Wave-by-wave nearshore wave breaking identification using U-Net

Sáez, Francisco Javier Llovera; Catalán, Patricio A.; Valle, Carlos

doi:10.1016/j.coastaleng.2021.104021

Cited by 18 publications

(6 citation statements)

References 22 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…In the five years since CIRN's inception, the use of machine learning algorithms to exploit coastal imagery has grown. Data-driven approaches have been used to extract nearshore parameters from imagery including bathymetry [78,79], wave heights [80], wave breaking type [81] and occurrence [82][83][84][85], shoreline position [86], and land cover [87]; to classify nearshore morphology [88]; and to identify dangerous flows, like rip currents [89]. These data-driven approaches can increase analysis capabilities, enabling rapid extraction of data and improving ease of use in engineering and science applications.…”

Section: Technological Advancements and Challengesmentioning

confidence: 99%

“…For example, 3-dimensional observations of individual breaking waves using lidar have revealed new information on the complex kinematics and geometry of spilling and plunging breakers and propagating bores [29,30,92,93]. Fusion between data sources will likely lead to an improved resolution of processes and better parameterization of those processes in numerical models [29,85]. A multi-modal approach camera calibration has been shown to be effective as well [27].…”

Section: Understanding Physical Processesmentioning

confidence: 99%

See 1 more Smart Citation

The Coastal Imaging Research Network (CIRN)

Palmsten

Brodie

2022

Remote Sensing

View full text Add to dashboard Cite

The Coastal Imaging Research Network (CIRN) is an international group of researchers who exploit signatures of phenomena in imagery of coastal, estuarine, and riverine environments. CIRN participants develop and implement new coastal imaging methodologies. The research objective of the group is to use imagery to gain a better fundamental understanding of the processes shaping those environments. Coastal imaging data may also be used to derive inputs for model boundary and initial conditions through assimilation, to validate models, and to make management decisions. CIRN was officially formed in 2016 to provide an integrative, multi-institutional group to collaborate on remotely sensed data techniques. As of 2021, the network is a collaboration between researchers from approximately 16 countries and includes investigators from universities, government laboratories and agencies, non-profits, and private companies. CIRN has a strong emphasis on education, exemplified by hosting annual “boot camps” to teach photogrammetry fundamentals and toolboxes from the CIRN code repository, as well as hosting an annual meeting for its members to present coastal imaging research. In this review article, we provide context for the development of CIRN as well as describe the goals and accomplishments of the CIRN community. We highlight components of CIRN’s resources for researchers worldwide including an open-source GitHub repository and coding boot camps. Finally, we provide CIRN’s perspective on the future of coastal imaging.

show abstract

Section: Technological Advancements and Challengesmentioning

confidence: 99%

Section: Understanding Physical Processesmentioning

confidence: 99%

The Coastal Imaging Research Network (CIRN)

Palmsten

Brodie

2022

Remote Sensing

View full text Add to dashboard Cite

show abstract

“…Thus, the minimal resolution was often too low to classify features smaller than the size of tiles and superpixels. Furthermore, Sáez et al (2021) used an U-net architecture for detecting wave-breaking nearshore while other studies had tried to validate semantic segmentation by comparing with other measurement methods. For example, by comparing the results with gauges in a physical model test, den Bieman et al (2020) showed that image segmentation by SegNet can reasonably predict surface elevation, run-up, and bed level from video images.…”

Section: Introductionmentioning

confidence: 99%

Stochastic properties of coastal flooding events – Part 1: CNN-based semantic segmentation for water detection

Kang

Feagin

Huff

et al. 2023

Preprint

View full text Add to dashboard Cite

Abstract. The frequency and intensity of coastal flooding is expected to accelerate in low-elevation coastal areas due to sea level rise. Coastal flooding due to wave runup affects coastal ecosystems and infrastructure, however it can be difficult to monitor in remote and vulnerable areas. Here we use a camera-based system to monitor wave runup as part of the after-storm recovery of an eroded beach on the Texas coast. We analyze high-temporal resolution images of the beach using Convolutional Neural Network (CNN)-based semantic segmentation to study the stochastic properties of runup-driven flooding events. In the first part of this work, we focus on the application of semantic segmentation to identify water and runup events. We train and validate a CNN with over 500 manually classified images, and introduce a post-processing method to reduce false positives. We find that the accuracy of CNN predictions of water pixels is around 90% and strongly depend on the number and diversity of images used for training.

show abstract

“…Image segmentation has become an increasingly important tool in Earth science research (Pally & Samadi, 2022; Sun et al., 2022; Yuan et al., 2021). In recent years, Deep Learning (LeCun et al., 2015) models based on the UNet (Ronneberger et al., 2015) and the Residual UNet (Liu et al., 2019; Zhang et al., 2018) have become the standard in state‐of‐the‐art Earth science applications involving image segmentation (Chen et al., 2020; Collins et al., 2020; Gupta et al., 2021; Hoeser & Kuenzer, 2020; Jin et al., 2022; Kattenborn et al., 2019; Kotaridis & Lazaridou, 2021; Li et al., 2022; Liu et al., 2019; Marangio et al., 2020; Nagi et al., 2021; Nalepa et al., 2019; Rafique et al., 2022; Sáez et al., 2021; Song et al., 2020; van der Meij et al., 2021; Verma et al., 2021; Xiao et al., 2021).…”

Section: Introductionmentioning

confidence: 99%

A Reproducible and Reusable Pipeline for Segmentation of Geoscientific Imagery

Buscombe

Goldstein

2022

Earth and Space Science

View full text Add to dashboard Cite

Segmentation of Earth science imagery is an increasingly common task. Among modern techniques that use Deep Learning, the UNet architecture has been shown to be a reliable for segmenting a range of imagery. We developed software–Segmentation Gym–to implement a data‐model pipeline for segmentation of scientific imagery using a family of UNet models. With an existing set of imagery and labels, the software uses a single configuration file that handles data set creation, as well as model setup and model training. Key benefits of this software are (a) the focus on reproducible data set creation and modeling, and (b) the ability for quick model experimentation through changes to a configuration file. Quick experimentation permits researchers to prototype different model architectures, sizes, and adjust common hyperparameters to find a suitable model. We demonstrate the use of the software using a data set of 419 labeled Landsat‐8 scenes of coastal environments and compare results across two model architectures, five model sizes, and three loss functions. This demonstration highlights that our software enables rapid, reproducible experimentation to determine optimal hyperparameters for specific data sets and research questions.

show abstract

Wave-by-wave nearshore wave breaking identification using U-Net

Cited by 18 publications

References 22 publications

The Coastal Imaging Research Network (CIRN)

The Coastal Imaging Research Network (CIRN)

Stochastic properties of coastal flooding events – Part 1: CNN-based semantic segmentation for water detection

A Reproducible and Reusable Pipeline for Segmentation of Geoscientific Imagery

Contact Info

Product

Resources

About