Color Harmonoy Based Multiple Target Image Color Transfer

As the information technology change rapidly today, the application of the computer reached more and more deeply into our daily life details. More and more people tend to consider the application of computer technology to process the image we capture to obtain ideal effect, at the same time, extracting the information we want. from the images. With the rapid development of image processing technology, computer image processing is more and more widely applied. The images in the real world, due to various reasons, such as excessive exposure or color difference of terminal device, are often unable to meet the needs of people. On the basis of previous studies, we propose a method of multi-objective image color transfer to get the image we want.Computer image processing is mainly refers to a series of techniques we use when processing the images we collected in our daily life. Including preliminary processing technology (translation, rotation, de-noising), intermediate processing technology (enhancement, restoration, segmentation) and post-processing techniques (identification, analysis, code etc.). Color transfer belongs to comprehensive processing. Color transfer refers to the use of color feature of images (target image) which look good to improve the color of images (source image) which is waiting to be processed. There are two kinds of color transfer:local color transfer and global color transfer. The color transfer we proposed in this paper allows the user to choose multiple target images or image areas, interacting simply, basing on image segmentation and region matching, to transfer the color of the region they like to improve the characteristics of the image regions waiting to be improved. Finally, considering the compatibility of the color of different images, the images which are improved will be further processed with color harmonious in order to achieve a more naturally and really feeling.This paper mainly has the following two aspects of contributions:1) We integrate the current color transfer method and put forward a new application:the user choose one or more target images or regions freely, based on image segmentation transfer the characteristics of color to the image which needs to be improved, The users have more flexible choice to achieve more ideal effect.2) We apply color harmony theory to color transfer. Considering the source of multiple target images, the scene, illumination, imaging principle and other factors, after the color transfer, the image appears to be unnatural and unreal, our method have an better effect.