Commonly deployed monitoring techniques (e.g. aerial surveys, camera traps), can be labour intensive, require substantial investment in equipment and personnel, carry safety risks, and rely on indices rather than enumeration. The application of thermal sensors to ecological and wildlife monitoring purposes is a rapidly growing area of investigation.
Thermal sensors have the potential to address common issues associated with traditional survey techniques such as visual acuity and observer fatigue, especially when attempting to detect cryptic targets or surveying large areas. However, recorded observations (thermal or otherwise) still generate hours of footage that requires time consuming and laborious analysis.
Automated computer software systems for detecting and identifying target objects from thermal imagery have the potential to quickly and accurately analyse large imagery datasets. This project will develop an automated thermal imagery analysis models that incorporate artificial intelligence and deep machine learning to further enhance low-cost and time-efficient processing. This work will provide a central analysis platform that is compatible to low, medium and high resolution thermal imagery, and equally accessible to all stakeholders and end-users.
The project receives funding from the Australian Government Department of Agriculture, Water and Environment