A survey of object recognition methods for automatic asset detection in high-definition video

Abstract

Asset management systems allow organizations to efficiently store data pertaining to the physical location of important assets. Asset detection is a key component of such systems, the automation of which greatly increases efficiency and for which object recognition techniques are an obvious choice. Recently, High-Definition video capturing equipment has become more prolific in these systems. Data captured with such hardware provides more information regarding distant assets, which can be taken advantage of in asset management systems. In this report, we present a survey of object recognition techniques applicable to the scenario of automatic asset detection despite asset distance from the camera. We also present an experimental comparison of a selection of methods with distance-variant asset data.