imclab/neuroproof

Tool for graph-based image segmentation and analysis

NeuroProof 1.1

##Toolkit for Graph-based Image Segmentation and Analysis

The NeuroProof software is an image segmentation tool currently being used in the FlyEM project at Janelia Farm Research Campus to help reconstruct neuronal structure in the fly brain. This tool provides routines for efficiently agglomerating an initial volume that is over-segmented. While NeuroProof has been tested in the domain of EM reconstruction, we believe it to be widely applicable to other application domains. In addition, to graph agglomeration tools, this package also provides routines for estimating the uncertainty of a segmentation Plaza, et al '12 and mechanisms to comparing with ground truth.

###Features

• Algorithm for efficiently learning an agglomeration classifier
• Algorithm for efficiently agglomerating an oversegmented volume
• Implementation of variation of information metric for comparing two labeled volumes
• Algorithm to estimate uncertainty of graph and is predicted edge confidences
• Tools to assess the amount of work required to edit/revise a segmentation
• GUI front-end for visualizing segmentation results; interface for merging segments together
• Simple and access-efficient graph-library implementation and straightforward conversion to powerful boost graph library
• Data stack implementation that allows one to leverage image processing algorithms in Vigra
• Python bindings to enable accessing the Rag, segmentation routines, and editing operations in various tool environments like in Gala and Raveler

Installation Instructions

NeuroProof is currently only supported on linux operating systems. It probably would be relatively straightforward to build on a MacOS. Documentation in NeuroProof follow Doxygen comment conventions; an html view can be created by running the following command:

% doxygen doxygenconfig.file

NeuroProof supports two mechanisms for installation:

Buildem

The most-supported and most push-button approach involves using the build system called buildem. The advantage of using buildem is that it will automatically build all of the dependencies and put it in a separate environment so as not to conflict with any other builds. These automatic builds contain versions of the dependencies that are know to work with NeuroProof. The disadvantage of buildem is that it will install a lot of package dependencies. The initial build of all of the dependencies could take on the order of an hour to complete.

To build NeuroProof using buildem

% mkdir build; cd build;
% cmake .. -DBUILDEM_DIR=/user-defined/path/to/build/directory
% make -j num_processors

This will automatically load the binaries into BUILDEM_DIR/bin. To run the test regressions, add this directory into your PATH environment and run make test in the build directory.

NeuroProof Only

NeuroProof can be built without Buildem with the following:

% mkdir build; cd build;
% cmake ..
% make

To successfully compile this code, you will need to install the following dependencies and make sure their headers and libraries can be found in the common linux search paths:

• jsoncpp
• boost
• vigra
• hdf5
• opencv
• vtk
• qt4

For more information on the supported versions of these dependencies, please consult buildem.

To run make test, the installed libraries will need to be in the default library search path or set in LD_LIBRARY_PATH. The PYTHONPATH environment variable will need to be set to the lib directory created within NeuroProof. The binaries produced will be placed in the NeuroProof bin directory.

NeuroProof Examples

The top-level programs that are built in NeuroProof are defined in the src/ directory. NeuroProof's capabilities are mainly accessed via command-line executables. A subset of these capabilities are exposed in a python interface. For some examples on how to run the tool, please consult the 'examples' directory. If you have just cloned the NeuroProof repository, you will need to add the example submodule. (The examples are not included by default due to the size of the directory.)

% git submodule init
% git submodule update

To Be Implemented

• Focused agglomeration training in stack viewer
• Add capability to call NeuroProof as a service in clustered environments
• Direct support for 2D datasets (2D data is generally allowable in the current implementation)
• Improved algorithms for handling EM reconstruction specific goals