Video Analysis of Runup 



Photographic techniques have been applied to the measurement of nearshore 

 waves and runup by a number of authors (Katoh 1981, Holman and Guza 1984, 

 Holman and Bowen 1984, and others). These techniques have the advantage of 

 being low-cost and logistically simple, and provide measurements over a large 

 longshore distance. The initial method of using photographic films was a time- 

 consuming process requiring manual digitization, which in turn had questionable 

 precision since it required subjective judgment of the operator. This is especially 

 true in the determination of rundown position, when percolation into the beach and 

 residual foam makes it difficult to determine an obvious swash edge. 



An improved method for runup measurements was developed using a video 

 image processor to automate the digitization of runup, which reduced the 

 processing time (U.S. Army Engineer Waterways Experiment Station (1990)). 

 This method measures the cross-shore position of the swash at several longshore 

 locations simultaneously. For each camera view a solution for converting two- 

 dimensional image coordinates to three-dimensional ground position is computed 

 using known geometrical transformations (Lippmann and Holman (1989), 

 summarized in the following section). Thus, swash positions digitized on known 

 beach profiles are transformed from image coordinates to a vertical swash 

 elevation. Since this technique relies on a change in image contrast between beach 

 and swash, it has difficulty in detecting swash positions when anomalous features 

 enter the viewing field (e.g., birds, persistent sea foam, or people). 



An analysis modification was made based on the "timestack" method described 

 by Aagaard and Holm (1989). The image processor digitizes picture element 

 (pixel) intensities along screen coordinates corresponding to a cross-shore transect 

 of the beach profile. Each sample of these intensities is then "stacked" in a matrix 

 and stored on disk. One dimension of the stack is cross-shore distance and the 

 other is time. Digitization can be done in real time (directly from a video camera), 

 or more typically post-processed from video tape. Subsequently, the data can be 

 retrieved, displayed on a monitor, and analyzed for a runup time series. The 

 timestack provides a two-dimensional image for which swash edge detection is 

 substantially improved over the multiple profile method. In addition, with the 

 timestack method, the image can be edited manually for checking and correcting 

 the edge detection before transforming to runup elevations. Although the 

 timestack method is presently limited to only a single profile, the results are 

 believed more reliable than the multiprofile method. This system uses an Imaging 

 Technologies video image processor (model ITI-15 1) interfaced to a Sun-4 host 

 computer. 



In a typical timestack (Figure 14) the swash is clearly visible as a sharp change 

 in intensity, from the darker beach (left) to the whiter foam of the swash (right). 

 Runup position is found using computer edge detection algorithms combined with 



32 Chapter 4 Field Observations 



