A New Edge Detection Arithmetic for Images Based on Fractal Error
This paper provides a new edge detection arithmetic based on fractal error to solve the shortcoming of traditional methods. Image data is transformed from gray space to fractal characteristic space which is also called error space. Then the difference of the dissimilar fractal models which are indications of target and background is used to the edge detection. When the value of fractal error becomes higher, its corresponding pixel is more consistent with FBM model. As a result, there is more possibility that the pixel belonges to the brim parts. The simulation results indicate the algorithm has the excellent performance in anti-noise and little target or multi-target detection.
Xinnan Fan Chunbo Zhao
College of Computer & Information Engineering Hohai University Changzhou, 213022
国际会议
南宁
英文
2007-07-20（万方平台首次上网日期，不代表论文的发表时间）