-
Sea ice detection using concurrent multispectral and synthetic aperture radar imagery
Authors:
Martin S J Rogers,
Maria Fox,
Andrew Fleming,
Louisa van Zeeland,
Jeremy Wilkinson,
J. Scott Hosking
Abstract:
Synthetic Aperture Radar (SAR) imagery is the primary data type used for sea ice map** due to its spatio-temporal coverage and the ability to detect sea ice independent of cloud and lighting conditions. Automatic sea ice detection using SAR imagery remains problematic due to the presence of ambiguous signal and noise within the image. Conversely, ice and water are easily distinguishable using mu…
▽ More
Synthetic Aperture Radar (SAR) imagery is the primary data type used for sea ice map** due to its spatio-temporal coverage and the ability to detect sea ice independent of cloud and lighting conditions. Automatic sea ice detection using SAR imagery remains problematic due to the presence of ambiguous signal and noise within the image. Conversely, ice and water are easily distinguishable using multispectral imagery (MSI), but in the polar regions the ocean's surface is often occluded by cloud or the sun may not appear above the horizon for many months. To address some of these limitations, this paper proposes a new tool trained using concurrent multispectral Visible and SAR imagery for sea Ice Detection (ViSual\_IceD). ViSual\_IceD is a convolution neural network (CNN) that builds on the classic U-Net architecture by containing two parallel encoder stages, enabling the fusion and concatenation of MSI and SAR imagery containing different spatial resolutions. The performance of ViSual\_IceD is compared with U-Net models trained using concatenated MSI and SAR imagery as well as models trained exclusively on MSI or SAR imagery. ViSual\_IceD outperforms the other networks, with a F1 score 1.60\% points higher than the next best network, and results indicate that ViSual\_IceD is selective in the image type it uses during image segmentation. Outputs from ViSual\_IceD are compared to sea ice concentration products derived from the AMSR2 Passive Microwave (PMW) sensor. Results highlight how ViSual\_IceD is a useful tool to use in conjunction with PMW data, particularly in coastal regions. As the spatial-temporal coverage of MSI and SAR imagery continues to increase, ViSual\_IceD provides a new opportunity for robust, accurate sea ice coverage detection in polar regions.
△ Less
Submitted 11 January, 2024;
originally announced January 2024.
-
Design of EMG-driven Musculoskeletal Model for Volitional Control of a Robotic Ankle Prosthesis
Authors:
Chinmay Shah,
Aaron Fleming,
Varun Nalam,
He,
Huang
Abstract:
Existing robotic lower-limb prostheses use autonomous control to address cyclic, locomotive tasks, but they are inadequate to operate the prosthesis for daily activities that are non-cyclic and unpredictable. To address this challenge, this study aims to design a novel electromyography (EMG)-driven musculoskeletal model for volitional control of a robotic ankle-foot prosthesis. This controller pla…
▽ More
Existing robotic lower-limb prostheses use autonomous control to address cyclic, locomotive tasks, but they are inadequate to operate the prosthesis for daily activities that are non-cyclic and unpredictable. To address this challenge, this study aims to design a novel electromyography (EMG)-driven musculoskeletal model for volitional control of a robotic ankle-foot prosthesis. This controller places the user in continuous control of the device, allowing them to freely manipulate the prosthesis behavior at will. The Hill-type muscle model was used to model a dorsiflexor and a plantarflexor, which functioned around a virtual ankle joint. The model parameters were determined by fitting the model prediction to the experimental data collected from an able-bodied subject. EMG signals recorded from ankle agonist and antagonist muscle pair were used to activate the virtual muscle models. This model was validated via offline simulations and real-time prosthesis control. Additionally, the feasibility of the proposed prosthesis control on assisting the user's functional tasks was demonstrated. The present control may further improve the function of robotic prosthesis for supporting versatile activities in individuals with lower-limb amputations.
△ Less
Submitted 17 February, 2022;
originally announced February 2022.
-
Optic disc and fovea localisation in ultra-widefield scanning laser ophthalmoscope images captured in multiple modalities
Authors:
Peter Robert Wakeford,
Enrico Pellegrini,
Gavin Robertson,
Michael Verhoek,
Alan Duncan Fleming,
Jano van Hemert,
Ik Siong Heng
Abstract:
We propose a convolutional neural network for localising the centres of the optic disc (OD) and fovea in ultra-wide field of view scanning laser ophthalmoscope (UWFoV-SLO) images of the retina. Images captured in both reflectance and autofluorescence (AF) modes, and central pole and eyesteered gazes, were used. The method achieved an OD localisation accuracy of 99.4% within one OD radius, and fove…
▽ More
We propose a convolutional neural network for localising the centres of the optic disc (OD) and fovea in ultra-wide field of view scanning laser ophthalmoscope (UWFoV-SLO) images of the retina. Images captured in both reflectance and autofluorescence (AF) modes, and central pole and eyesteered gazes, were used. The method achieved an OD localisation accuracy of 99.4% within one OD radius, and fovea localisation accuracy of 99.1% within one OD radius on a test set comprising of 1790 images. The performance of fovea localisation in AF images was comparable to the variation between human annotators at this task. The laterality of the image (whether the image is of the left or right eye) was inferred from the OD and fovea coordinates with an accuracy of 99.9%
△ Less
Submitted 24 April, 2020;
originally announced April 2020.
