Impact of Gap Filling on Quality of Road Networks

Abstract

High quality and updated road network maps provide important information for many domains. Many small segments appear on the road surface in VHR images. Most road extraction systems have problem in extraction of these small segments and usually they appear as gaps in the final extracted road networks. However, most approaches skip filling these gaps. This is on account of the fact that usually overall length of the missing parts of the road extraction results is very short relative to the total length of the whole road network. This leads to an indiscernible impact of filling these gaps on geometrical quality criteria. In this paper, using two different VHR satellite datasets and a gap-filling approach which is based on tensor voting, we show that utilizing an effective road gap filling can result in a quite tangible topological improvement in the final road network which is highly demanded in many applications

Automatic Road Extraction from High-Resolution Remote Sensing Images Using a Method Based on Densely Connected Spatial Feature-Enhanced Pyramid

“…At the same time, the first two indicators correspond to the completeness and quality of classic road extraction [58]- [59]. They can evaluate the geometric quality of the extracted road network [60,61]. This paper considers road extraction as a two-class classification problem, and the prediction results are divided into two categories: road and nonroad.…”