IncredibleEye 2015
This is a sample created with the objective of demonstrating the functionalities of our SDK in the facial recognition area.
Compiled executables can be found in the IncredibleEye Developer Zone (click here).
You'll need to install:
- Qt 5.4.x
- CMake 2.8.11+
- OpenCV 2.4.x
- IncredibleEye SDK
To build, you'll need to run these commands in the FacialRecognitionSample directory:
Linux:
$ mkdir build
$ cd build
$ cmake -D HRDSDK_DIR="<path to the sdk>" -D OpenCV_DIR="<opencv directory>/opencv/build" ..
$ make
Windows (for MS Visual Studio):
$ mkdir build
$ cd build
$ cmake -G "Visual Studio 12 2013" -D HRDSDK_DIR="<path to the sdk>" -D OpenCV_DIR="<opencv directory>\opencv\build" ..
$ start FacialRecognitionSample.sln
If CMake complains about Qt5Widgets, that means there is no environment variable set to the Qt directory, and you'll need to add this parameter to the cmake command:
-D CMAKE_PREFIX_PATH="<path to qt>"
where the path leads to the directory containing Qt's /bin directory.
To run the generated executable, you'll need to place the following files and directories in the same directory as the executable:
- config (from the IncredibleEye SDK)
- shared (from the IncredibleEye SDK)
- plugins (from the IncredibleEye SDK) (not needed on Linux)
- platforms (from the IncredibleEye SDK)
- Resources (available here in the /src directory)
- properties.txt file
- (If running on Windows) the .dll files from the IncredibleEye SDK
- The first time the application is ran, you will be taken through a series of steps in order to be able to test the sample: Camera selection, camera validation and user creation are key for the sample to function, so make sure you do all of those steps.
- Once that is done, you can test the recognition process, and add more users to comprove it's accuracy.
Check out the website for more information on how to train correctly
Note that in Linux, due to OpenCV, the cameras might only work the first time you change them. Restarting the application will make them work again.