Code Farm - Portrait Tattoos

This technology takes a portrait photo, a reference portrait tattoo and a skin image as input. The region in the portrait photo that will be transformed to a synthetic portrait tattoo is selected by the user manually. The region in the skin image where the portrait tattoo should be painted is also selected by the user manually. A series of image processing methods including alpha-matting, sketch extraction, facial landmarks detection are applied on the portrait photo to enhance the image and extract the portrait detail information. A colour mapping algorithm based on histogram mapping is designed to calculate the colour for tattooing portrait tattoo image. A tattoo needle model is applied to generate high resolution portrait tattoo images. The size and orientation of the model is decided automatically by the algorithm. The colour used in the tattoo needle is selected from the colour mapping results. The tattoo needle setting is decided by the user in order to generate the portrait tattoo that meets the user’s expectation.