PHYSICAL ENVIRONMENT IMPACT ON MINE AND IMPROVISED EXPLOSIVE DEVICE DETECTION TECHNOLOGIES

The military community involved with range cleanup activities such as unexploded ordnance (UXO) or munitions of explosive concern (MEC), warfighter support, and humanitarian demining programs focused on explosive remnants of war (ERW) share many of the same technical issues. One common theme is the desire to inexpensively, rapidly detect, and identify UXO, MEC, mines, and improvised explosive devices (IEDs) spread over large geospatial areas. Over the past decade, the Army Corps of Engineers (USACE), Engineer Research and Development Center (ERDC) and others within the military community have made considerable progress on identifying useful sensor modalities for detection surface and subsurface objects of interest. To that end, there has been recent recognition of the role the physical environment plays in sensor selection and object detectability. The presentation will discuss an improved understanding on how the physical environment (soil, weather) affects the visibility/detectability of buried targets by sensor modality. Considering the environmental phenomenology affecting sensor performance, a new approach using computer vision (CV) techniques combined with machine learning (ML) algorithms to improve detection performance is discussed. This synergy of environmental understanding and CV/ML approaches has been incorporated into ERDCs prototype two-step automatic target detection and recognition (AiTD/R) environment informed algorithm, referred to as AiTD/R Envi.