Synthetic Imagery for the Automated Detection of Rip Currents

Pitman, Sebastian J.; Gallop, Shari L.; Haigh, Ivan D.; Mahmoodi, Sasan; Masselink, Gerd; Ranasinghe, Roshanka

doi:10.2112/si75-183.1

Cited by 11 publications

(6 citation statements)

References 9 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…These data support the validation of a rip current forecast model to alert people to potential hazards [13]. One method that has promise for visually detecting rip currents is the use of "time exposure" or Timex images [31] [32] [33] [34]. These are obtained by simply averaging frames of a video clip, usually over a period of 10 minute intervals.…”

mentioning

confidence: 77%

Flow-Based Rip Current Detection and Visualization

et al. 2022

View full text Add to dashboard Cite

Rip currents are like rivers of fast-moving water that can quickly carry the unwary out to sea. They are not easy to recognize, especially to the untrained observer. Other than in-situ current measurements, there exists a number of methods that analyzes images and videos to detect rip currents. Most of these techniques base their detection on the appearance of rip currents such as foamy pattern, discoloration of water, and locations of breaking waves. Leading methods use either image processing or machine learning of images and/or video input. In this paper, we analyze the behavior of water movement rather than simply its appearance to detect rip currents. Specifically, we investigated several flow visualization methods and tune them to detect rip currents. Based on our study, we recommend two methods that allowed us to detect rip currents where other methods have failed. And because the methods originated as visualization techniques, any presence of rip currents are automatically highlighted. We also evaluated these two methods against previously annotated results by rip current experts, and found that our detections were sufficiently sensitive that some expert annotations were relabeled.

show abstract

mentioning

confidence: 77%

Flow-Based Rip Current Detection and Visualization

et al. 2022

View full text Add to dashboard Cite

show abstract

“…They modified the color matching method for optical flow to detect rip currents. Pitman et al (2016) [5] developed a method for automatic rip current detection based on synthetic images. They segmented the "original" images into essential sections, including sand bars, estuaries, coastlines, and the ocean, which are required for the generation of a synthetic image.…”

Section: Literature Reviewmentioning

confidence: 99%

Rip Current: A Potential Hazard Zones Detection in Saint Martin’s Island using Machine Learning Approach

Islam

Shampa²

2023

ELCVIA

View full text Add to dashboard Cite

Beach hazards would be any occurrences potentially endanger individuals aswell as their activity. Rip current, or reverse current of the sea, is a typeof wave that pushes against the shore and moves in the opposite direction,that is, towards the deep sea. The management of access to the beach sometimes accidentally push unwary beachgoers forward into rip-prone regions,increasing the probability of a drowning on that beach. The research suggestsan approach for something like the automatic detection of rip currents withwaves crashing based on convolutional neural networks (CNN) and machinelearning algorithms (MLAs) for classification. Several individuals are unableto identify rip currents in order to prevent them. In addition, the absenceof evidence to aid in training and validating hazardous systems hinders attempts to predict rip currents. Security cameras and mobile phones have stillimages of something like the shore pervasive and represent a possible causeof rip current measurements and management to handle this hazards accordingly. This work deals with developing detection systems from still beachimages, bathymetric images, and beach parameters using CNN and MLAs.The detection model based on CNN for the input features of beach imagesand bathymetric images has been implemented. MLAs have been applied todetect rip currents based on beach parameters. When compared to other detection models, bathymetric image-based detection models have significantlyhigher accuracy and precision. The VGG16 model of CNN shows maximumaccuracy of 91.13% (Recall = 0.94, F1-score = 0.87) for beach images. Forthe bathymetric images, the highest performance has been found with anaccuracy of 96.89% (Recall= 0.97, F1-score=0.92) for the DenseNet model of CNN. The MLA-based model shows an accuracy of 86.98% (Recall=0.89,F1-score= 0.90) for random forest classifier. Once we know about the potential zone of rip current continuosly generating rip current, then the coastalregion can be managed accordingly to prevent the accidents occured due tothis coastal hazards.

show abstract

“…Time averaged images are a routine method for analyzing video in oceanic research, with 10 minutes being a common integration period [27,28,38,47]. This method is popular because averages often make identification of rip channels easier for the human eye.…”

Section: Related Workmentioning

confidence: 99%

Automated Rip Current Detection with Region based Convolutional Neural Networks

Silva

Mori

Dusek

et al. 2021

Preprint

View full text Add to dashboard Cite

This paper presents a machine learning approach for the automatic identification of rip currents with breaking waves. Rip currents are dangerous fast moving currents of water that result in many deaths by sweeping people out to sea. Most people do not know how to recognize rip currents in order to avoid them. Furthermore, efforts to forecast rip currents are hindered by lack of observations to help train and validate hazard models. The presence of web cams and smart phones have made video and still imagery of the coast ubiquitous and provide a potential source of rip current observations. These same devices could aid public awareness of the presence of rip currents. What is lacking is a method to detect the presence or absence of rip currents from coastal imagery. This paper provides expert labeled training and test data sets for rip currents. We use Faster-RCNN and a custom temporal aggregation stage to make detections from still images or videos with higher measured accuracy than both humans and other methods of rip current detection previously reported in the literature.

show abstract

Synthetic Imagery for the Automated Detection of Rip Currents

Cited by 11 publications

References 9 publications

Flow-Based Rip Current Detection and Visualization

Flow-Based Rip Current Detection and Visualization

Rip Current: A Potential Hazard Zones Detection in Saint Martin’s Island using Machine Learning Approach

Automated Rip Current Detection with Region based Convolutional Neural Networks

Contact Info

Product

Resources

About