|
|
|
|
|
Wacquet, Guillaume; Lefebvre, Alain; Colas, Florent; Grosjean, Philippe. |
Today, the most widely used method for the analysis of samples within the REPHY network, is the microscopy. However, this method has several drawbacks: different skill level depending on the analyst, taxa may be misidentified, analyst subject to fatigue and loss of concentration, non-quantifiable errors. That is why different axis of development have been proposed to adapt the FlowCAM and ZooPhytoImage to the requirements of REPHY network which are the accuracy and repeatability of the measurement. The purpose of these new tools being to provide a tangible gain of time for analysts compared to the observations with optical microscope. The analysis time being a key point of this system, studies were conducted in order to achieve a fast acquisition. To have... |
Tipo: Text |
|
Ano: 2014 |
URL: http://archimer.ifremer.fr/doc/00363/47440/47465.pdf |
| |
|
| |
|
|
Grosjean, Philippe; Wacquet, Guillaume. |
This report details the work accomplished to enhance the Zoo/PhytoImage software to optimize its use for the analysis of phytoplankton samples in general, but more particularly, in the framework of an operational survey of coastal seawater (REPHY, IFREMER). Zoo/PhytoImage allows to analyze “numerically recorded” plankton samples, that is, by using digital images gathered with specialized devices such as the FlowCAM, or the FastCAM (see report 3). A machine learning approach allows to automatically classify the digitized particles into various taxonomic groups. Once this is done, global statistics are calculated on each sample, including the number of particles, the biomass, and the size spectrum per taxonomic group. Two major changes are introduced in the... |
Tipo: Text |
Palavras-chave: Phytoplancton; REPHY; Analyse d'image; Classification supervisée; Dénombrement de cellules; Apprentissage actif; Manche; Atlantique.; Phytoplankton; REPHY; Image analysis; Machine learning; Cells enumeration; Active learning; The Channel; Atlantic Ocean. |
Ano: 2016 |
URL: http://archimer.ifremer.fr/doc/00389/49990/50578.pdf |
| |
|
| |
|
|
Wacquet, Guillaume; Grosjean, Philippe; Colas, Florent; Hamad, Denis; Artigas, Luis Felipe. |
The coupled system FlowCAM/ZooPhytoImage has become a real operational tool in 2014. However, to be fully adapted to the observations of phytoplankton performed in the context of the REPHY observation network and in order to better respond to present and future requests concerning the evaluation of quality of coastal and marine waters within the European requirements, such as the WFD and MSFD, new functionalities must be integrated into existing tools. Therefore, different axis of development have been proposed by UMONS and Ifremer to adapt both the digitization device and the Zoo/PhytoImage software to the constraints defined by the REPHY. First, version 5 of Zoo/PhytoImage contains recent innovations such as the development of routines to automatically... |
Tipo: Text |
Palavras-chave: Plancton; Analyse automatisée; Analyse d'image; Classification supervisée; Apprentissage actif; Dénombrement de cellules; Plankton; Automated analysis; Image processing; Supervised classification; Active learning; Cells counting. |
Ano: 2015 |
URL: http://archimer.ifremer.fr/doc/00389/49986/50573.pdf |
| |
|
|
|