scholarly journals A Deep Learning Approach to Assist Sustainability of Demersal Trawling Operations

2021 ◽  
Vol 13 (22) ◽  
pp. 12362
Author(s):  
Maria Sokolova ◽  
Adrià Mompó Alepuz ◽  
Fletcher Thompson ◽  
Patrizio Mariani ◽  
Roberto Galeazzi ◽  
...  
Keyword(s):  
Absolute Error ◽  
Monitoring Tool ◽  
Video Footage ◽  
Trawl Fishery ◽  
Augmentation Techniques ◽  
Image Acquisition System ◽  
Catch Composition ◽  
First Time ◽  
Trawl Fisheries ◽  
Resolution Dataset

Bycatch in demersal trawl fisheries challenges their sustainability despite the implementation of the various gear technical regulations. A step towards extended control over the catch process can be established through a real-time catch monitoring tool that will allow fishers to react to unwanted catch compositions. In this study, for the first time in the commercial demersal trawl fishery sector, we introduce an automated catch description that leverages state-of-the-art region based convolutional neural network (Mask R-CNN) architecture and builds upon an in-trawl novel image acquisition system. The system is optimized for applications in Nephrops fishery and enables the classification and count of catch items during fishing operation. The detector robustness was improved with augmentation techniques applied during training on a custom high-resolution dataset obtained during extensive demersal trawling. The resulting algorithms were tested on video footage representing both the normal towing process and haul-back conditions. The algorithm obtained an F-score of 0.79. The resulting automated catch description was compared with the manual catch count showing low absolute error during towing. Current practices in demersal trawl fisheries are carried out without any indications of catch composition nor whether the catch enters the fishing gear. Hence, the proposed solution provides a substantial technical contribution to making this type of fishery more targeted, paving the way to further optimization of fishing activities aiming at increasing target catch while reducing unwanted bycatch.

scholarly journals Combining in-trawl video with observer coverage improves understanding of protected and vulnerable species by-catch in trawl fisheries

2014 ◽  
Vol 65 (9) ◽  
pp. 830 ◽  
Author(s):  
Vanessa F. Jaiteh ◽  
Simon J. Allen ◽  
Jessica J. Meeuwig ◽  
Neil R. Loneragan
Keyword(s):  
Video Camera ◽  
Fishing Effort ◽  
Vulnerable Species ◽  
Independent Observer ◽  
Video Footage ◽  
Trawl Fishery ◽  
Scalloped Hammerhead ◽  
Sea Snakes ◽  
By Catch ◽  
Trawl Fisheries

Assessments of incidental wildlife mortality resulting from fishing rarely account for unobserved by-catch. We assessed by-catch of protected and vulnerable wildlife species in an Australian trawl fishery by comparing in-trawl video footage with data collected by an on-board observer. Data were obtained from 44 commercial trawls with two different by-catch reduction devices (BRDs). Eighty-six individuals from six major taxa (dolphins, sharks, rays, sea snakes, turtles and sygnathids) were documented from video analysis, including the endangered scalloped hammerhead shark (Sphyrna lewini) and the critically endangered green sawfish (Pristis zijsron). On the basis of the 2008–2009 fishing effort of 4149 trawls and scaling from these results, we estimated the annual catch of protected and vulnerable species (± 1 s.e.) at 8109 ± 910 individuals. Only 34% of by-catch was expelled through the BRDs. Independent observer data for the 44 trawls showed that 77% of the landed by-catch from these taxa were dead when discarded. The results indicate that unaccounted by-catch in trawl fisheries can be substantial, and that current methods of recording by-catch on-board vessels are likely to underestimate total fishing mortality. We recommend gear modifications and their validation through dedicated observer coverage, combined with in-trawl video camera deployments to improve current approaches to by-catch mitigation.

scholarly journals Generative Adversarial Networks to Improve the Robustness of Visual Defect Segmentation by Semantic Networks in Manufacturing Components

