Face image processing system and method

Abstract

The invention discloses a face image processing system and method, the system comprising: a model image storage module for storing a model image; a face image acquisition module for acquiring a face image; an image transmission module for transmitting the face image to an image processing module; the image processing module configured to perform face image synthesis; and an image display module configured to display the acquired face image, the recommended model image, and the final composite image. The method mainly detects the face regions in the input image and the reference image and extracts face feature points, and performs triangulation and the like on the two face images according to the same law. The system and method of the invention can separate the image acquisition module fromthe image processing module, facilitates the selection of an optimal acquisition location, and enables a customer to see the acquired face image in real time and select an appropriate facial expression and the best photographing location.

Description

Technical field [0001] The present invention relates to an image processing system and method, in particular to a system and method for face image processing. Background technique [0002] Image processing technology is used in many fields including medical, military, and production. The application of image processing technology in facial image recognition enables people to obtain relevant information more conveniently in many fields, so as to make more accurate judgments on relevant situations. [0003] In many scenarios, such as in a hair salon, customers want to have an intuitive judgment on whether the hairstyle is suitable for them before getting a haircut, so that they can make a more suitable choice. In addition, this method of facial image synthesis also has broad application prospects in privacy protection, virtual fitting, entertainment and leisure. [0004] The main disadvantages of the current existing technology are: [0005] First, if the brightness, contrast, and ton...

AI Technical Summary

Problems solved by technology

[0005] One is that if the brightness, contrast, and hue of the image to be synthesized are inconsistent, the authenticity of the synthesized image will not be high, so that useful information cannot be obtained

[0006] Second, when applied to the above-mentioned barber shop scene, because the hairstyle will block the facial features, the facial features of the customer cannot be seamlessly fitted with the hairstyle of the model, resulting in an unnatural image effect and degrading the customer experience.

[0007] The 3rd, existing technology can't change the facial profile of synthesized image, generally speaking, if the model provided is a round face, but the customer's face is a square face, existing synthesis technology can only make the synthesized image still be a round face

[0009] "Face Swapping under Large Pose Variations: A3D Model Based Approach" by Lin Y, Wang S, Lin Q et al. (in IEEE, International Conference on Multimedia and Expo, 2012) adopts the establishment of 3D head model (3D morphable models, 3dMM ) method, due to the immaturity of the technology, can only generate part of the facial features, which will cause the final composite effect to be unreal, and the method based on this 3D model will take a lot of time and cannot meet most application scenarios

Images