The purpose of this study is to propose a real-time footstep adaptation mechanism for humanoid robots that can be integrated into a conventional bipedal walking planner and increase the robustness of walking against disturbances. One of the main advantages of the proposed methods is that it can take into consideration restrictions such as allowable landing region and relationship between foot movement distance and step duration. In order to meet the strict real-time constraint of humanoid robot control, the proposed method computes viable capture basins in the design phase. This pre-computed data can be used at run-time to modify foot placement and center-of-mass movement in response to applied disturbances with small computation cost. The performance of the proposed method is evaluated in simulation experiments.