This is on one hand because of changing conditions during image acquisition even within one dive (changing light conditions from different altitude variable backscatter of particles in the water over time) and between dives (potentially different light and camera configurations), but also because domain expertise is needed to annotate a large number of training examples for fitting an image classification model. Despite this benefit, the implementation of automated workflows is still challenging. This is because manually inspecting each photo is infeasible and non-scalable 2 and thus, automated seafloor classification workflows are preferred because of the benefit of repeatability, which also reduces subjectivity and bias 3. ![]() ![]() Such cameras generate a high-volume of seafloor images in number and storage space, whose analysis requires automated workflows. The cameras recording these images are usually attached to platforms such as towed camera frames as in the Ocean Floor Observation System (OFOS), Automated Underwater Vehicles (AUVs) and Remotely Operated Vehicles (ROVs). Together with bathymetric information, underwater optical images are typically used to characterize the seafloor by partitioning it into classes, and assigning the relevant semantic seafloor- or habitat-label to each class 1. Understanding the geological and ecological characteristics of the seafloor is key for monitoring and managing marine ecosystems. On the other hand, the German area primarily comprises nodules that only partly cover the seabed, and these occur alongside turned-over sediment (artificial seafloor) that were caused by the settling plume following a dredging experiment conducted in the area. Our results show that the seafloor in the Belgian area predominantly comprises densely distributed nodules, which are intermingled with qualitatively larger-sized nodules at local elevations and within depressions. Based on this, we provide seafloor classifications along the camera deployment tracks, and discuss results in the context of seafloor multibeam bathymetry. As a case study, we applied the workflow to an example seafloor image dataset from the Belgian and German contract areas for Manganese-nodule exploration in the Pacific Ocean. It further includes semi-automatic generation of the training data set for fitting the seafloor classifier. AI-SCW incorporates laser point detection for scale determination and color normalization. The workflow aims to classify the seafloor into habitat categories based on automated analysis of optical underwater images with only minimal amount of human annotations. Here, we present a generic implementation of an Automated and Integrated Seafloor Classification Workflow (AI-SCW). However, in order to provide consistent and repeatable analysis, these automated workflows need to address e.g., underwater illumination artefacts, variances in resolution and class-imbalances, which could bias the classification. Automated workflows have been proposed as a solution, in which algorithms assign pre-defined seafloor categories to each image. With the increasing capabilities to record high-resolution underwater images, manual approaches for analyzing these images to create seafloor classifications are no longer feasible. ![]() Habitat mapping relies on seafloor classification typically based on acoustic methods, and ground truthing through direct sampling and optical imaging. Mapping and monitoring of seafloor habitats are key tasks for fully understanding ocean ecosystems and resilience, which contributes towards sustainable use of ocean resources.
0 Comments
Leave a Reply. |
AuthorWrite something about yourself. No need to be fancy, just an overview. ArchivesCategories |