Positional accuracy control of extracted roads from VHR images using GPS kinematic survey: a proposal of methodology

摘要

This article describes a methodology developed for analysing the positional accuracy of extracted roads from very high resolution remote-sensing images by means of a Global Positioning System (GPS) kinematic survey (as control). Traditionally, this control has been performed by using a set of points located on the image, comparing their positions with those obtained from more accurate sources. In this work, we propose the use of lines for this purpose. These control elements are obtained by means of a more accurate kinematic GPS survey. Thus we analyse the main aspects of this methodology such as the population, size of sample, process (for obtaining the control set of lines), edition of lines, control methodologies based on lines, etc. We have applied the proposed methodology to a real example from an orthorectified Quickbird image. The results obtained show that the proposed methodology is more efficient than those based on isolated points.