FishEye: Marine species' recognition and visualization

2017 ◽  
Author(s):  
Tiago Nascimento ◽  
Sandra Gama
Keyword(s):  
Species Recognition ◽  
Marine Species

scholarly journals Fish species recognition using transfer learning techniques

2021 ◽  
Vol 7 (2) ◽  
pp. 188
Author(s):  
Jaisakthi Seetharani Murugaiyan ◽  
Mirunalini Palaniappan ◽  
Thenmozhi Durairaj ◽  
Vigneshkumar Muthukumar
Keyword(s):  
Transfer Learning ◽  
Fish Species ◽  
Species Recognition ◽  
Marine Species ◽  
Superior Performance ◽  
Marine Animals ◽  
Large Dataset ◽  
Underwater Environment ◽  
Learning Techniques ◽  
Learning Technique

Marine species recognition is the process of identifying various species that help in population estimation and identifying the endangered types for taking further remedies and actions. The superior performance of deep learning for classification is due to the property of estimating millions of parameters that have to be extracted from many annotated datasets. However, many types of fish species are becoming extinct, which may reduce the number of samples. The unavailability of a large dataset is a significant hurdle for applying a deep neural network that can be overcome using transfer learning techniques. To overcome this problem, we propose a transfer learning technique using a pre-trained model that uses underwater fish images as input and applies a transfer learning technique to detect the fish species using a pre-trained Google Inception-v3 model. We have evaluated our proposed method on the Fish4knowledge(F4K) dataset and obtained an accuracy of 95.37%. The research would be helpful to identify fish existence and quantity for marine biologists to understand the underwater environment to encourage its preservation and study the behavior and interactions of marine animals.

A morphometric diagnosis using continuous characters of Pinnunavis edkuensis, sp. nov. (Bacillariophyta: Bacillariophyceae), a brackish-marine species from Egypt

Phytotaxa ◽  
2015 ◽  
Vol 212 (1) ◽  
pp. 1 ◽  
Author(s):  
ROBERT K. EDGAR ◽  
AMAL I. SALEH ◽  
STACY M. EDGAR
Keyword(s):  
Surface Area ◽  
Species Recognition ◽  
Cladistic Analysis ◽  
Central Area ◽  
Distinct Group ◽  
Marine Species ◽  
Morphological Characters ◽  
Similar Species ◽  
Continuous Variables ◽  
Marine Habitats

Pinnunavis edkuensis is reported as a new diatom species from Egypt from brackish and marine habitats. Its description and diagnoses focus on 12 continuous morphological characters: valve breadth, valve mantle shape and surface area, axial-central area shape and surface area, axial area breadth, distal and proximal interstriae distances, striae slopes, magnitude and position of maximal slope along a semivalve transect, and inter-raphe distance. These characters are considered along their ontogenetic trajectories, using a size vector, by means of regression analyses. Shape was assessed using geometric morphometric methods. Diagnostic comparisons of P. edkuensis were made with P. yarrensis, for which a lectotype was chosen and Grunow’s unpublished description and drawings presented, and P. zalatii, a new name replacing the illegitimate name, Navicula aegyptiaca. P. edkuensis was demonstrated to be a distinct group compared to these two similar species using 11 of the 12 characters. The diagnoses emphasize species recognition by means of anova, regression and randomization analyses and specimen determination primarily by means of regression-based prediction intervals. Additional comments are made on the statistical nature of gaps associated with characters described by discrete or continuous variables, the importance of reporting sample sizes as part of diatom diagnoses and other comparisons, and the role of size in specimen comparability and taxon identification. The generic distinctness of Pinnunavis and Pinnularia is discussed based on a preliminary cladistic analysis of morphological characters.

FishEye - An Integrated Marine Species' Visualization

2018 ◽  
Vol 9 (2) ◽  
pp. 1-17
Author(s):  
Tiago Nascimento ◽  
Sandra Pereira Gama
Keyword(s):  
Cognitive Load ◽  
Image Recognition ◽  
Information Visualization ◽  
Species Recognition ◽  
Marine Species ◽  
Data Interpretation ◽  
Relevant Information ◽  
Usability Tests ◽  
Animal Data ◽  
Tracking Devices

Despite the historical limitations associated with the study of marine species, current technology makes it possible to collect ocean animal data in a more accessible way, with a variety of tagging and tracking devices. Hence, such information is nowadays generated in large amounts, often in textual formats, making it difficult to interpret and analyze. Information visualization, due to its potential to represent large amounts of data while alleviating cognitive load associated with data interpretation, may help overcome this limitation. This article presents the visualization of a marine species that allows the representation and interactive exploration of species' telemetric data through an integrated dashboard with coordinated views. A species image recognition module was implemented together with the described visualization, enabling species recognition. Usability tests have validated its potential in making important patterns immediately perceivable and also showed that FishEye provides exploration and comparison mechanisms to obtain further relevant information.

