Adaptive Segmentation Algorithm of Complicated Image
There are a number of similar objects and noises in complicated image,with which classical watershed has no corresponding mechanism to cope.The paper proposes adaptive segmentation algorithm for complicated image.Firstly two windows adaptive binary technique is adopted to generate binary image.Seed region of each object is extracted by distance transform.Seed regions grow outward continuously in parallelism until the boundary of segmented objects is met.The algorithm resolves over segmentation and under segmentation problem of watershed and improves the accuracy of segmentation.
distance transforming image segmentation region growing adaptive threshold
Zhu Hong Zhang Guoying
Dept of Computer Science and Technology China University of Mining & Technology,Beijing Beijing,China
国际会议
太原
英文
65-68
2011-02-26（万方平台首次上网日期，不代表论文的发表时间）