EllipTrack is a new global-local hybrid cell tracker optimized for tracking modern large-scale movies. The key innovation of EllipTrack is the introduction of a local track correction module, which systematically corrects the tracking mistakes of the global track linking algorithm from Magnusson and colleagues. Benchmark reveals that EllipTrack outperforms the existing state-of-the-art cell trackers and can identify nearly error-free cell lineages from multiple large-scale movies.
-
Hardware: A modern computer, at least 8GB RAM.
-
Software: MATLAB (Mathworks), R2017a or later.
Required toolboxes: System Identification Toolbox, Image Processing Toolbox, Computer Vision Toolbox, Statistics and Machine Learning Toolbox, Parallel Computing Toolbox, and MATLAB Distributed Computing Server.
A C++ compiler in MATLAB with C++11 support. Install a free minGW compiler here.
-
Download EllipTrack: Click "Clone or download" on the top of this page, select "Download ZIP", and extract files.
-
Compile generate_tracks.cpp: Navigate MATLAB to the functions/track_linking folder, and execute
mex -largeArrayDims generate_tracks.cpp -
Install BioformatsImage: Follow the instruction here.
Documentation is available here.