Install
WARNING: This page was written to install CellTracking 0.9x (not supported anymore), it may contain some inaccuracies for the installations of v1.x.
This page describe how to install CellTracking for various platforms. Once all the dependencies are installed, you only have to unpack the CellTracking archive to start working with it.
Installation instructions are available for the following systems:
1 Debian
- Install Python libraries
# aptitude install python-numpy python-scipy python-gtk2
- Install OpenCV with Python bindings
Note that OpenCV version >2.0 (and maybe more) is required, you should be able to import the
cv2
module in python.Debian-like (including Ubuntu) used to have obsolete Python bindings for OpenCV, which required the user to compile OpenCV. This is no longer the case since the binding is not buggy anymore. If you have a modern version of the operating system (i.e.: Python bindings for OpenCV work properly, simply install the packages using the command line (2.1). If this doesn't work, try to compile OpenCV (2.2).
2.1 Install everything thanks to
aptitude
# aptitude install python-opencv
Test
$ python > import cv2
If Python returns you an error, you might have an obsolete Python binding (see below 2.2).
2.2 Compile OpenCV If the first option doesn't work (obsolete Python binding)
In order to compile the Python bindings,
python-dev
should be installed. If not,import cv
will fail, even if the OpenCV install succeeded. python-dev can be installed by# aptitude install python-dev
.To get a video playback, you must install libgtk2.0-dev:
# aptitude install libgtk2.0-dev
.Detailed instructions can be found here and here. These pages have a clear detail of the required packages.
If OpenCV compilation fails quite early, you should consider upgrading your
cmake
version, according to this post. - Install
CellTracking
- Download the CellTracking archive: Downloads.
- Unpack it.
- Launch
./CellTracking.sh
2 MacOS X
This procedure relies on MacPorts to install the majority of the required software. This is probably not compulsory, but this ensure an easy way to install OpenCV and to have all the components installed for the same version of Python.
- Install XCode
- XCode will provide all the compiling tools required to make OpenCV work properly. Download the latest version compatible with your version of MacOS (https://developer.apple.com/downloads/, Apple requires you to create an account). Note that OpenCV is unlikely to compile if XCode version is below 3.2.
- During the installation, make sure that you check the box to install the Command line tools. For XCode 4, you must launch XCode and go to: Preferences->Downloads->Download Command Line Tools.
- To compile Scipy, you need to agree some xcode license
sudo xcodebuild -license agree
. (Not required in version 1.x)
- Install MacPorts: http://www.macports.org/install.php
- Install the required libraries
sudo port selfupdate sudo port -v install opencv +python27 sudo port install py27-gtk sudo port install py27-pil sudo port install py27-scipy
- Install
CellTracking
- Download the CellTracking archive: Downloads.
- Unpack it.
- Launch
./CellTracking.command
Note : Using Leopard (and maybe some other version), an error might occur when launching CellTracking.command :
Fatal IO error 35 (Resource temporarily unavailable) on X server
. An explanation is that the automatically launched version of X11 is buggy. A simple workaround is to:- Open /Applications/MacPorts/X11,
- then go to Applications->Terminal,
- and
cd
to your CellTracking folder - Finally, type
./CellTracking.command
to launch CellTracking - More information here.