The reliability of rock mass geometry characterisation is dependent on the quality and quantity of discontinuity measurement. Time constraints especially at a working rock face and size of the sampling window severely limit the amount of data which can be collected. This paper presents the outline of an aligorithm utilising digital image processing techniques for the development of an automated system which traces discontinuities from photographs via digital imagery. The algorithm is currently being written in the C programming language to form part of an integrated rock mass characterisation computer application. The new system will also benefit existing software which model rock masses and model rock mass permeability.
La fidelite de la characterisation de la geometric d'une masse rocheuse est dependante sur la qualite des measures des discontinuites. Les contraintes de temps, en particulier a une face rocheuse active. et lu tailIe de la fenêtre d'ehantillonage limitent sevèrement la quantite de donnee pouvant être ressemhlee. Ce papier preseuic l'aperçu d'un algorithme utilisant les techniques de traitement d'image digitale pour la development d'un systeme automatise qui trace les discontinuites a partir des photographies via I'imagerie digitale. L'algorithme est en train d'être ecrit en language de programmation C pour former part d'une application integre sur I'ordinateur de characterisation des masses rocheuses. Le nouveau systeme beneficiera aussi d'un logiciel existant qui modèle les masses rocheuses et leurs permeabilites.
Die Zuverlaßigkeit einer geometrischen Charakterisierung des Gebirges hangt von der Anzahl und Qualitat der Trennflachenmessungen ab, Sowohl der Zeitbedarf insbesondere an einer in Abbau befindlichen Felswand und die Große der erreichbaren Meßflache begrenzen entscheidend die Anzahl der einmeßbaren Daten. Dieser Aufsatz umreißt einen Algorithmus fuer die Entwicklung eines automatischen Systems, bei dem mit Hilfe von digitaler Bildverarbeitung die Spuren von Trennflachen auf Photos erfasst werden. Der Algorithmus wir zur Zeit in der Programmiersprache C erstellt und wird ein Teil eines Computerprograms zur ganzheitlichen Gebirgscharakterisierung. Das neue System wird auch fuer bestehende Programme von Nutzen sein, mit denen das mechanische und hydraulische Verhalten des Gebirges simuliert wird.
Discontinuities within a rock mass typically lead to Complex structures on the engineering scale, as illustrated in Figure 1. It is recognised that discontinuities can have a dominant influence on the stability, deformability and permeability of a rock mass. An increase in the quantity of discontinuity geometrical measurements will lead to an Improvement in the reliability of rock mass characterisation and ultimately to the betterment of engineering design. Current techniques for rock mass geometrical characterisation involve measurement of a limited sample of discontinuity spacings, orientations, extent and surface roughness followed by the adoption of rock mass classification schemes and/or stochastic modelling. Measurement of these parameters is a time consuming process and hence typically only a small amount of data is obtained. Assuming that discontinuity spacing values follow a negative exponential distribution, Priest and Hudson [10] have shown that several hundred measurements are required to obtain reasonably reliable estimates of the mean discontinuity spacing in a rock mass. Obtaining very reliable estimates requires a dramatic increase in the number of measurements. For example if we wish to be 99% confident that measurements of discontinuity spacing fall within ± I% of the mean spacing value, then it is necessary to make at least 66358 measurements of spacing. Existing measurement techniques, as suggested by the ISRM [5], are inadequate in coping with this demand.
