基于蜂鸟羽毛的色彩提取及应用

Color extraction and application based on hummingbirds' feathers

鸟类羽毛色彩进化了上万年,形成了一套天然美感的搭配法则,能够为设计师提供设计灵感。为探寻其中的配色规律并加以利用,文章基于蜂鸟羽毛的色彩进行了一系列的仿生色彩提取研究。首先针对K-MEANS聚类算法,对比分析了色相模式与灰度模式两种不同的初始聚类中心的设置方法,并在此基础上提出了一种经过优化的初始聚类中心的设置方法;然后使用优化过后的K-MEANS聚类算法,对10种蜂鸟共200张羽毛色彩照片进行羽毛色彩的主色系提取。通过三次聚类计算,探讨分析蜂鸟羽毛色彩的配色规律,制得相应的占比配色色卡,为设计师在以蜂鸟羽毛色彩为灵感来源的配色方案上提供了一定的理论依据。

The colors of birds＇ feathers have evolved for thousands of years,and a set of collocation tules witrh natural beauty has formed,which can provide designers with design inspiration. In order to explore the color matching law and use it,a series of bionic color extraction and researches have been made based on colors of hummingbirds＇ feathers. Firstly,the K-MEANS clustering algorithm was used to analyze the setting method of two different initial clustering centers（ hue mode and gray mode）. On this basis,an optimized setting method of initial clustering center was proposed. Then,this optimized K-MEANS clustering algorithm was used to extract the main color lines of feather colors from a total of 200 feather color photos of 10 species of hummingbirds. Through the clustering calculation for three times,the matching rules of colors of hummingbirds＇ feathers were analyzed and corresponding proportion color cards were gained. This paper provides certain theoretical basis for the designers to apply the color of hummingbirds＇ feathers as the source of color scheme inspiration.