Using drones to map habitats: data fusion and classification

Using drones to map habitat (training course)

Date:

This course last ran in January 2019. The next course will run in summer/ autumn 2019 subject to demand

Please express your interest online here so we can fix a date and venue!

Location:

CEH Lancaster, Lancaster Environment Centre, Library Avenue, Bailrigg, Lancaster, LA1 4AP Location map & travel directions http://www.ceh.ac.uk/lancaster

or CEH Wallingford (near Oxford)

Cost: 

 standard £599 (Early Bird, later £649)

 students £499 (Early Bird, later £549)

Refreshments and lunches are included in the cost. 

Short Course Description:

This two-day course will provide participants with the skills needed to take images collected from a UAV platform such as a quadcopter, process the images into a scene suitable for image processing, and generate classified images. We will talk you through the steps necessary to produce high quality and accurate maps suitable for the mapping of vegetation and land forms.

It will cover the following topics:

  • UK Civil Aviation Authority guidelines and the law
  • Drone technology
  • Aerial photography
  • Image stitching software
  • Image classification

Course Objectives:

  • Gain skills to create high quality and spatially accurate habitat classification maps from data you have captured using your UAV platform.
  • Learn how to create digital elevation and surface models.
  • Learn how to process these data to map the heights and locations of trees and vegetation or other features
  • Learn how to operate your UAV platform within the law
  • Gain  insight into some of the problems you may encounter when flying your aircraft
  • Learn how to bring together the data collected by your aircraft into a spatially accurate and informative map suitable for mapping habitats and change.

Software used during this course

We will use a number of different software for the different stages of processing the images. Key principles will apply to whatever software you are using.

  • R open source programming language for the classification of data
  • ESRI/ArcGIS for the spatial data aspects
  • PhotoScan for a demonstration of the photogrammetric processing of digital images. On request, we can also do a short demonstration in PIX4D 

Some prior knowledge of these will be an advantage, but not essential to benefit from the course. All of this software is available as short free trial versions.

Accommodation:

Accommodation is not included in the cost.  Please see our guide to accommodation near CEH Lancaster.

Places: 20

Level:

Intermediate - Course participants should be familiar with the manipulation and fusion of spatial data, GIS and/or image processing and UAV platforms for data capture.

 

Course Leader:

Dr. Paul Scholefield, Ecological Modeller and Spatial Analyst, Centre for Ecology & Hydrology

Paul is researching the use of Unmanned Aerial Vehicles (UAVs)  for fine-scale mapping and habitat classification of remote environments. He is a qualified RPA-S pilot, and proficient in the use of UAV derived imagery to generate high-resolution habitat classification maps.

Partner organisations and other trainers:

Dr Michael James, Senior Lecturer in Lancaster Environment Centre, Lancaster University

Michael uses field measurements of active processes using Unmanned Aerial Vehicles (UAVs) and other technology. Michael is currently exploring the use of structure-from-motion algorithms for facilitating 3D surface generation from photographs. 

 

Relevant links and key publications:

Science takes to the skies with ‘Hollywood-style’ remote sensing drone technology.