Alluvial substrate mapping by automated texture segmentation of recreational-grade side scan sonar imagery

Abstract

Side scan sonar in low-cost ‘fishfinder’ systems has become popular in aquatic ecology and sedimentology for imaging submerged riverbed sediment at coverages and resolutions sufficient to relate bed texture to grain-size. Traditional methods to map bed texture (i.e. physical samples) are relatively high-cost and low spatial coverage compared to sonar, which can continuously image several kilometers of channel in a few hours. Towards a goal of automating the classification of bed habitat features, we investigate relationships between substrates and statistical descriptors of bed textures in side scan sonar echograms of alluvial deposits. We develop a method for automated segmentation of bed textures into between two to five grain-size classes. Second-order texture statistics are used in conjunction with a Gaussian Mixture Model to classify the heterogeneous bed into small homogeneous patches of sand, gravel, and boulders with an average accuracy of 80%, 49%, and 61%, respectively. Reach-averaged proportions of these sediment types were within 3% compared to similar maps derived from multibeam sonar.

A review of seagrass detection, mapping and monitoring applications using acoustic systems

“…Finally, coherent scales between backscatter and topography are analysed using cospectra. Hamill et al (2018) used linear least squares and GMM (2 models) for automated segmentation of bed textures using recreational-grade SSS. In this study, SSS imagery is used to derive first order statics and GLCM textures.…”

“…Sabol et al, 2014). In recent years, recreational grade side sonar has become popular for substrate mapping studies in rivers (Buscombe, 2017;Hamill, Buscombe, & Wheaton, 2018). This low-cost technology has also been applied for measuring seagrass cover in shallow environments by Greene, Rahman, Kline, and Rahman (2018).…”

Assessment of a Low-Cost Side-Scan Sonar for River Estuary Underwater Imaging

“…However, the recording option was chosen during the survey, and the Humminbird unit recorded the data continuously using a micro SD storage card. This method was also used by (Hamill, Buscombe, & Wheaton, 2018) in their study of river substrate mapping. In addition, visual examination of the Humminbird recording screen was also conducted during the survey.…”

Probabilistic Substrate Classification with Multispectral Acoustic Backscatter: A Comparison of Discriminative and Generative Models

“…Instead of using Equation (13) to find the posterior probabilities of each label, one may model the joint distribution of x and y together, P(x, y), and then normalize to obtain posterior probabilities. An example of one such approach is the Gaussian Mixture Model (GMM) [6,7,24,38], which is a weighted sum of M component Gaussian probability density functions, g, with unknown parameters, given by (e.g., [7])…”