Automatic Change Detection for Road Networks from Imagery Based on Gis

Automatic change detection and updating road networks is a key work for updating geo-spatial information especially in developing countries. Old GIS data can make road networks detection and extraction easier. However, detailed strategy and algorithms of automatic change detection for road networks based on GIS is not discussed in detail. Aiming at this, a strategy for automatic change detection for images with low resolution and high resolution is presented firstly in this paper. Then a new so-called buffer detection change detection algorithm is proposed for detecting changes of features. Considering the difficulty of detecting road in high-resolution image a new profile tracing algorithm is developed in this paper. Some ideas and algorithms for using GIS prior information such as invariant objects like vector river boundaries, substructures of road like cars in high-resolution images to assist in road detection and extraction are discussed in the paper. The summary is given in the last part of the paper.