2021 ◽  
Vol 11 (14) ◽  
pp. 6368
Author(s):  
Fátima A. Saiz ◽  
Garazi Alfaro ◽  
Iñigo Barandiaran ◽  
Manuel Graña
Keyword(s):  
Ad Hoc ◽  
Data Augmentation ◽  
Semantic Network ◽  
Semantic Networks ◽  
Stereo Image ◽  
Generative Adversarial Networks ◽  
Specific Class ◽  
Adversarial Networks ◽  
Augmentation Techniques ◽  
Image Acquisition System

This paper describes the application of Semantic Networks for the detection of defects in images of metallic manufactured components in a situation where the number of available samples of defects is small, which is rather common in real practical environments. In order to overcome this shortage of data, the common approach is to use conventional data augmentation techniques. We resort to Generative Adversarial Networks (GANs) that have shown the capability to generate highly convincing samples of a specific class as a result of a game between a discriminator and a generator module. Here, we apply the GANs to generate samples of images of metallic manufactured components with specific defects, in order to improve training of Semantic Networks (specifically DeepLabV3+ and Pyramid Attention Network (PAN) networks) carrying out the defect detection and segmentation. Our process carries out the generation of defect images using the StyleGAN2 with the DiffAugment method, followed by a conventional data augmentation over the entire enriched dataset, achieving a large balanced dataset that allows robust training of the Semantic Network. We demonstrate the approach on a private dataset generated for an industrial client, where images are captured by an ad-hoc photometric-stereo image acquisition system, and a public dataset, the Northeastern University surface defect database (NEU). The proposed approach achieves an improvement of 7% and 6% in an intersection over union (IoU) measure of detection performance on each dataset over the conventional data augmentation.

scholarly journals Approximate analytical fractional view of convection–diffusion equations

Open Physics ◽  
2020 ◽  
Vol 18 (1) ◽  
pp. 897-905
Author(s):  
Hassan Khan ◽  
Saima Mustafa ◽  
Izaz Ali ◽  
Poom Kumam ◽  
Dumitru Baleanu ◽  
...  
Keyword(s):  
Fractional Order ◽  
Variational Iteration Method ◽  
Absolute Error ◽  
Form Solution ◽  
Diffusion Equations ◽  
Nonlinear Convection ◽  
Convection Diffusion ◽  
Suggested Technique ◽  
First Time ◽  
Modified Variational Iteration Method

Abstract In this article, a modified variational iteration method along with Laplace transformation is used for obtaining the solution of fractional-order nonlinear convection–diffusion equations (CDEs). The proposed technique is applied for the first time to solve fractional-order nonlinear CDEs and attain a series-form solution with the quick rate of convergence. Tabular and graphical representations are presented to confirm the reliability of the suggested technique. The solutions are calculated for fractional as well as for integer orders of the problems. The solution graphs of the solutions at various fractional derivatives are plotted. The accuracy is measured in terms of absolute error. The higher degree of accuracy is observed from the table and figures. It is further investigated that fractional solutions have the convergence behavior toward the solution at integer order. The applicability of the present technique is verified by illustrative examples. The simple and effective procedure of the current technique supports its implementation to solve other nonlinear fractional problems in different areas of applied science.

scholarly journals Crowd density estimation using deep learning for Hajj pilgrimage video analytics

F1000Research ◽  
2021 ◽  
Vol 10 ◽  
pp. 1190
Author(s):  
MD ROMAN BHUIYAN ◽  
Dr Junaidi Abdullah ◽  
Dr Noramiza Hashim ◽  
Fahmid Al Farid ◽  
Dr Jia Uddin ◽  
...  
Keyword(s):  
Absolute Error ◽  
Visual Surveillance ◽  
Small Space ◽  
Video Footage ◽  
Crowd Control ◽  
Crowd Density Estimation ◽  
Crowd Density ◽  
Precise Analysis ◽  
Density Map ◽  
Control Study

