基于小波分层的多方向图像边缘检测

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的体现出优越性. 本文结合目前先进的小波理论, 将图像进行小波变换, 得到具有单一性边缘的子图像, 再将传统边缘检测算子的方向性与这些子图像对应起来分别进行检测, 最后分别得到不同强度(层次)图像边缘, 并且这些边缘可以进行合成, 得到较好的图像边缘. 该算法操作简单, 具有很好的效果.

Multidirectional Image Edge Detection Based on Wavelet Laid

Image edge detection plays a very important role in image processing. It can serve as a basis for modal distinction, image division and image scene analysis. Traditional edge image detection is easy to do and its direction adaptation is very good. However, the image edge is diverse, e.g. it has different directions, different robustness, etc, which limits its uses, But it also has its own disadvantages. This paper is an attempt to get single-edge branch images using wavelet transform, These branch images can be detected separately by edge detection operation, whose direction adaptation is very close to that of single-edge detection. Image edges of different layers can be achieved finally. And these edges can be synthesized, thereby giving rise to better image edges, The operation is easy and the result is good.