The K-tree software library has been written in C++. It has been tested on a *NIX working environment, but should also work on other platforms such as Windows.

Hardware

K-tree has been designed to operate on affordable off-the-shelf workstations that cost between $5k and $10K. It has been tested on a server with 256GB memory, which allows for the equivalent of 2 billion pixels to be clustered (120 GeoTiff images from the Landsat 5 and 7 archives at 1 x 1 degree/4000 x 4000 pixel resolution).

Software

Compiling

It is recommended to use GCC 4.8+ or equivalents to compile the K-tree source code.

Software dependencies

The K-tree library and programs relies on following third party libraries:

Boost (any up to date version)
Intel Theading Building Blocks
GDAL
Java 8+ (optional) for a GUI tool plotting the k-tree search results
Third party libraries installation

GDAL

For CentOS/RHEL/Oracle Linux users,

#yum install gdal*

for Debian/Ubuntu/Linux Mints/Other Debian derivatives Linux users,

$sudo apt-get install gdal*

If you are using CentOS/RHEL 6.0, please read the GDAL installation guide for CentOS 6.

TBB

Please visit their website to download the latest version, but sometimes their new version won't be compiled by the compiler. In this case, you might need to choose an achieved version such as tbb42_20140601oss. After the installation, you will need put a line similar as the following in your shell environment file (eg. ~/.bashrc), or run it before compiling:

source /opt/tbb/tbb42_20140601oss/bin/tbbvars.sh intel64

Boost

In most Linux distributions, boost can be installed via system package management tool:

#yum install boost*

or,

$sudo apt-get install boost*

There are four particular boost libraries required:

-lboost_system -lboost_thread -lboost_timer \
                    -lboost_chrono -lboost_filesystem