Researchers from the Institute for Environmental Solutions (IES) use drone data and other new technologies to develop the methodology for wild animal counting in Latvia. It will enable objective and research-based decision making on sustainable wildlife management, additionally helping to save costs for forest management professionals.
Wildlife is Latvia's valuable natural resource and ungulates such as red deer, roe deer, elk and wild boar are a part of it. According to data of European Union, wild animals give an annual contribution of above 394 million Euros to the economy through game meat production. SIA “Forest Owners’ Consulting Centre” forestry expert Gvido Prudņikovičs describes the wild animal counting in Latvia: “management of wild ungulates is based on estimates from field-based animal counting campaigns in which standard methods such as animal footprint tracking in snow, animal counting at the feeders, counting of excrements during winter season and estimations from damages done by animals are used. These methods are not efficient and cost-effective. Furthermore, landowners (state and private) do not get accurate information about the real density of ungulates in their territories,”
Previous research shows that the usual field-based wild animal counting techniques do not meet the necessary precision and efficiency goals. To improve this process IES researchers in cooperation with SIA “Forest Owners’ Consulting Centre” and Latvian State Forest Research Institute "Silava" test various technologies - drones, motion-activated camera traps and passive acoustic sensor networks as well as deer tracking with GPS transmitters.
“Currently there are no such technology based wild animal counting solutions developed that would allow to do this process automatically and remotely. When managing large forest areas, it is hard to oversee the whole territory and determine the size of wildlife populations. Therefore, I see great potential in these remote sensing solutions, especially animal tracking with drones. The advantage of drones is the ability to cover large territories in a short time span and to reach places that are difficult to access by foot,” explains forestry expert Gvido Prudņikovičs.
“All the technologies involved complement and validate each other. Camera traps and microphone networks allow us to remotely see the processes that take place in the forest from a human point of view. Drones equipped with cameras give an additional perspective on the research area - from above. This dimension allows us to increase the precision of wild animal counting methods. For example, drone data shows the animals that are not captured by camera view length,” IES remote sensing expert Jevgēņijs Fiļipovs talking about technologies used in this research.
After the first technology tests, IES researchers have concluded that the use of drones gives relatively precise information about the size of animal population in the territory of interest – up to 90%. The precision measurements were done in deer garden within Rāmuļi parish in Latvia. During these tests the researchers used drones equipped with two cameras – thermal imaging and visible light (RGB) cameras. While a thermal camera is used for spotting the location of a warm animal on the cold ground, a visible light camera is used to identify target objects in the data – detect whether the warm object is an animal and identify the animal species.
“For now, identification of animals and their species in drone data is done manually but we are working on automatization of data processing. By using machine-learning technologies, we develop a computer-vision algorithm. We “feed” this algorithm with manually processed drone data samples of wild animals. As a result, the algorithm will be able to automatically recognise animals and their species in drone data without the assistance of the researcher,” Jevgēņijs Fiļipovs describes the data processing.
IES’s scientists are certain that this research will lead to the development of automated solutions for remote evaluation of wildlife populations. This approach will support objective and research-based decision making on sustainable wildlife management. Moreover, it will help to save costs for forest management professionals.
The research is a part of the project “ICT-based wild animal census approach for sustainable wildlife management” (No. 220.127.116.11/18/A/146) is part of European Regional Development Fund, 1.1.1 "Improve research and innovation capacity and the ability of Latvian research institutions to attract external funding, by investing in human capital and infrastructure" 18.104.22.168. measure “Support for applied research”.
Find more about this project here.