3699245 orig

2015 Hydrological station and UAV Imagery

Hydrological data, using well distributed and low cost stations

System of measuring and helping to monitor waterways based on image analysis and processing.

It uses the images supplied by a surface-based image acquiring device installed very close to a waterway or canal to measure the water’s height and surface flow velocity and calculate the discharge.

The non-intrusive nature (not submerged) of this solution means that it continues to function during extreme events such as flooding. Designed as a hydrometric information service it is a "4 in 1" data and measurement display system (height, velocity, discharge) making it possible to:

-      create and update a rating curve using a number of discharge estimates calculated by the software using the images;

-      obtain two discharge calculations: the discharge estimated by the current rating curve (Q) systematically, as well as the discharge estimated using the surface velocity measured (Qcalc) on an ad hoc basis;

-      detect changes in hydraulic behaviour (i.e. rating curve shift)


Hydromed Station hydromedv2


Simple schema to show low cost hydrological stations.

Camera's and staff gauge (left side red line)position

Prototype of low cost station.

Waterproof enclosure and electronics elements.

UAV (unmanned aerial vehicle) and Near Infra Red Sensor

An UAV (unmanned aerial vehicle) is an unpiloted aerial vehicle and a remotely piloted aircraft. It's a modern platform which is increasingly used as alternative of traditional aerial vectors such as aircraft and helicopter for different utilisation.

UAV photogrammetry strength:

-      Flexible in time. It's deployable during narrow meteorological windows

-      Flexible in geographic footprint. It can be design of specific flight patterns such river, and vegetation path

-      Really high pixel size


Bing satellite image (résolution 50 cm ) Potentiel3.0 UAV image (résolution 4cm)

For large area > 1 km2:

Potentiel3.0 's uses a very small professional mapping UAV: senseFly’s Ebee. It's operates with pre-programmed flight plans with automatic take-off, in-flight high-definition photography and landing. It can cover 1km2 in 30 minutes (battery life).

Since 2011, more than 300 km2 has been covered in Haiti

Technical specifications of the Ebee.More information on : https://www.sensefly.com/drones/ebee.html

For small a < 1km2

Potentiel3.0 's uses quadcopter or octocopter, exemple: DJI Phantom 3


Geospatial image processing

Here below some of the main image processing steps applied on raw UAV images:

a)   Initial processing to convert the aerial images into geo-referenced 2D mosaics and 3D models. 

 Processing software. Pix4d and/or Micmac (IGN Open source software).



Small zoomed area into Tisous basin, Haiti. Orthmosaic image and on Digital Surface Model extraction.


b)   Post processing. Rectification of the Orthomosaic spectral bands, image's fusion, Digital Surface Model cleaning,

Processing software.OrfeoToolBox (Remote Sensing software) and Qgis (GIS software), both are Open-source software.




Small zoomed area into Tisous basin, Haiti.  Different filter and spectral band's combination.




 The UAV images offer many opportunities for analysis.

 In 2016 , we hope to deploy hydrological stations to refine flood simulations to Improve risk reduction and mitigation project




Where we Work

Voir en plein écran