系统工程与电子技术

• 软件、算法与仿真 • 上一篇    下一篇

基于自适应分数阶微积分的图像去噪与增强算法

李博, 谢巍   

  1. 华南理工大学自动化科学与工程学院, 广东 广州 510641
  • 出版日期:2016-01-12 发布日期:2010-01-03

Image enhancement and denoising algorithms based on adaptive fractional differential and integral

LI Bo, XIE Wei   

  1. Automation Science and Technology, South China University of Technology, Guangzhou 510640, China
  • Online:2016-01-12 Published:2010-01-03

摘要:

针对传统的图像去噪算法容易忽视图像纹理细节的问题,首先提出一种全局自适应分数阶积分去噪算法。该算法可以在去除图像噪声的同时,对图像的纹理进行一定的保留。其次在全局自适应分数阶算法的基础上,针对一类低强度椒盐噪声提出另一种基于小概率策略的自适应分数阶微积分图像去噪与增强算法,该算法将图像中噪声点的出现视为小概率事件并进行分割,然后再采用自适应分数阶积分对噪声点进行处理的同时,采用自适应分数阶微分对图像的纹理进行增强和保留。实验结果表明,两种方法都可以达到较好的去噪效果,其中基于小概率策略的自适应分数阶算法在去噪的同时更具有增强图像的边缘的效果。

Abstract:

Due to the problem that traditional algorithms would be easy to ignore the texture of the image while image denoising, a new image denoising algorithm based on global adaptive fractional integral is proposed, which preserves image texture information while removes the noises of image. Secondly, the image enhancement and denoising algorithm based on adaptive fractional differential and integral using small probability strategy is proposed for low intensity salt and pepper noise image. The appearance of salt and pepper noise points is regarded as a small probability event, and the adaptive fractional integral algorithm is used to process the image’s noises while the adaptive fractional differential algorithm is used to enhance the texture of the image. Experimental results show that both the two presented methods have good image denoising effect, and the second algorithm has better effect than other methods on enhancing the image edges.