Special purpose machines have been recently developed to support prompt rescue and initial recovery in the event of a disaster or accident. The machine can perform various tasks such as holding the obstacle, spreading it on both sides. And It can move the unstable field freely through the four crawlers. On the other hand, it is difficult to grasp the topographical information on the moving route only by the visual or image of the driver due to the residue of building, dust, weather, etc. This may result in delays in responding to the initial operation of the machine or a second accident such as an overturning. This paper develops terrain cognitive technology using heterogeneous sensors such as image sensor and distance measurement sensor to provide driver with topographical information of movement route. This paper proposes a two types topography recognition method.