Research&Design Of Example Based Digital Facial Makeup

In ancient times, women used natural dyes to color their lips and cheeks, in order to increases their physical attractiveness. Thus it can be seen that the concept of makeup has a long history. Nowadays, putting on makeup has become many people’s normal routine. With the popularity of intelligent mobile terminals with camera and the development of image processing technology, people have begun to use the applications on intelligent mobile terminal to beautify their own images and even put on digital facial makeup.In this thesis, we study the example based digital facial makeup algorithm deeply. After that, we improve the facial image layer decomposition algorithm and facial image layer composition algorithm. On this basis, we implement a digital makeup App on i OS platform, with the combination of digital makeup algorithm and mobile Internet.The major contributions of this thesis are listed as below:Firstly, we presented the overseas and domestic research status of digital makeup. We study the face detection, facial landmark location and image morphing algorithm in detail before we choose the proper one for our digital makeup algorithm.Secondly, we introduced the facial image layer decomposition algorithm based on the improved Guided Filter and facial image layer composition algorithm. We improved the Guided Filter in order to get different filtering effect in different position, and then we use it for facial image layer decomposition. Due to this improvement, we can both reduce the time complexity of the algorithm and obtain good layer decomposition results. In the aspect of facial image layer composition, we improved the composition algorithm for detail layer, which makes better makeup results.Finally, we implemented a digital makeup App with “makeup in one key” function on i OS platform by combining digital makeup algorithm and mobile Internet technology together. We introduced the example-based super resolution to handle the problem of different resolution between makeup example image and input facial image. Besides, we compare the makeup results of our algorithm with several existing applications.Digital makeup technology has a great application potential in daily life. For examples, it can help comestics E-commerce companies to provide virtual trial for their potential customers; it can help makeup service providers to offer best solution to customers; it can help people to choose proper makeup styles by themself; and people can use digital makeup before they share their pictures to social network.