Background: This paper focuses on advances in crowd control study with an emphasis on high-density crowds, particularly Hajj crowds. Video analysis and visual surveillance have been of increasing importance in order to enhance the safety and security of pilgrimages in Makkah, Saudi Arabia. Hajj is considered to be a particularly distinctive event, with hundreds of thousands of people gathering in a small space, which does not allow a precise analysis of video footage using advanced video and computer vision algorithms. This paper aims to propose an algorithm based on a Convolutional Neural Networks model specifically for Hajj applications. Additionally, the work introduces a system for counting and then estimating the crowd density. Methods: The model adopts an architecture which detects each person in the crowd, spots head location with a bounding box and does the counting in our own novel dataset (HAJJ-Crowd). Results: Our algorithm outperforms the state-of-the-art method, and attains a remarkable Mean Absolute Error result of 200 (average of 82.0 improvement) and Mean Square Error of 240 (average of 135.54 improvement). Conclusions: In our new HAJJ-Crowd dataset for evaluation and testing, we have a density map and prediction results of some standard methods.

Is a Head-Worn Inertial Sensor a Valid Tool to Monitor Swimming?

2021 ◽  
pp. 1-4
Author(s):  
Stephanie J. Shell ◽  
Brad Clark ◽  
James R. Broatch ◽  
Katie Slattery ◽  
Shona L. Halson ◽  
...  
Keyword(s):  
Correlation Coefficient ◽  
Video Analysis ◽  
Inertial Sensor ◽  
Absolute Error ◽  
Wearable Device ◽  
Measurement Unit ◽  
Fisher Exact Test ◽  
Mean Velocity ◽  
Video Footage ◽  
Stroke Type

Purpose: This study aimed to independently validate a wearable inertial sensor designed to monitor training and performance metrics in swimmers. Methods: A total of 4 male (21 [4] y, 1 national and 3 international) and 6 female (22 [3] y, 1 national and 5 international) swimmers completed 15 training sessions in an outdoor 50-m pool. Swimmers were fitted with a wearable device (TritonWear, 9-axis inertial measurement unit with triaxial accelerometer, gyroscope, and magnetometer), placed under the swim cap on top of the occipital protuberance. Video footage was captured for each session to establish criterion values. Absolute error, standardized effect, and Pearson correlation coefficient were used to determine the validity of the wearable device against video footage for total swim distance, total stroke count, mean stroke count, and mean velocity. A Fisher exact test was used to analyze the accuracy of stroke-type identification. Results: Total swim distance was underestimated by the device relative to video analysis. Absolute error was consistently higher for total and mean stroke count, and mean velocity, relative to video analysis. Across all sessions, the device incorrectly detected total time spent in backstroke, breaststroke, butterfly, and freestyle by 51% (15%). The device did not detect time spent in drill. Intraclass correlation coefficient results demonstrated excellent intrarater reliability between repeated measures across all swimming metrics. Conclusions: The wearable device investigated in this study does not accurately measure distance, stroke count, and velocity swimming metrics or detect stroke type. Its use as a training monitoring tool in swimming is limited.

Relative effects of trap versus trawl fisheries on population dynamics of the northern cod (Gadus morhua) stock

1998 ◽  
Vol 55 (1) ◽  
pp. 76-85 ◽  
Author(s):  
L Fahrig ◽  
S E Pope ◽  
K M Henein ◽  
G A Rose
Keyword(s):  
Population Dynamics ◽  
Gadus Morhua ◽  
Simulation Modelling ◽  
Trawl Fishery ◽  
Trawl Catch ◽  
Stock Biomass ◽  
Northern Cod ◽  
Trap Fishery ◽  
Using Data ◽  
Trawl Fisheries