Deep Learning for Marine Species Recognition

2019 ◽  
pp. 129-145 ◽  
Author(s):  
Lian Xu ◽  
Mohammed Bennamoun ◽  
Senjian An ◽  
Ferdous Sohel ◽  
Farid Boussaid
Keyword(s):  
Deep Learning ◽  
Species Recognition ◽  
Marine Species

High predatory efficiency and abundance drive expected ecological impacts of a marine invasive fish

2020 ◽  
Vol 637 ◽  
pp. 195-208 ◽  
Author(s):  
EM DeRoy ◽  
R Scott ◽  
NE Hussey ◽  
HJ MacIsaac
Keyword(s):  
Invasive Species ◽  
Native Species ◽  
Ecological Impact ◽  
Marine Species ◽  
Ecological Impacts ◽  
Negative Effects ◽  
Marine Predator ◽  
Eastern Gulf Of Mexico ◽  
Red Grouper ◽  
Per Capita

The ecological impacts of invasive species are highly variable and mediated by many factors, including both habitat and population abundance. Lionfish Pterois volitans are an invasive marine species which have high reported detrimental effects on prey populations, but whose effects relative to native predators are currently unknown for the recently colonized eastern Gulf of Mexico. We used functional response (FR) methodology to assess the ecological impact of lionfish relative to 2 functionally similar native species (red grouper Epinephelus morio and graysby grouper Cephalopholis cruentata) foraging in a heterogeneous environment. We then combined the per capita impact of each species with their field abundance to obtain a Relative Impact Potential (RIP). RIP assesses the broader ecological impact of invasive relative to native predators, the magnitude of which predicts community-level negative effects of invasive species. Lionfish FR and overall consumption rate was intermediate to that of red grouper (higher) and graysby grouper (lower). However, lionfish had the highest capture efficiency of all species, which was invariant of habitat. Much higher field abundance of lionfish resulted in high RIPs relative to both grouper species, demonstrating that the ecological impact of lionfish in this region will be driven mainly by high abundance and high predator efficiency rather than per capita effect. Our comparative study is the first empirical assessment of lionfish per capita impact and RIP in this region and is one of few such studies to quantify the FR of a marine predator.

Geolocation estimation of marine species using a high resolution light tag

OCEANS 2009 ◽  
2009 ◽  
Author(s):  
Prajas John ◽  
Jaison Peter ◽  
Adrine Antony Correya ◽  
M. H. Supriya ◽  
P. R. Saseendran Pillai
Keyword(s):  
High Resolution ◽  
Marine Species

Natural Compounds and Drug Discovery: Can Cnidarian Venom Play a Role?

2019 ◽  
Vol 19 (2) ◽  
pp. 114-118
Author(s):  
Gian Luigi Mariottini ◽  
Irwin Darren Grice
Keyword(s):  
Biological Activity ◽  
Marine Species ◽  
Natural Compounds ◽  
Therapeutic Agents ◽  
Industrial Applications ◽  
Bioactive Molecules ◽  
Sea Anemones ◽  
Therapeutic Approaches ◽  
Ongoing Research ◽  
Biological Compounds

Natural compounds extracted from organisms and microorganisms are an important resource for the development of drugs and bioactive molecules. Many such compounds have made valuable contributions in diverse fields such as human health, pharmaceutics and industrial applications. Presently, however, research on investigating natural compounds from marine organisms is scarce. This is somewhat surprising considering that the marine environment makes a major contribution to Earth's ecosystems and consequently possesses a vast storehouse of diverse marine species. Interestingly, of the marine bioactive natural compounds identified to date, many are venoms, coming from Cnidarians (jellyfish, sea anemones, corals). Cnidarians are therefore particularly interesting marine species, producing important biological compounds that warrant further investigation for their development as possible therapeutic agents. From an experimental aspect, this review aims to emphasize and update the current scientific knowledge reported on selected biological activity (antiinflammatory, antimicrobial, antitumoral, anticoagulant, along with several less studied effects) of Cnidarian venoms/extracts, highlighting potential aspects for ongoing research towards their utilization in human therapeutic approaches.

Sign in / Sign up

Export Citation Format

Share Document