Figure 3
From: Deep neural networks for active wave breaking classification
Example of the application of the method. (a) Results of the naïve wave breaking detection (thresholding + DBSCAN) for La Jument data (03/01/2018 09:39). Note the great amount of passive foam being detected as active wave breaking. (b) Results of active wave breaking detection using VGG16 as backbone for La Jument data (03/01/2018 09:39). Note the significant reduction in the amount of passive foam being detected. In both plots, number of clusters refers to the number of clusters identified by DBSCAN. The red dashed rectangle indicates the region of interest which, in these examples, was the same as the stereo-video reconstruction area. (c) Results of the naïve wave breaking detection (thresholding + DBSCAN) for Black sea data (04/10/2011 09:38). (d) Results of active wave breaking detection using VGG16 as backbone for Black Sea data (04/10/2011 09:38). Animations of these results are available at https://github.com/caiostringari/deepwaves. In this particular example, the image was subdivided into blocks of \(256\times 256\) pixels for processing. Note that identical results were seen using other architectures other than VGG16 to classify these data.
