基于人眼视觉特性的景观图像高频细节增强方法  

A High Frequency Detail Enhancement Method of Landscape Image Based on Human Visual Characteristics

作　　者：夏弘睿[1] 赵静 XIA Hong-rui;ZHAO Jing(Department of Art and Design,Ma’anshan Normal College,Ma’anshan 243000,China)

机构地区：[1]马鞍山师范高等专科学校艺术设计系,安徽马鞍山243000

出　　处：《宜春学院学报》2022年第12期39-43,共5页Journal of Yichun University

基　　金：安徽省2019年度高校优秀青年人才支持计划项目(编号:gxyq2019193);安徽省教学研究重点项目“本科层次职业教育人才培养思路研究——以“职教20条”引领艺术设计类专业为例”。

摘　　要：由于图像获取设备、大气环境等多种因素的影响,致使景观图像存在高频细节模糊、人眼难以分辨细节信息等现象。为了解决上述问题,提出基于人眼视觉特性的景观图像高频细节增强方法。划分景观图像视觉区域,提取景观图像视觉显著区域。应用双边滤波算法剥离出高频细节区域的景观图像。利用改进自适应直方图均衡化算法(Improved Adaptively Increasing the Value of Histogram Equalization, IAIVHE)重新定义景观图像概率密度函数,通过累积映射实现景观图像高频细节的增强。实验数据显示:提出方法获得的信息熵最大值为8.45,峰值信噪比数值最大值为22.48 dB,证实了提出方法高频细节增强效果更佳。Due to the influence of image acquisition equipment, atmospheric environment and other factors, there are many phenomena in landscape images, such as high-frequency detail blur.It is difficult for human eyes to distinguish detail information.In order to solve the above problems, a high-frequency detail enhancement method of landscape image based on human visual characteristics is proposed.Firstly, it needs to divide the visual area of the landscape image and extract the visual significant area of the landscape image.Then, the landscape image of high-frequency detail area is stripped by using bilateral filtering algorithm.Finally, the probability density function of landscape image is redefined by using Improved Adaptively Increasing the Value of Histogram Equalization(IAIVHE),and the high-frequency details of landscape image are enhanced through cumulative mapping.The experimental data show that the maximum value of information entropy obtained by the proposed method is 8.45.The maximum value of peak signal-to-noise ratio is 22.48 dB.It proves that the proposed method has a better effect of high-frequency detail enhancement.

关 键 词：景观图像 高频细节 人眼视觉特性 图像增强处理 滤波去噪 

分 类 号：TP399[自动化与计算机技术—计算机应用技术]