What's the context?
The current preferred method of herbivore population assessment is expensive and has a high carbon footprint — we need to find greener and more cost-effective ways of collecting, collating, quality assuring and sharing the data relating to this task.
Scottish National Heritage Deer Counting method to date
The main method of deer counting uses helicopters (with a range of relevant staff/ stakeholders and equipment) to fly over properties / parcels of land to photograph and count deer. This approach is effective, but is costly in terms of financial and carbon costs. At present Scottish National Heritage (SNH) undertake 30–40 helicopter days per year, to a cost of around £200-300K per year. Other methods of deer counting also include dung assessment and foot counts but these also have limitations.
How are the team at Cubica helping?
We will look to use UAS technology and machine learning to provide semi-automated and automated deer counting and classification. We will apply technology to estimate herbivore populations and their impacts across Scotland in a greener and more cost effective way.
“We are pleased to announce that Cubica are part of a winning team for an exciting CivTech project that looks to use UAS technology and machine learning to provide semi-automated and automated deer counting and classification for Scottish Natural Heritage.
This important project seeks to exploit technology to help speed up and reduce the carbon footprint of existing deer management activities.
We are looking forward to getting stuck into designing the system which will use Deep Learning Convolutional Neural Networks and statistical analysis.”
Full project overview is available to read here: https://www.civtechalliance.org/civtech-5-challenge-2-scottish-natural-heritage