一种基于形态学的有噪彩色图像边缘检测方法

摘要

Edge detection is the foundation of image analysis and processing which include image segmentation, image enhancement , image restoration, pattern recognition, image compression and so on. Edge detection is also a hot topic in the field of digital image processing. The method of noise color image edge detection based on morphology in HIS space is proposed. By improving the traditional morphological gradient edge detection, the edge information of H, S, I is calculated, then the edge information of color image is obtained by combining three components' edge information according to their specific weights. The experimental result shows that the edge which detected by the proposed method meets the visual edge, this algorithm is better than traditional ones in the terms of resistance to noise and is able to keep more complete outline of the original color image and has better practicality and versatility.%边缘检测技术是图像分割、图像增强、图像复原、模式识别、图像压缩等图像分析和处理的基础,也是图像处理领域研究的热点问题.本文提出基于HSI颜色空间形态学的有噪彩色图像边缘检测方法.在传统形态学梯度检测边缘的基础上进行改进,计算H、S、I三个分量的边缘信息,然后根据H、S、I所占比重对三分量进行加权融合得到彩色图像边缘.实验结果表明,这种方法所检测的边缘符合视觉边缘,在抗噪声方面的效果比传统方法更佳,能够更完整地保留原彩色图像的轮廓,有更好的实用性和通用性.