This report describes a technique to automatically augment photographs by adding masks and hats (or ears) to people’s faces and to additionally complement the scene by adding Instagram-like effects such as vignetting, gradient blurring and color filtering. Unlike commercially available applications, this approach can be used to make any photographed or animated face into a mask. The augmentation of the target photograph is fully automatic and takes 1-3 seconds, depending on the size of the mage. As next steps, this algorithm could be implemented on a mobile device in real time for video-chat applications.
The project will be implemented in Matlab and utilize openCV libraries. I aim to create three themes in total, which include a combination of a mask and a mood.