In this paper, the movement behavior of a virtual human based on realistically limited perception (RLP) is proposed to be human-like. As an interface between perception and movement-path generation, a mapping module is a fundamental component needed by a virtual human. Research of the mapping based on RLP was performed by Hill et al. However, their research was conducted using only a camera’s view point. In this present research, a virtual human’s integration with Hill et al’s mapping and other variables (e.g., enemy emergence) is considered in the context of a reconnaissance mission. The loci of the movement paths that were generated by human subjects and a virtual human based on RLP are compared with each other.