一种基于逆Bilateral滤波的边缘提取方法

陈亮, 杨宁, 杜亚勤, 郭雷

西北工业大学自动化学院



摘要:

为了增强图像中边缘的响应,进而获得具有鲁棒性的边缘提取算法,提出了一种逆Bilateral滤波的图像预处理方法,将图像中的边缘很好地显现出来,在此基础上添加策略进行边缘提取.Bilateral滤波器能够进行边缘保持,鉴于其在低频和高频的良好响应,给出新的含义,构造逆Bilateral滤波器进行边缘增强处理,然后在梯度模图像上,采用自适应阈值获得和非最大抑制策略,提取图像中的边缘.实验证明,逆向思维的应用能够提高光照不一致的条件下的边缘提取效果,以及增强对噪声的低敏感性。

关键词:  图像处理  边缘提取  逆Bilateral滤波  梯度模图像

DOI：10.11918/j.issn.0367-6234.2010.11.031

分类号:TP391.41

基金项目:国家自然科学基金资助项目(60675015)



An edge extraction method based on inverse bilateral filtering

CHEN Liang, YANG Ning, DU Ya-qin, GUO Lei

School of Automation,Northwestern Polytechnical University,Xi’an 710129,China

Abstract:

To enhance the edge response of the image and then extract the edge robustly,a pre-processing method based on inverse bilateral filtering is proposed,which will make the edge appear perfectly.For bilateral filter having good capability to preserve the edge and good response on low and high frequencies,a new concept for inverse bilateral filter is constructed.And then,strategies of self-adaptive threshold and non-maximum suppress are adopted to extract the edge from the gradient module map of enhanced image.Experiments show that the proposed method has better extraction result with variable illuminance and less sensibility to the noise.

Key words:  image processing  edge extraction  inverse bilateral filtering  gradient module map