Computing Aesthetics And Color Migration Based Color-Assessment And Application Of Watercolor Images

The purpose of the computational image aesthetics research is to endow computer with the ability to assess the aesthetics value of images as human beings do.The results can be used in many fields,for example fusing the subjective perception,image aesthetics evaluation,image aesthetics retouching,photograph aesthetics prediction and art works style analysis.The research of this paper focuses on still life water-color images.First the research divided the image into Entire Region and Key Region,and extracted effective low-level visual features,high-level aesthetic features and regional features;then based on people's subjective aesthetic score,it used machine learning methods to establish an image beauty classifier and a classify-evaluation model of aesthetics score.With the model people can achieve a automatic cla ssification,and can automatically evaluate the aesthetic score of the image.Computational image aesthetics is a new interdisciplinary advanced topic with good developing prospect,while it involves different subjects,including Aesthetics,Art,Computer Science,Psychology,and so on.This paper will use the computing aesthetic theories and techniques to calculate the score of watercolor works in HSV color space.We use the technology of Color Migration to migrate the high-aesthetic-score color to new pictures.This paper will complete color migration of the watercolor works in l?? color space,realizing analysis,extraction and application of color material of watercolor works.