visionary

Various types of detection implementations with OpenCV and C++.

Building Repository

To build the repository you need CMake. You can download from here. You can create a shortcut cmake command for macOS as below

sudo mkdir -p /usr/local/bin
sudo /Applications/CMake.app/Contents/bin/cmake-gui --install=/usr/local/bin

After cloning repository to your own local machine go to project root folder and run

cmake .

and then

cmake --build .

to run unit tests for UNIX machines

cd test
./visionary-test --gtest_color=yes

Sample Usage

Motion-Detection

To detect motions from camera

MotionDetector motionDetector;
motionDetector.setCeil(15);
motionDetector.openCamera();
motionDetector.run();

To detect motions from given local video

MotionDetector motionDetector;
motionDetector.setCeil(15);
motionDetector.setVideoPath("YOUR_VIDEO_PATH");
motionDetector.openCamera();
motionDetector.run();

Face-Detection

To detect face from camera

FaceDetector faceDetector("YOUR_CASCADE_PATH",
                          source::type::camera,
                          nullptr);
faceDetector.run();

To detect face from given local video

FaceDetector faceDetector("YOUR_CASCADE_PATH",
                          source::type::video,
                          "YOUR_VIDEO_PATH");
faceDetector.run();

To detect face from given image

FaceDetector faceDetector("YOUR_CASCADE_PATH",
                          source::type::image,
                          "YOUR_IMAGE_PATH");
faceDetector.run();

Face & Eye Detection

To detect eye and face from given image

EyeDetector eyeDetector;
eyeDetector.setEyeCascadePath("YOUR_EYE_CASCADE_PATH");
eyeDetector.setFaceCascadePath("YOUR_FACE_CASCADE_PATH");
eyeDetector.setSourcePath("YOUR_IMAGE_PATH");
eyeDetector.setSourceType(source::type::image);
eyeDetector.run();

To detect eye and face from camera

EyeDetector eyeDetector("YOUR_EYE_CASCADE_PATH",
                        "YOUR_FACE_CASCADE_PATH",
                        source::type::camera,
                        nullptr);
eyeDetector.run();

To detect eye and face from given local video

EyeDetector eyeDetector("YOUR_EYE_CASCADE_PATH",
                        "YOUR_FACE_CASCADE_PATH",
                        source::type::video,
                        "YOUR_VIDEO_PATH");
eyeDetector.run();

Object Tracking

Object tracking from camera

ObjectTracker objectTracker(tracker::type::mil,
                            source::type::camera,
                            nullptr);
cv::Rect2d boundingBox(x, y, width, height);
objectTracker.run(boundingBox);

Object tracking from given local video

ObjectTracker objectTracker(tracker::type::mil,
                            source::type::video,
                            "YOUR_VIDEO_PATH");
cv::Rect2d boundingBox(x, y, width, height);
objectTracker.run(boundingBox);

Sample Videos & Images

Videos

Motion Detection
Face Detection
Face & Eyes Detection

Images

NOTE

OpenCV should be installed to your computer before running, take a look at .travis.yml file for sample installation for macOS and Linux.