>> 自然科学版期刊 >> 2009年01期 >> 正文
基于小波变换的Canny检测算法
供稿: 余发山;李睿;陈步威 时间: 2019-04-22 次数:

作者:余发山;李睿;陈步威

作者单位:河南理工大学电气工程学院

摘要:在图像处理中,边缘检测具有很重要的作用,它可作为模式识别、图像分割及图像场景分析的基础.传统的图像边缘算法具有算法简单,方向适应性强的优势,然而由于图像边缘具有多样性(方向的不一致性、边缘强弱的不相同等),这些传统算法不能很好地体现出优越性.论文提出利用小波变换对图像进行分解,增强和重构后,再利用传统的Canny算子对变换后的图像进行边缘检测,以缓解图像处理中噪声抑制与精细边缘检测的矛盾.结果证明,该方法提取的图像边缘失真小,能有效地拟制噪声,对多边缘特征的图像具有良好的适应性.

关键词:小波变换;Canny算子;边缘提取;

DOI:10.16186/j.cnki.1673-9787.2009.01.026

分类号:TP391.41

anny operator detection based on wavelet transform

Abstract:Image edge detection plays a very important role in image processing, it can serve as a basis for modaldistinction, image division and image scene analysis.Traditional edge image detection is easy to do and its direction adaptation is good.However, the image edge is diverse.E.g.it has different directions, different robustness, etc, which limits its uses.In the paper, image deposition, enhance and restruction is analysed based on wavele transform, then image edge is detected based on traditional Canny operator.The experiment results show that the proposed method has a better ability in precise location and noise restraining, which adapts to the images with more edge characteristics.

最近更新