robotics as well as internal indoor mapping in collaboration with Mapxus. GeoXpert is both the project coordinator and the platform to store and organize project deliverables.
In this project, GDAL was integrated with Cesium to convert and display the BIM model.
Aerial survey & tree counting
This project, led in partnership with Insight Robotics, consisted in surveying a tree plantation in China and conduct some analytics including:
- Tree counting;
- Health assessment;
- Height measurement; and
- Species identification.
Deep learning algorithms were used for that purpose.
Both tensorflow and YOLO algorithm have been applied in this project.
Kevin Chan, as head of Insight Robotics, has led many projects in plantation surveying in South East Asia, including tree counting and health assessment.
With more recent projects, Insight Robotics has been using GeoXpert online platform to deliver results to clients.