Quick Start - asolis/vivaTracker GitHub Wiki

Requirements

Before using the vivaVideo base project, you need to install OpenCV, Eigen libraries and CMake tool for your platform:

After installing OpenCV, CMake and EigenLibrary, clone the vivaTracker repository:

https://github.com/asolis/vivaTracker.git

Assuming you cloned the git repository in a vivaTracker folder inside your home folder:

  • Mac: /Users/username/vivaVideo
  • Linux: /home/username/vivaVideo
  • Windows: C:\Users\username\vivaVideo

(username is your system username)

###CMake

Open the CMake GUI and point the source code to your working copy path:

(e.g., /Users/username/vivaVideo)

Point CMake's build folder (i.e. "Where to build the binaries") to a build folder path:

(e.g., /Users/username/vivaVideo/build/)

CMake will ask for permission to create the build folder for you if it doesn't exist. Click yes.

Add OpenCV to your vivaTracker project

Click the Configure button and specify the generator for this project (e.g., Xcode, Visual Studio, make files, ... etc)

Point the OpenCV_DIR variable to your installed OpenCV folder containing these files:

  • OpenCVConfig-version.cmake
  • OpenCVConfig.cmake

Add Eigen to your vivaTracker project

Point the EIGEN3_INCLUDE_DIRS variable to your installed Eigen header files folder library. (e.g., "/usr/local/include/eigen3/Eigen")

Add annotated datasets to your project

Specify if you wish to download any of the available datasets by turning on the cmake options "download_datasetname". This will download and extract the annotated datasets and make them available for testing with the tracking methods.

(e.g., download_vot2013, download_vot2014, ....)

Add tracker implementations to your project

The CMake rules will look inside the trackers folder searching for possible tracking implementations. For each found tracker an option WITH_{TRACKER_NAME} will be available for integration (i.e., WITH_SKCF, WITH_NCC). The TRACKER_NAME will be the folder name all in uppercase. Before generating your code, you should select all the tracking algorithms that you will like to include in the final project. The

Generating the project

Click the Configure button again and make sure CMake finds your OpenCV and Eigen installation. Click the Generate button.

if everything went OK you will see the following messages with the list of available trackers founded by the cmake rules. In the case of selecting to download any of the available datasets a progress bar will report the download/unpacking progress for each dataset.

    Project Name: vivaTracker
    OpenCV version: 3.1.0 
    libraries: ...
    include path: ...
    tracker included: kcf
    tracker included: kcf2
    tracker included: ncc
    tracker included: opentld
    tracker included: skcf
    tracker included: struck
    downloading dataset: vot2013 10%complete...
    Configuring done
    Generating done

Your built project is now created inside your build folder. You should use the selected generator (i.e., Xcode, Visual Studio, make files, etc...) to build the vivaTracker executable.

IDE details:

#####Visual Studio:

  1. Select vivaTracker as your StartUp project.
  2. Open Project > Properties > Configuration Properties > Debugging.
  3. The Working Directory entry would be $(ProjectDir) by default. Change it to $(SolutionDir)$(Configuration)\

#####Make files

  1. After compiling program using the make. Move the vivaTracker executable to the Debug or Release folder before executing it.