3D extraction of the relief of road surface through image analysis

We have developed a technique for determining the 3D mapping of the texture of road pavement. From the point of view of image processing, road surfaces can be considered as textured images including both micro and macro-textures. The extraction of relief is based on modeling the photometric behavior of the surface. This approach has its limits when the analyzed surface contains specular aggregates. In this case, the Lambertian model is no longer valid and the relief extract is unreliable. One solution may then be the development of photometric models according to specular phenomena. But these models are difficult to use in terms of inversion of equations. For this, we propose an original method in four steps: the first step determines the different image areas, specular and diffuse ones. In a second step we consider the surface as a set of micro-facet and their slopes are determine from the geometrical configuration of the image shooting system. A third step determines the slopes of micro facets in diffuse areas, from the model of Lambert. The last step reconstruct from these slopes the relief of the entire surface.


  • English

Media Info

  • Pagination: 8p
  • Monograph Title: Transport Research Arena (TRA) 2014 Proceedings

Subject/Index Terms

Filing Info

  • Accession Number: 01530658
  • Record Type: Publication
  • Source Agency: ARRB
  • Files: VTI, TRIS, ATRI
  • Created Date: Jul 23 2014 12:53PM