We compared the effects of the inshore trap and the offshore trawl fisheries on the population dynamics of the northern cod (Gadus morhua) stock using data analyses and simulation modelling. We first statistically characterized the catch versus stock biomass relationships for the two fisheries (1977-1986). We found a significant (P < 0.0001) relationship between the trawl catch at time t and the stock biomass at time t - 2. No temporal lag was evident in the trap catch versus stock biomass relationship. The variability in these two relationships was similar. We then modelled the catch and stock biomass dynamics of the two fisheries in parallel, incorporating the observed catch versus stock biomass relationships, but assuming equal mean catches, to examine the effects on cod population dynamics of the temporal lag and variability in the catch versus stock biomass relationships. The results suggest that, for the same amount of fish taken, a quota-based trawl fishery presents a much greater risk of collapse to the cod stock than does an inshore trap fishery. Current management methods overestimate the "safe" catch for the trawl fishery because they do not incorporate the consequences of the lag in the relationship between stock biomass and trawl catch.

scholarly journals Prediction of delayed mortality using vitality scores and reflexes, as well as catch, processing, and post-release conditions: evidence from discarded flatfish in the Western Baltic trawl fishery

2018 ◽  
Vol 76 (1) ◽  
pp. 330-341 ◽  
Author(s):  
S B M Kraak ◽  
A Velasco ◽  
U Fröse ◽  
U Krumme
Keyword(s):  
Scientific Evidence ◽  
Handling Time ◽  
Fishing Ground ◽  
High Survival ◽  
Trawl Fishery ◽  
Fishing Vessels ◽  
Reflex Action ◽  
Delayed Mortality ◽  
Catch Composition ◽  
The Eu

Abstract The EU discard ban and its high-survival exemption exposed our lack of scientific evidence on discard survival in the fisheries. Discard survival is known to be highly variable and influenced by numerous factors, including conditions during the catch, on-board the fishing vessels, and post-discard. Therefore, obtaining unambiguous results in discard survival experiments is challenging. We conducted the first systematic year-round discard survival study of flatfish in the Western Baltic Sea on-board a commercial stern trawler under realistic fishing conditions (13 monthly hauls from May 2015 to May 2016) to test whether delayed mortality can be predicted by vitality scores and reflex action mortality predictor (RAMP) scores in combination with variable conditions during catch, processing, and post-release. The factors vessel type, gear, haul duration, fishing ground, depth, handling time, and processing procedures were kept constant as much as possible. On-board, live individual flatfish were tested for vitality and the presence of reflexes for RAMP, then “discarded” and kept in cages on the bottom of the seafloor for about a week after which the delayed mortality was determined. The proportions of “discarded” plaice, flounder, and dab that were dead after being kept in the cages ranged from 5% to 100%, 0% to 96%, and 33% to 100%, respectively. Higher mortalities occurred in summer-autumn when air and water temperatures were higher, catches smaller, and catches contained smaller proportions of roundfish. Relationships between RAMP scores and mortality probabilities varied substantially across the monthly trials. Indeed, in addition to RAMP or vitality scores and individual reflexes, the factors air and water temperature and catch weight and catch composition were significant in logistic GLMs explaining delayed mortality. Cross-validations indicated that delayed mortality could be predicted by these models with a reasonable accuracy. Nevertheless, the presence of possible confounding effects calls for caution in inferring causality and extrapolating the conclusions on predictability.

scholarly journals Catch composition of trawl fisheries in Mersin Bay with emphasis on catch biodiversity

2016 ◽  
Vol 40 ◽  
pp. 522-533 ◽  
Author(s):  
Gökhan GÖKÇE ◽  
İsmet SAYGU ◽  
Ahmet Raif ERYAŞAR
Keyword(s):  
Catch Composition ◽  
Trawl Fisheries

scholarly journals Scampi (Metanephrops challengeri) emergence patterns and catchability

