Survey of UAV mounted landmine detection technology capabilities with consideration for soil and vegetation types

One of the greatest threats posed to human life after a conflict is unexploded ordinances in the form of landmines, improvised explosive devices (IEDs), and bombs. In 2021 alone, it is estimated that over 5,500 people died from unmarked landmines, most of them civilians (Billy, 2023). The traditional method of landmine detection is slow, expensive, labor intensive, and places human lives at risk. Several current studies use remote arial sensing to identify landmines with infrared cameras, RGB cameras, radar, and other technologies. Many of these studies have used machine learning systems to map the locations of landmines but failed to detect fully covered explosives and have been tested in a limited number of terrains. This paper aims to gather data on the various detection methods currently being tested and identify their limitations and opportunities, by completing a systematic literature review of the current literature.