A Method of Eye Image Processing Based on Image Editing

Abstract

The invention discloses an eye image processing method based on image editing. Firstly, a palpebra superior boundary of a reference image serves as a key point, and mapping is carried out on a single-fold eyelid face input image through a thin-plate spline function to adjust the shape and the ratio of an eye. Secondly, a double-fold eyelid folded area of the reference image is combined into the input face image seamlessly through Poisson image editing, and the optimized effect of integrating the shape of the eye of the input image and the features of a double-fold eyelid of the reference image is simulated. The eye image processing method based on the image editing is not only capable of simulating the shapes and appearances of various natural double-fold eyelids according to the difference of reference images, but also simple in realization process, and has great significance for face photo entertainment, cosmetic design and plastic preoperative predetermination.

Description

technical field [0001] The invention relates to an image processing method, in particular to an eye image processing method based on image editing. Background technique [0002] The eyes are one of the most important organs of the human face. In the aesthetic concept of modern people, it is generally believed that the double eyelid shape of the face is more beautiful than the single eyelid shape. In order to directly simulate the shape effect of double eyelids without real plastic surgery, researchers have proposed a variety of methods, which can be mainly divided into two categories. [0003] The first type is an image-based method, where the user can adjust the position, brightness, and width of the line after manually drawing the double eyelid crease line on the image. The ByuPAQ system implemented by Zou Beiji et al. also uses the curve drawn on the image to represent the double eyelid folds. In order to make the "double eyelid" look natural, the author proposes an imag...

AI Technical Summary

Problems solved by technology

The cutting line of this method is to draw a plane curve on the frontal image, and then project it onto the model mesh of the upper eyelid, which will have a certain error with the real face surface

[0005] The first type of method is to draw a curve on the image to represent the double eyelid fold. It is easy to implement and the amount of calculation is small, but the simulation effect is less realistic.

The second type of method simulates the shape effect of double eyelids by changing the grid model to generate a 3D model of "double eyelid folds". The simulation effect is improved to a certain extent, but it is necessary to re-establish the 3D model of "double eyelid folds" for each different user , the implementation process is complex

In addition, the current method does not focus on considering the changes in the shape and proportion of the eyes after the single eyelid shape is changed to double eyelid shape, and the realism of the simulated double eyelid shape effect needs to be further improved

Images