In this paper, we propose a novel inverse Dynamic Reachability Map (iDRM) that allows a floating base system to find valid end-poses in complex and dynamically changing environments in real-time. End-pose planning for valid stance pose and collision-free configuration is an essential problem for humanoid applications, such as providing goal states for walking and motion planners. However, this is non-trivial in complex environments, where standing locations and reaching postures are restricted by obstacles. Our proposed iDRM customizes the robot-to-workspace occupation list and uses an online update algorithm to enable efficient reconstruction of the reachability map to guarantee that the selected end-poses are always collision-free. The iDRM was evaluated in a variety of reaching tasks using the 38 degree-of-freedom (DoF) humanoid robot Valkyrie. Our results show that the approach is capable of finding valid end-poses in a fraction of a second. Significantly, we also demonstrate that motion planning algorithms integrating our end-pose planning method are more efficient than those not utilizing this technique.