Laser-based intelligent perception method for tunnel invasion

Abstract

Laser-based visual perception technology is an effective way to enhance the intelligent level of tunnel operation and maintenance, which could improve the efficiency of the automatic identification and intelligent positioning of different diseases. Although various sensing and data processing methods for tunnel inspection have been studied, the research on invasion detection based on laser scanning technology is still rare. In this paper, the invasion detection of the tunnel and train is carried out based on terrestrial laser scanning data. A feature-based registration is proposed in order to fuze the tunnel and the train models accurately and automatically in the unified coordinate system, where the local contact points are set as the features and identified by means of the normal vector descriptor. The invasion analysis is achieved with the convex points, which are paired through the convex hull and polar principle. A complete and comprehensive image of the invasion analysis is generated, which offers a reality model base for the design and decision-making. The method proposed can be extended to other similar scale scenarios of invasion analysis and spatial position inspection. It is a reference for the engineering applications of the intelligent operation and maintenance of the tunnel.