iwyoo / groundtruth-drosophila-vnc Goto Github PK

The ground truth stack 1 contains 20 sections from serial section Transmission Electron Microscopy (ssTEM) of the Drosophila melanogaster third instar larva ventral nerve cord. The cube measures 4.7 x 4.7 x 1 microns approx, with a resolution of 4.6 x 4.6 nm/pixel and section thickness of 45-50 nm.

Use stack 1 to train pixel classifiers for the purpose of automatic segmentation of neural structures in ssTEM. For instance, use 80% of the sections to train, and 20% of the sections for validation. Apply the classifier to the raw data of stack 2 (test data) and send us the results, so we're able to compute pixel-wise and other error measures in the subsequent neuron reconstruction pipeline.

This work is licensed under a Creative Commons Attribution-NonCommercial-ShareAlike 3.0 Unported License.

You are free to use this data set for the purpose of generating or testing non-commercial image segmentation software. If any scientific publications derive from the usage of this data set, you must cite:

Segmented anisotropic ssTEM dataset of neural tissue. Stephan Gerhard, Jan Funke, Julien Martel, Albert Cardona, Richard Fetter. figshare. Retrieved 16:09, Nov 20, 2013 (GMT) http://dx.doi.org/10.6084/m9.figshare.856713

Stephan Gerhard

/stack1/
   /raw: The raw 8-bit greyscale images
   /membranes: Series of binary images of segmented membranes
   /mitochondria: Series of binary images of segmented mitochondria
   /synapses: Series of binary images of segmented synapses
   /labels: Series of merged labels including oriented membranes, membrane junctions,
            mitochondria and synapses. The pixels are labeled as follows:
            0   -> membrane | (0°)
            32  -> membrane / (45°)
            64  -> membrane - (90°)
            96  -> membrane \ (135°)
            128 -> membrane "junction"
            159 -> glia/extracellular
            191 -> mitochondria
            223 -> synapse
            255 -> intracellular

/stack2/
  /raw: The raw 8-bit greyscale images