Tracking systems

As part of my PhD thesis at de Polavieja Lab, I have developed two videotracking systems:

idTracker

Using a very sensitive recognition algorithm, idTracker distinguishes each individual in each frame. It can therefore maintain its identity during the whole video, regardless of the complexity of the crossings with other individuals, and even if it temporarily disappears from view.

idTracker: Summary of features & performance


Main features of idTracker:

  • Provides the trajectory of each individual with typically >99% of frames with correct identities (validated against human)
  • It is applicable to any species, and so far we have validated it for fruitflies (Drosophila melanogaster), mice (Mus musculus), zebrafish (Danio rerio), nine-spine stickleback (Pungitius pungitius), medaka (Oryzias latipes), European sea bass (Dicentrarchus labrax), Hemigrammus rhodostomus and ants (Atta cephalotes and Atta columbica)
  • Reference images to recognize each individual are automatically extracted from the problem video, so it does not require any additional recordings.

Related publication:

A. Pérez-Escudero, J. Vicente-Page, R.C. Hinz, S. Arganda, G.G. de Polavieja (2013) idTracker: Tracking individuals in a group by automatic identification of unmarked animals (submitted)

More info at www.idtracker.es

multiwellTracker

This software follows isolated individuals in wells. It can analyze many wells simultaneously and with low resolution and low contrast between animals and background. It auto-detects circular wells in the video.

More info at www.multiwelltracker.es