Integration of GPS and Remote Sensing into GIS: A Case Study of Rectifying Satellite Imagery Using Uncorrected Coordinates in Different Scenes

Abstract

This paper explores the feasibility of integrating GPS and remote sensing data into a GIS database through rectification of satellite images of both urban and rural areas using uncorrected GPS coordinates. A GPS receiver was used to log the coordinates of 25 ground control points (GCPs) in a rural area and 20 GCPs in an urban area. Their coordinates were also read from topographic maps. Both sets of coordinates were used to rectify Landsat TM, and SPOT panchromatic (PAN) and multispectral (XL) images. It is found that none of the GPS coordinates are able to achieve an acceptable accuracy level of integration unless some inaccurate GCPs are excluded. The percentage of useable GCPs varies from less than 50 to nearly 90, depending on the spatial resolution of the image and the ease of GCP identification. On average, the difficulty in identifying GCPs on the ground and the image degrades integration accuracy by up to one pixel.