Recognition of human activities using handheld thermal systems

摘要

Development of low-cost lightweight thermal sensors necessitates the re-evaluation of detection, classification and recognition criteria as applied to human performance with thermal imaging systems. The need exists to assess spatial and motion characteristics of man targets, rather than vehicle targets, and how these differing characteristics affect human perception and performance models. Higher order discriminations, such as determining activities and objects being carried are useful in determining range performance of specific sensors, and reflect the needs of the man-portable sensor user.