Now a day, it is a growing demand for three-dimension (3-D) spatial information, especially in the fields related to geometric applications. Information achieved from 3-D visualization brings a great improvement in the job of disaster simulation, microclimate investigator, urban and environmental planning. The Advanced Land Observe Satellite (ALOS) has been newly developed (2006), however, has shown its important role in the fields of mapping, land observation, disaster monitoring and resource surveying. One of three ALOS sensors is Panchromatic Remote-sensing Instrument for Stereo Mapping (PRISM), which composes of three sets of optical system with different angles of view (nadir, forward, and backward views). Thanks for such optical components, PRISM could make a snapshot from three angles of view, and then, record 3-D terrain images with a high accuracy and frequency. However, with 2.5 meter ground resolution, ALOS PRISM triple image requires a professional method to read and explore 3-D information. This paper presents to audience a modern method, which allows extracting information automatically from ALOS PRISM triple image. The accuracy of this method has been checked by many experiment evaluations on different spatial structure areas.