The main two approaches for vision based mobile augmented reality systems are either those employing fiducial markers or those which track natural features in the environment to estimate camera pose information. Whilst marker based systems are relatively simple to implement and are robust they present difficulties for wide scale deployment as they are obtrusive and their size is proportional to the distance from which they need to be used. However, the alternate approaches of marker less systems present significant computational challenges, can be highly problematic in poor light conditions, and are independent of scale. In the paper we present a novel solution using Infra Red LED’s as markers that overcomes many of these limitations in that they are: invisible to the human sight but can tracked by phone camera optics; can be used in varied light conditions; structured to provide scale; and significantly reduce the computational overhead.