| 基于区域合并的动态阈值分割算法 |
| |
| 引用本文： | 王萍,苏秀琴,刘雅轩.基于区域合并的动态阈值分割算法[J].光子学报,2004,33(3):378-381. |
| |
| 作者姓名： | 王萍 苏秀琴 刘雅轩 |
| |
| 作者单位： | 中科院西安光学精密机械研究所,西安,710068;中科院西安光学精密机械研究所,西安,710068;中科院西安光学精密机械研究所,西安,710068 |
| |
| 摘 要： | 在传统分块分割再合并阈值法的基础上,引入图像区域合并的思想,以邻近区域阈值相似性作为判断区域合并的依据,借助人机交互式系统,运用待识别目标灰度与完成区域合并的阈值作加权,消除区域灰度不连续,实现目标与背景的有效分离,减少算法的运算量,提高目标识别精度.实验证明,该算法实现简单且行之有效.
|
| 关 键 词： | 阈值 区域合并 相似性 加权 |
| 收稿时间： | 2003-03-10 |
| 修稿时间： | 2003年3月10日 |
Dynamic Thresholding Based Region Combination |
| |
| Institution: | (Xi’an Institute of Optics and Precision Mechanics of CAS, Xi’an 710068) |
| |
| Abstract: | The dynamic thresholding based on region combining is proposed. The conception of region combination is introduced, the threshold of adjacent region is taken as the rule of judging region combining, and the gray level of object that depend on IIPS(Interactive Information Process Systems) is weighed with the threshold of each region. The discontinuity of gray is decreased; the performance of the object recognition is improved. The operation of thresholding is decreased and the precision of object recognition is enhanced. The verification testing shows that the method is simple and valid. |
| |
| Keywords: | Threshold Region combining Similarity Weigh |
| 本文献已被 CNKI 维普 万方数据 等数据库收录！ |
| 点击此处可从《光子学报》浏览原始摘要信息 |
| 点击此处可从《光子学报》下载免费的PDF全文 |
|