2015 ◽  
Vol 72 (suppl_1) ◽  
pp. i199-i210 ◽  
Author(s):  
Ian D. Tuck ◽  
Darren M. Parsons ◽  
Bruce W. Hartill ◽  
Stephen M. Chiswell
Keyword(s):  
New Zealand ◽  
Stock Assessment ◽  
Time Of Day ◽  
Major Step ◽  
Emergence Patterns ◽  
Photographic Survey ◽  
Strong Emergence ◽  
First Time ◽  
Trawl Fisheries ◽  
Acoustic Tagging

Abstract Catchability is often a key source of uncertainty with any stock assessment, but especially for burrowing species, as their emergent behaviour is often poorly understood. Quantification of catchability will provide a major step towards improvements in the assessment for many species. Scampi (Metanephrops challengeri) are widely distributed around New Zealand, and as with Nephrops (Norway lobster), they occupy burrows in muddy substrate, and are exploited through trawl fisheries, but are only available to these fisheries when emerged on the seabed. Burrow emergence is known to vary over daily and longer cycles. Uncertainty over trawl catchability associated with emergence patterns has led to the development of photographic survey approaches for scampi, based on the counts of burrows. Both survey approaches require an understanding of burrow occupancy and emergence rates to estimate trawl/photo survey catchability, which is a key source of uncertainty. We used acoustic tagging to examine levels of and patterns in the emergence of Metanephrops, using hydrophone receivers moored close to the seabed. Strong emergence cycles were apparent in relation to tidal current (higher emergence with inshore water flow across the slope) and time of day (peaking just after dawn). These data have subsequently been used within a length-based stock assessment of New Zealand scampi, which uses emergence data in conjunction with burrow and animal counts from photographic surveys, for the first time, to inform priors for trawl (mean 0.094) and photographic (mean 0.46) survey catchability, and for burrow occupancy (mean 49.3%).

Early Recognition of Lumbar Overdrainage by Lumboventricular Pressure Gradient

Neurosurgery ◽  
2011 ◽  
Vol 68 (5) ◽  
pp. 1187-1191 ◽  
Author(s):  
Dimitre Staykov ◽  
Verena Speck ◽  
Bastian Volbers ◽  
Ingrid Wagner ◽  
Marc Saake ◽  
...  
Keyword(s):  
Pressure Gradient ◽  
Clinical Symptoms ◽  
Early Recognition ◽  
Aneurysmal Subarachnoid Hemorrhage ◽  
External Ventricular Drainage ◽  
Monitoring Tool ◽  
Management Protocol ◽  
Complete Reversal ◽  
Promising Treatment ◽  
First Time

Abstract BACKGROUND: Lumbar drainage (LD) represents a promising treatment strategy for prevention of vasospasm after aneurysmal subarachnoid hemorrhage (SAH). OBJECTIVE: To report on transient herniation caused by lumbar overdrainage in 3 patients with severe SAH who were treated with early LD within an ongoing feasibility study. METHODS: Patients with first-time aneurysmal SAH received LD within 72 hours of symptom onset, after aneurysm clipping or coiling. LD, with a target drainage amount of 5 to 10 mL, was continued for 6 to 9 days. External ventricular drainage (EVD) was begun on admission when hydrocephalus was present. With both catheters in place, intracranial pressure (ICP) and lumbar pressure (LP) were monitored simultaneously. RESULTS: Three of 22 patients developed a progressive lumboventricular pressure gradient, likely due to cerebrospinal fluid (CSF) overdrainage. Two patients showed signs of herniation. Clamping of LD resulted in complete reversal of symptoms in those patients. The lumboventricular pressure gradient began to evolve at least 12 hours before clinical symptoms developed, and gradually disappeared in all 3 patients after LD clamping. CONCLUSION: Lumbar overdrainage should be avoided in severe SAH, and lumboventricular pressure measurement may be a useful monitoring tool. Herniation due to lumbar overdrainage is a feared complication that can be avoided by following a strict LD management protocol.

Sign in / Sign up

Export Citation Format

Share Document