Item

Abstract

This investigation is focused on the feasibility to conduct population statistics in developing countries, based on remote sensing data and consecutive field examination. Hereby, the average number of inhabitants per house unit (plot) is evaluated in field. These values are projected by numbers of plots evaluated on aerial photographs/maps and MOMS-02 spaceborne data to an estimated population number of areas formed by homogeneous types of houses (characterized by a size of 6 m x 3 m and up to 15 m x 7 m). Usually, this kind of RS-information is only provided by photographic systems e.g. aerial photographs, KVR-1000. A resolution of 10 m x 10 m like SPOT-HRV is still insufficient to this problem. The high resolution of MOMS-02 panchromatic data (4,5 m x 4,5 m) however, allows to distinguish clearly the single plots in high density sub-urban areas. Based on this information, population numbers are calculated for training and test are as of different but homogeneous housing units. Data are then compared to census data and evaluations using aerial photographs/maps. The results show that MOMS-02 panchromatic data in connection with the developed methodology can be used for this purpose as only minor deviations occur.