Investigating the potential for hazard tree identification using thermal imagery from the DJI Mavic 2 Enterprise Dual

Abstract

The Mavic 2 Enterprise Dual was recently released by DJI in December of 2018. It is a small compact device with enormous potential for field work in a variety of industries, one of which has been investigated in this undergraduate thesis report. Public parks and recreation areas are becoming an important part of our health and well-being during an increasingly technological and urban time. However, safety is always a concern with public participation in any activity, and one which wardens and managers are constantly trying to improve upon. Approximately 11% of deaths or injuries that occur during outdoor recreational activities have been the result of falling trees or tree branches (Brookes 2007). Trail inspections, in an attempt to identify hazard trees that are dead or rotting before causing issue, can be infeasible due to a number of conditions, making the rise in remote sensing and drone technology potentially revolutionary to this field. The Mavic 2 Enterprise Dual is equipped with dual thermal and visual cameras. Thermal imagery is incredibly useful for identifying objects that are less visible with traditional imagery by using different heat signatures. Thermography has been used across a range of disciplines including engineering, medicine and perhaps most relevant, arboriculture. Although not thoroughly researched, numerous case studies have shown that zones of decay can be seen inside standing trees using thermal imagery at ground level (Catena & Catena 2008). Assessing individual trees in this manner may not be particularly useful for identifying hazard trees in large public parks but it begs the question of whether aerial thermal imagery could potentially be implemented in the same manner and if so, would the Mavic 2 Enterprise Dual be an appropriate tool for the task.