Local thresholding approach is slow and prone to give ghost image. We present a hybrid threshold approach which can avoid these disadvantages. The way of the hybrid approach to synthesize global thresholding and local area information and to construct and extended membership function to describe how much the pixel belongs to the stroke or background is studied. Simulation results indicate that the hybrid approach is robust to noise and makes the stroke connectivity better.
|Number of pages||4|
|Journal||Xibei Gongye Daxue Xuebao/Journal of Northwestern Polytechnical University|
|Publication status||Published - 1 Aug 2002|
- Document image
- Membership function
ASJC Scopus subject areas