ON SOME APPLICATIONS OF STATISTICAL-METHODS IN REMOTE-SENSING

In the recent years Remote Sensing of earth has gained wide popularity in view of its regional and repetitive coverage in a cost effective m anner. Remote sensing data being bulk in nature is subjected to statis tical analysis for deriving meaningful information pertaining to mappi ng, monitoring and assessment of natural resources in addition to moni toring of natural hazards. Classification techniques are extensively , applied for generation of thematic maps of soil, forestry applications . The sampling techniques are used for assessing the accuracy of thema tic maps and generation of forest resources inventory. Relaxation tech niques are being utilised in image analysis for edge detection and enh ancement, shape recognition etc. Texture analysis of remote sensing da ta reveals regions of interest in an image. The multivariate statistic al analysis of geological maps derived from remote sensing data are be ing applied for mineral exploration studies. The soil expert system de veloped utilising satellite image can be put to use as a tool for trai ning soil scientistis. The present paper deals with many of these stat istical techniques as applied in remote senging data for natural resou rces mapping and management.