Automated Tree Mapping Automated Tree Mapping Automated Tree Mapping

Dr Kavitha Muthu

MSc, PhD

kavitha-muthu

Biography

Kavitha is the Geographical Information Systems (GIS) Specialist combining data from different sources (thermal, biomass: satellite, aerial) to carry out spatial analysis and produce output maps.

Dr Kavitha Muthu’s paper on Automated Tree Mapping.

Download pdf

Automated Tree Mapping.

Download pdf

Automated Tree Mapping

The Problem

Leicester City Council is using tree maps in order to accurately quantify the amount of tree canopy that currently exists in order to plan healthier, more sustainable urban environments. Traditional manual extraction of tree maps from aerial imagery is a labour intensive and time consuming process. G-STEP is working with local aerial photography company, Bluesky, to create tree maps using an automated classification scheme.

The Solution

The procedure involves the use of high resolution aerial photography to provide information on land cover classes and LiDAR  (Light Detection and Ranging) to provide accurate height data for the measurement of land cover structure. LiDAR technology uses light sensors to measure the distance between an aircraft and the ground, including objects such as buildings and vegetation. Processing of the height data produces a normalised digital surface model (nDSM), this model removes the effect of terrain elevation to give only surface feature information – representing the absolute height values of all objects. An aerial photograph of the same area is used to extract a greenness index, which is used to separate vegetation from non vegetated areas. Objects are classified as trees using the elevation data from the nDSM and the greenness index derived from the aerial photograph.