Diatom detection and identification

Work in progresss

Objective

Diatoms are a type of  unicellular microalgae commonly used as bio-indicators for monitoring the ecological status of watercourses, particularly in the context of the implementation of the European Water Framework Directive. Current biological indices based on diatoms rely on morphological criteria (shape and ornamentation of the siliceous exoskeleton, the frustule) that are sometimes difficult to characterize using conventional optical methods.
The objective of the project is to develop a framework to automatically identify diatoms from virtual slide images using deep learning.
 

The identification process can be divided into 2 main steps:

  1. Detection – locating diatoms on the virtual-slide images
  2. Classification – assigning diatoms into their corresponding taxa

Diatom Detection

Using Bounding Box 

Using Image Segmentation

© 2019 – 2021 DREAM Lab – Georgia Tech Lorraine