20 Aug 2025

SandSnap Filtering Techniques

Abstract: The aim of this Coastal and Hydraulics Laboratory Special Report is to elucidate the new SandSnap image filters. These SandSnap filters distinguish between high-quality and poor-quality images and enhance accuracy in high-quality images. To achieve this goal, a dataset of 5,000 photos was created and curated for this endeavor. Images were collected that had varying levels of focus, sedimentological conditions, foreign objects present, distances from the sediment bed, coin types, and geographic locations. This dataset was used to train multiple quality control check models and uncover beneficial correlations. Additionally, an existing dataset of high-quality images was analyzed using various filtering techniques to highlight key features, leading to higher-accuracy scores. Using the findings from both the high-quality and poor-quality datasets, SandSnap was updated to increase usability and efficiently identify images that may lead to poor results. This ensures that user results can be calculated in less than a minute, emphasizing the commitment to maintaining a fast and responsive model.

8 Sep 2023

Applicability of CoastSnap, a Crowd-Sourced Coastal Monitoring Approach for US Army Corps of Engineers District Use

Abstract: This US Army Engineer Research and Development Center, Coastal and Hydraulics Laboratory, technical report details the pilot deployment, accuracy evaluation, and best practices of the citizen-science, coastal-image monitoring program CoastSnap. Despite the need for regular observational data, many coastlines are monitored infrequently due to cost and personnel, and this cell phone-image-based approach represents a new potential data source to districts in addition to providing an outreach opportunity for the public. Requiring minimal hardware and signage, the system is simple to install but requires user-image processing. Analysis shows the CoastSnap-derived shorelines compare well to real-time kinematic and lidar-derived shorelines during low-to-moderate wave conditions (root mean square errors [RMSEs] <10 m). During high-wave conditions, errors are higher (RMSE up to 18 m) but are improved when incorporating wave run-up. Beyond shoreline quantification, images provide other qualitative information such as storm-impact characteristics and timing of the formation of beach scarps. Ultimately, the citizen-science tool is a viable low-cost option to districts for monitoring shorelines and tracking the evolution of coastal projects such as beach nourishments.

13 May 2020

PUBLICATION NOTICE: Technical Feasibility of Creating a Beach Grain Size Database with Citizen Scientists

ABSTRACT:  The goal of this Coastal and Hydraulics Engineering Technical Note (CHETN) is to investigate the feasibility of collecting beach grain size information with images collected by citizen scientists to build a globally accessible database. Engaging citizen scientists in scientific information collection through crowdsourcing has become a more popular and cost-effective way to collect large amounts of data while increasing interest in the research through public engagement (Irwin 2018). Citizen scientists equipped with their personal smartphones allow for very large datasets to be collected that would otherwise be financially or logistically impossible. Additionally, it provides an opportunity to educate and engage the general public.