Premium
Visual rhythm for particle analysis in sample‐in‐flow systems: Application for continuous plankton monitoring
Author(s) -
Matuszewski Damian J.,
Cesar Roberto M.,
Strickler J. Rudi,
Baldasso Luis F.,
Lopes Rubens M.
Publication year - 2015
Publication title -
limnology and oceanography: methods
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.898
H-Index - 72
ISSN - 1541-5856
DOI - 10.1002/lom3.10058
Subject(s) - plankton , computer science , frame (networking) , sample (material) , segmentation , artificial intelligence , computer vision , environmental science , ecology , biology , telecommunications , chemistry , chromatography
Planktonic organisms are important components of aquatic ecosystems and there is a growing interest in their exploitation for industrial purposes, such as in biofuel production using microalgae as a biomass source. The evaluation of plankton composition and abundance usually relies on time‐consuming microscopic analysis. However, automatic imaging systems are becoming increasingly available to perform such tasks. Here we present a novel method based on the visual rhythm (VR) approach to detect particles in fluidic systems through imaging techniques. It was tested on large datasets of plankton images and its performance was compared to a traditional frame‐by‐frame (FBF) approach. Performed tests demonstrated that the VR‐based approach is faster and much more precise for plankton detection and segmentation (accuracy above 96% and precision above 95%), while having the same accuracy as the FBF method. Although designed for continuous monitoring of plankton samples, the method can be easily adapted to different applications in which targets to be detected and identified move in a unidirectional flow.