交通警示性图片颜色特征选取仿真

摘要

对交通警示性图片颜色特征特征进行选取,能够有效提高交通图标的警示作用,对人们安全出行十分关键.对交通警示图片颜色特征特征的选取,需要采用K均值聚类方法对交通警示性图片中的颜色特征进行聚类,对聚类函数进行最优处理,完成交通警示性图片颜色特征高精准度选取.传统方法通过对颜色特征分量进行分类最佳排序,并采用不同颜色特征分类的区分性进行验证,但忽略了对图片颜色特征的聚类处理,导致颜色特征选取精度偏低.提出考虑视觉传达约束的交通警示性图片颜色特征选取模型,根据HSI颜色模型完成背景颜色的特性描述,计算滤波器窗口中每个矢量与滤波器窗口中剩余矢量的距离和,对交通警示性图片中的颜色进行聚类,优化聚类函数,对交通警示性图片的类间独立进行处理,完成交通警示性图片中颜色特征选取.实验证明,上述方法能够有效提高交通图标警示效果.%Traditional selection method of color feature of traffic warning image always ignores cluster process for the color feature,which leads to poor selection precision.In order to solve the problem,by considering constraint of visual communication,a new selection model is proposed.According to HIS color model,characteristic description of background color is completed and distance sum of each vector and residual vector in filter window is calculated.Cluster process is carried out for color in the warning image and clustering function is optimized.Inter-class independent of the image is processed.Thus,the selection of color feature is completed.Simulation proves that the meth-od can improve warning effect of traffic icon effectively.