Verification of a 3-D LiDAR point cloud viewer for measuring discontinuity orientations

LiDAR (Light Detection and Ranging) scanners are increasingly being used to measure discontinuity orientations on rock cuts to eliminate the bias and hazards of manual measurements which are also time consuming and somewhat subjective. Typically LiDAR data sets (point clouds) are analyzed by sophisticated algorithms that break down when conditions are not ideal, eg. when some of the discontinuities are obscured by vegetation, or when significant portions of the rock face are composed of blast fractures, weathering generated surfaces, or anything that should not be identified as a discontinuity for the purposes of slope stability analysis. This paper presents a simple LIDAR point cloud viewer that allows the user to view the point cloud, identify discontinuities, pick 3 points on the surface (plane) of each discontinuity, and generate discontinuities orientations using the three point method. A test of our 3-D LiDAR viewer for discontinuity orientations on three rock cuts in the Golden Gate Canyon Road area of Colorado is also presented. large blocks of rock to fall or slide down from these steep rock cuts. The greater the number discontinuity planes present in the rock mass, the higher the chances of failure since many of the failures result because of release along discontinuity planes. Whether or not failure occurs can depend largely on the orientation of the discontinuities, individually or in combinations (Figure 1). Thus, knowing the orientations of the discontinuities can lead to stability prediction based on well established analytical tools as described by Hoek and Bray [7]. Orientations are typically measured manually in the field using a compass and clinometer. These methods are manual and have disadvantages which include the introduction of erroneous data because of sampling difficulties and human bias, considerable safety risks since measurements are sometimes carried at the base of existing slopes or during quarrying, tunneling or mining operations or along busy highways, difficult or impossible access to some sections of rock faces, and are time consuming and labor intensive which make them costly [8]. Laser scanning and digital images can be less costly, more objective and more precise and accurate in determining discontinuity orientations [9,10]. Figure 1. A rock mass showing a discontinuity along which a rock block slid. The block at the top left corner is also likely to slide with time. For a given rock mass, measured discontinuity planes can be assigned by using cluster analysis. Cluster analysis techniques are described in detail by Maerz and Zhou [6, 11, 12, 13]. Once having identified discontinuity clusters, graphical or computational techniques can be used to determine the kinematic feasibility of failure (Figure 2) and standard modeling techniques such as limiting equilibrium analysis can be used to determine if failure will indeed take place (Figure 3). Figure 2: Planar failure geometry (left) and graphical method of determining if slide failure is kinematically possible [7]. Figure 3: Limiting equilibriums analysis applied to planar features (left) and wedge features (right) [7]. 1.2. LiDAR Scanning A LiDAR (Light Detection and Ranging or Light RADar) scanner uses either a time of flight or phase shift sensors to generate a 3-D image of a surface. It involves the emission of light pulse from a source, which reflects off surface of the object is reflected and returns to the source which then receives and measures it [9]. A high precision counter measures the travel time and intensity of the returned pulse. The pulse source also measures the angle at which the light pulse is emitted and received, these enables the spatial location of a point on a surface to be calculated [9]. The result is a million of points reflected from the surface. The points are represented by xyz coordinates, these xyz coordinates and their associated intensity values are known as a “Point cloud”. The LiDAR 3-D technology is becoming increasingly useful in geology and engineering. Kemeny et al. characterized rock masses using LiDAR and automated point cloud processing, and also analyzed rock slope stability using LiDAR and digital images [14, 15], including measuring and clustering discontinuity orientations. LiDAR was used by Mikos et al. to study rock slope stability [16]. Lim et al used photogrammetry and laser scanning to monitor processes active in hard rock coastal cliffs [17]. High resolution LiDAR data was used by Sagy et al. to quantitatively study fault surface geometry [18]. Enge et al. illustrated the use of LiDAR to study petroleum reservoir analogues [19]. Using a combination of LiDAR and aerial photographs, Labourdette and Jones studied elements of fluid depositional sequences using LiDAR [20]. Figure 4: Rock faces with 100% coverage of natural joint surfaces (a) and with significant ambiguity as to the location of natural joint surfaces (b). Automated algorithms used to generate discontinuity orientations are in general fairly sophisticated and can give excellent results under certain conditions. In places where rock faces are virtually 100% bounded by discontinuities they work well; in places obscured by vegetation, rock projections, or surfaces created by recent fracturing because of blasting or weathering not so well (Figure 4). In the latter case the algorithms will break down. Although vegetation removal algorithms could be used, this adds another layer of difficulty to both the data collection and analysis sides. It is often better just to manually identify discontinuities on the LiDAR point cloud. Figure 5: Leica ScanStation 2 LiDAR unit. Table 1: Features and specifications of the ScanStation 2 unit (modified from Leica webpage, 2012) Feature Specification Laser scanning type Pulsed; proprietary microchip