Noise reduction and edge detection in digital images


Noise reduction for a digital image often compromises the results of subsequent edge detection due to blurring the edges. However, there is no established quantitative measurement to describe such impact from different noise reduction methods so that they can be compared objectively. In this paper, the author introduces a quantitative measurement for this purpose, and uses this measurement to analyze the effects of various existing noise reduction methods to the edge detection algorithms. The experimental results show that this measurement represents the impact of noise reduction well and matches our empirical judgment without such measurement. Therefore, it can be used as a guide for developing new noise reduction methods that have less negative impact on edge detection.


Computer Science

Document Type

Conference Proceeding

Publication Date


Journal Title

2005 Beijing International Conference on Imaging: Technology and Applications for the 21st Century