In recent years, crisis management's response to terrorist attacks and natural disasters, as well as accelerating rescue operations have become an important issue. In rescue operations of fire disaster, one of the biggest problems is that the firefighter?s view is obstructed by dense smoke. We considered the most important task for firefighters is to understand the inside situation of the dense smoke space. Therefore, we aim to develop a system visualizing the situation of the space. First, we scanned target space by using millimeter-wave radar combined with a gyro sensor. Then, due to detect humans and objects, we construct a 3D map from signal-reflection datasets using 3D image processing techniques. In this paper, we introduce our system and report the results of the measurement experiment in the real smoke space situation.