Analysis of Rip Current Flow Fields in CCTV Images Using Optimal LSPIV Parameters

Jeong, S.-H.; Kwon, J.H., and Lee, J., 2023. Analysis of rip current flow fields in CCTV images using optimal LSPIV parameters. In: Lee, J.L.; Lee, H.; Min, B.I.; Chang, J.-I.; Cho, G.T.; Yoon, J.-S., and Lee, J. (eds.), Multidisciplinary Approaches to Coastal and Marine Management. Journal of Coastal Research, Special Issue No. 116, pp. 240-244. Charlotte (North Carolina), ISSN 0749-0208. Research on rip currents can be conducted through indoor studies with numerical simulation tests or by studies that analyze data by installing instruments in the actual ocean. For flow field studies using measuring instruments, data are collected by installing an Acoustic Wave Current Measurement (AWAC) or Acoustic Doppler Current Profiler (ADCP). It is not appropriate in terms of effectiveness, cost, and accuracy to analyze data obtained in a specific area by applying it to the entire flow field. In this study, the applicability and limitations of the Large Scale Particle Image Velocimetry (LSPIV) technique were derived for Closed-circuit Television (CCTV) images of rip currents in Haeundae by applying the optimal velocity calculation parameters. For LSPIV analysis, the tilted CCTV image was tilt-corrected using 208 GCPs and Affine Transformation and stabilized using feature points. The flow rate calculated with 18 parameters by combining 6 interrogation areas and 3 searching areas was compared with the average flow rate of the normal distribution confidence interval. As the size of the parameter changes, the error rate changes along with its characteristics. Compared to the average flow rate, three parameters with errors within 2% were determined as LSPIV flow rate calculation parameters. The flow velocities calculated with the determined parameters were 0.196 m/sec, 0.205 m/sec, and 0.207 m/sec within a maximum range of 5.31%. In addition, it was found that the flow velocity increased, and the error decreased as the size of the interrogation area increased. The LSPIV velocity vector overlapped with the rip current's foam image in the tilt-corrected CCTV image to confirm the degree of matching. In the future, if several high-resolution CCTVs are used and stereo image-based object detection and 3D location tracking algorithms are applied to the analysis area, it can be a good alternative as a direct measurement analysis verification method of marine data.