Journal of Coastal Research, Special Issue 56, 2009 Journal of Coastal Research SI 56 1484 - 1488 ICS2009 (Proceedings) Portugal ISSN 0749-0258 Using Stereo Photogrammetry to Measure Coastal Waves S. de Vries†, D. Hill‡, M.A. de Schipper† and M.J.F. Stive† †Section of Hydraulic Engineering Delft University of Technoloy, Delft 2628 CN, The Netherlands Sierd.deVries@TUdelft.nl M.A.deSchipper@TUdelft.nl M.J.F.Stive@TUdelft.nl ‡ Civil and Environmental Engineering Pennsylvania State University, University Park, PA, 16802, USA Dfh4@psu.edu ABSTRACT DE VRIES, S., HILL, D., DE SCHIPPER, M.A. and STIVE, M.J.F., 2009. Using Stereo Photogrammetry to Measure Coastal Waves. Journal of Coastal Research, SI 56 (Proceedings of the 10th International Coastal Symposium), 1484 – 1488. Lisbon, Portugal, ISSN 0749-0258. This paper summarizes a field campaign designed to investigate the application of stereo photogrammetry to measure water surface elevations. The goal of the field test was to show that, using this technique, wave transformations in near shore waters can be measured with high spatial and temporal resolutions in a large spatial domain (O(10 4 m 2 )). These data can then be used to analyse the propagation and transformation of waves and to formulate and verify numerical models. Also, the technique can be used as a real time monitoring technique in the coastal zone. Previous studies show that stereo photogrammetry has great potential. The contributions of the present study are that (i) it is the first true field-scale test of the technique and that (ii) the stereo video algorithms have been modified to tackle the specific challenges of the problem. Most stereo matching is based upon well lit and very high texture surfaces. We face greater challenges, with variable lighting characteristics and highly variable texture. The measurements took place in Scheveningen on the South Holland coast. Two cameras imaged an overlapping area of about 5e3 m 2 . Stereo images were acquired for 5 minutes at a rate of 8 Hz, yielding a very complete record of the water surface evolution. In this paper we present the results and application of the technique as a whole. Hardware requirements, calibration, correlation and triangulation techniques. Initial Results of reconstructions prove to be very promising. ADITIONAL INDEX WORDS: Remote sensing, Video imaging, Surf zone INTRODUCTION Remote sensing using video imagery techniques is common in coastal studies and as a management tool. For example, HOLLAND et al. (2001) applied particle image velocimetry (PIV) methods to the swash zone. This technique uses two images of the same field of view, but taken at different times, in order to determine the spatial displacements (and therefore velocities) of ‘tracers’ in the images. These tracers provide visual texture in the images and, in the coastal environment, are provided by bubbles, foam, ripples, and other features. As another example, the Argus imaging technique is used to monitor morphologic behaviour of coastal zones (HOLMAN and STANLEY, 2007). The initial focus of the Argus program was on time exposure photographs where the image intensity at each pixel could be taken as a proxy for water depth. Interpreted in this way, the Argus data provided inexpensive non-contact bathymetry information. The automated nature of the Argus instrumentation meant that individual coastal sites could be sampled repeatedly over periods of days, months, and years. Recently, there have been some efforts using a stereo camera setup to measure the elevations of a dynamic water surface. HOLTHUIJSEN (1983) provided an initial description of the method and of an operational system using two airborne cameras. The lack of digital cameras and automated analysis methods at the time of the study led to prohibitive post-processing times. PIEPMEIER and WATERS (2004) recently discussed a laboratory implementation of the stereo imaging of waves. In their study, they considered only monochromatic waves and found it necessary to artificially ‘roughen’ the water surface in order to provide adequate visual texture. WANEK and WU (2005) investigated the use of trinocular imaging for the purposes of imaging waves in a field environment. However, their study was limited to a very small field of view (O(1 m 2 )) and it relied upon a delicate calibration method that may not be suitable for large scale applications. Moving up in spatial scale, Benetazzo (2006) presented results of the stereo imaging of waves from two field campaigns, one 4 m 2 in area, the other 400 m 2 in area. SANTEL et al. (2004) describe a true field test of stereo photogrammetry which covers a domain of 40,000 m 2 in area. While promising, the accuracy of their camera setup was limited mainly due to the high distance between the cameras and the area of interest. The goal of the present paper is to further the ability of stereo imaging methods to measure waves in the surf zone. An ideal system will be able to measure waves over a large area and with good spatial and temporal resolution. Additionally, an ideal system will be able to be rapidly installed and calibrated. Finally, an ideal system will have reasonable image processing times, with an eye towards real-time acquisition and processing. 1484