一种利用空间梯度信息进行阈值分割与边缘优化的新算法
A Novel Thresholding Segmentation and Edge Optimization Method Using Spatial Gradient Information
李晓兵 1阎志军 1许梅梅1
作者信息
- 1. 南京航空航天大学应用物理系,南京 211106
- 折叠
摘要
提出一种结合灰度直方图与空间梯度信息提取低对比图像边缘的新方法.该方法分为三部分:在消噪过程中,通过分析边缘点与噪点特征在消除噪点的同时尽可能地保留边缘点;结合灰度直方图与空间梯度信息获得分割阈值;再次应用空间梯度信息对初始边缘进行优化.实验结果显示,该方法对具有不同特征的低对比度图像都展示出较好的结果.
Abstract
Proposes a novel edge detection method combining gray histogram with spatial gradient information for low contrast image.This method gets improved results via three key steps:in de-noising process,preserves most of the edges by analyzing the character of noisy point and edge point;combined the spatial gradient information with gray histogram to obtain the thresholds of segmentation;applies the spatial gradient information again to optimize the extracted edges.Experimental results show that the proposed method obtains better results in low contrast images of various characters.
关键词
低对比度图像/边缘提取/灰度直方图/空间梯度信息Key words
Low Contrast Image/Edge Detection/Gray Histogram/Spatial Gradient Information引用本文复制引用
出版年
2015
NETL
NSTL
维普
万方数据