- Marine and fisheries research
- Water Quality Monitoring Technologies
- Fish Ecology and Management Studies
- Marine animal studies overview
- Underwater Vehicles and Communication Systems
- Coral and Marine Ecosystems Studies
- Image Enhancement Techniques
- Video Surveillance and Tracking Methods
- Marine Bivalve and Aquaculture Studies
- Underwater Acoustics Research
- Advanced Image and Video Retrieval Techniques
- Identification and Quantification in Food
- Robotics and Sensor-Based Localization
- Ichthyology and Marine Biology
- Cephalopods and Marine Biology
- Food Industry and Aquatic Biology
- Fish biology, ecology, and behavior
- Ergonomics and Musculoskeletal Disorders
- Occupational health in dentistry
- Anomaly Detection Techniques and Applications
- Digital Imaging for Blood Diseases
- Marine Biology and Ecology Research
- Marine Biology and Environmental Chemistry
- Arctic and Antarctic ice dynamics
- Target Tracking and Data Fusion in Sensor Networks
NOAA National Marine Fisheries Service Alaska Fisheries Science Center
2014-2024
NOAA National Marine Fisheries Service
2014-2024
National Oceanic and Atmospheric Administration
2013-2024
University of Washington
2010-2016
Live fish recognition is one of the most crucial elements fisheries survey applications where vast amount data rapidly acquired. Different from general scenarios, challenges to underwater image are posted by poor quality, uncontrolled objects and environment, difficulty in acquiring representative samples. In addition, existing feature extraction techniques hindered automation due involving human supervision. Toward this end, we propose an framework that consists a fully unsupervised...
Non-extractive fish abundance estimation with the aid of visual analysis has drawn increasing attention. Unstable illumination, ubiquitous noise and low frame rate video capturing in underwater environment, however, make conventional tracking methods unreliable. In this paper, we present a multiple system for low-contrast low-frame-rate stereo videos use trawl-based camera system. An automatic segmentation algorithm overcomes issues by adopting histogram backprojection approach on double...
Fishery surveys that call for the use of single or multiple underwater cameras have been an emerging technology as a nonextractive mean to estimate abundance fish stocks. Tracking live in open aquatic environment posts challenges are different from general pedestrian vehicle tracking surveillance applications. In many rough habitats, monitored by installed on moving platforms, where is even more challenging due inapplicability background models. this paper, novel algorithm based deformable...
Abstract The standard allometric weight‐length relationship W = aL b is widely used in fisheries science to estimate the weight of fish known length and compute body condition indices. This abundance surveys such as acoustic convert at into estimates population biomass. Although fitting this data over a broad range sizes common practice, fit often does not receive careful scrutiny. We explored model well alternative relationships from walleye pollock Theragra chalcogramma North Pacific found...
A distinct body of literature supports the association between clinical postures dental practitioner and work-related musculoskeletal disorders (WRMD). Several aids or devices have been tested to improve posture in interest decreasing WRMD. The use magnification lenses while performing procedures may increase quality work decrease likelihood problems. To date, only anecdotal personal opinions had existed regarding benefits using lenses, no empirical evidence authenticated contention that...
Abstract Handegard, N. O., and Williams, K. 2008. Automated tracking of fish in trawls using the DIDSON (Dual frequency IDentification SONar). – ICES Journal Marine Science. 65: 636–644. An application for automated dual-frequency, identification sonar (DIDSON) data was developed tested on observations taken midwater trawls. The process incorporates target detection, multiple tracking, extraction behaviour information such as speed direction from track data. automatic tracker evaluated three...
Abstract Climate-driven changes in the timing of spawning or migration can affect availability fish to surveys designed monitor their abundance, complicating efforts assess stock status and sustainably manage fisheries. From 2017 2019, trends biomass estimates from four used Gulf Alaska pollock diverged. These conflicting increased uncertainty assessment occurred during a time rapid environmental change. We hypothesized that spawn affected winter survey targets pre-spawning aggregations. To...
This paper describes an automatic segmentation algorithm for fish sampled by a trawl-based underwater camera system. To overcome the problem caused very low brightness contrast between and their background with dynamically changing luminance, our proposed adopts innovative histogram backprojection procedure on double local-thresholded images to ensure reliable shape boundaries. The thresholded results are further validated area variance criteria remove unwanted objects. Finally,...
To understand and manage marine ecosystems, long-term monitoring of fish biomass is needed. A challenge in estimating using sampling nets varying catchability with habitat, weather, or vessel traffic conditions. Underwater stereo cameras have shown promise providing a non-lethal, efficient, cost-effective method to observe measure areas that cannot be sampled otherwise. These methods, however, yet demonstrated for mid-water pelagic semi-pelagic fishes. We designed, built, tested camera its...
Automated fish species identification in open aquatic habitats based on video analytics is the primary area of research camera-based fisheries surveys. Finding informative features for these analyses, however, fundamentally challenging due to poor quality underwater imagery and strong visual similarity among species. In this paper, we compare two different feature extraction methods, namely supervised unsupervised approaches, which are then applied a hierarchical partial classification...
Abstract Diel vertical migration (DVM) in marine organisms is an evolved response to maximize foraging opportunities and minimize predation risk. This pattern distribution a widely observed phenomenon common pelagic forage fish species. Modelling these dynamics has important relevance understanding ecosystem dynamics, predator–prey interactions, food webs. We examine DVM sand lance—an ecologically taxa of fishes throughout northern hemisphere systems target commercial fisheries. Much remains...
Global fisheries and the future of sustainable seafood are predicated on healthy populations various species fish shellfish. Recent developments in collection large-volume optical data by autonomous underwater vehicles (AUVs), stationary camera arrays, towed has made it possible for fishery scientists to generate species-specific, size-structured abundance estimates different marine organisms via imagery. The immense volume collected such devices quickly exceeds manual processing capacity...
Abstract Williams, K., Punt, A. E., Wilson, C. D., and Horne, J. K. 2011. Length-selective retention of walleye pollock, Theragra chalcogramma, by midwater trawls. – ICES Journal Marine Science, 68: 119–129. Midwater trawls are commonly used during acoustic surveys fish abundance to determine species length compositions acoustically sampled aggregations. As selective samplers, catches can be unrepresentative populations lead biased estimates. Length-dependent pollock was estimated using...
Live fish recognition in open aquatic habitats suffers from the high uncertainty many of data. To alleviate this problem without discarding those data, system should learn a species hierarchy so that high-level labels can be assigned to ambiguous In paper, systematic hierarchical partial classification algorithm is therefore proposed for underwater recognition. Partial applied at each level coarse-to-fine categorization stops once decision confidence low. By defining exponential benefit...
Fish abundance estimation with the aid of visual analysis has drawn increasing attention based on underwater videos from a remotely-operated vehicle (ROV). We build novel fish tracking and counting system followed by tracking-by-detection framework. Since may keep entering or leaving field view (FOV), an offline trained deformable part model (DPM) detector is adopted to detect live video data. Besides that, multiple kernel approach used associate same object across consecutive frames for...
Abstract Spatial management of vulnerable benthic ecosystem components such as deep-sea corals and sponges requires adequate maps their distribution. These are often based on statistical models survey data. The objective this project was to validate the predictions existing presence or absence abundance in Aleutian Islands that were bottom trawl Model validation conducted by comparing model observations an situ camera at randomly selected locations. measures goodness fit...