Roughness Characterization of an Indoor Environment

Douiri, Nisrine; Khoudeir, Majdi; Olivier, Christian

doi:10.5566/ias.v20.p225-229

Cited by 2 publications

(1 citation statement)

References 9 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Studies characterising roughness by analysis of grey level images [1], [2], [3] generally assume that the variations in grey levels in an image are representative of variations in relief. This hypothesis, which is justified in the case of uniformly coloured surfaces, may no longer be valid in the case of surfaces which are not uniformly coloured.…”

Section: : Introductionmentioning

confidence: 99%

Extraction of relief information on textured and coloured rough surfaces for the optimisation of a propagation model for wireless transmission

Anis

Khoudeir

Jacques

First Canadian Conference on Computer and Robot Vision, 2004. Proceedings.

View full text Add to dashboard Cite

This study is carried out within the framework of the development of simulation systems of high frequency radio wave propagation (60 GHz) for wireless local area networks in an indoor environment. Thus the corresponding wavelength is in millimeters. At this frequency, in order to optimise models of radioelectric wave propagation, it is important to have information related to the 3D roughness of the main reflective surfaces encountered during the transmission. But the estimation of the relief of 3D textured surface is generally made on grey level images. This supposes that variations in grey levels are representative of local variations in the relief. This assumption is justified in the case of uniformly coloured surfaces, but is no longer valid when these surfaces present variations of colour or aspect. The corresponding image will then present variations in grey levels which can be related to colour variations or relief variations or both. It becomes difficult in this case to evaluate relief based on image analysis. Before any study of roughness, it is therefore necessary to devise a method for separating the information linked to colour variation from the information linked to relief variation. In this paper, we propose to carry out this separation through a photometric stereovision system. The method we have developed is based first on the acquisition of three images of the studied surface, obtained under different light conditions, and second on the photometric model of the surface. So we have established the relationship between the local relief of the surface, its colour aspect and the corresponding grey level image. Then, for the studied surface, we have extracted an image representative only of its colour aspect and an image representative only of its local relief. Finally, we have extended the proposed method to the case of coloured images. 1: IntroductionThe study below is a contribution to the analysis of roughness of coloured, textured, 3D surfaces. Studies characterising roughness by analysis of grey level images [1], [2], [3] generally assume that the variations in grey levels in an image are representative of variations in relief. This hypothesis, which is justified in the case of uniformly coloured surfaces, may no longer be valid in the case of surfaces which are not uniformly coloured. In the latter case, the image of the surface presents variations in grey levels due to variations in relief or to variations in colour or to both. Before any study of roughness, it is therefore necessary to separate within the image that information linked to variations in colour from information linked to variations in relief. Different approaches with filtration using wavelets [4], [5] have been developed to isolate the relief information in these images. However, these methods may result in a smoothing effect on relief and may be ineffective in cases where the spatial frequency of information linked to colour is of the same order as that linked to relief. To address this problem, we propose to use a method of ...

show abstract

Section: : Introductionmentioning

confidence: 99%