Tree Structure Reconstruction

Datasets for Tree Structure Reconstruction

This page contains two datasets of tree structures: 3D confocal microscopy stacks of neurons and aerial images of roads.

All ground truth tracings available in this page are copyrighted by CVLab – EPFL. You are free to use them for research purposes. If you use them to publish results, please cite the reference below.

Olfactory Projection Fibers

This dataset is taken from the DIADEM Challenge. The Z spacing of the stacks are set to 1.0 to make fiber cross sections appear circular. For details, please refer to the dataset’s website.

Stacks and Ground Truth Tracings (18 MB)

Aerial Images of Roads

Images and Ground Truth Tracings (18 MB)


Our semi-automated tracing tool for acquiring the ground-truths is available here.


The datasets on this page has been used for our automated tree reconstruction algorithm described in the following paper:

Automated Reconstruction of Curvilinear Networks from 2D and 3D Imagery

E. Türetken / P. Fua (Dir.)  

Lausanne, EPFL, 2013. 

Automated Reconstruction of Tree Structures using Path Classifiers and Mixed Integer Programming

E. Türetken; F. Benmansour; P. Fua 

2012. IEEE Conference on Computer Vision and Pattern Recognition (CVPR), Providence, Rhode Island, June 16-21, 2012. p. 566-573. DOI : 10.1109/CVPR.2012.6247722.