| 数字图像最佳灰度门限的确定 |
| |
| 引用本文： | 罗红宇,高鹏. 数字图像最佳灰度门限的确定[J]. 吉林工程技术师范学院学报, 2003, 0(9) |
| |
| 作者姓名： | 罗红宇 高鹏 |
| |
| 作者单位： | 吉林工程技术师范学院信息工程学院 吉林长春130052(罗红宇)，中科英华电工器材有限公司 吉林长春130012(高鹏) |
| |
| 摘 要： | 利用基于模式识别决策理论的最大类间距准则,给出了求解最佳分割灰度门限的方法。以三氧 化二铝衬底芯片为算例,利用UMA测度对最大类间距准则和标准双峰准则在分割质量上的差异进行定 量比较。结果表明:最大类间距准则在分割质量和鲁棒性两方面性能均优于标准双峰准则。
|
| 关 键 词： | 数字图像处理 图像分割 最大类间距准则 门限值 |
Method based on maximum class separation distance criterion for gray threshold of digital image of LED''''s chips |
| |
| LUO Hongyu,Gao Peng. Method based on maximum class separation distance criterion for gray threshold of digital image of LED''''s chips[J]. Journal of Jilin Teachers Institute of Engineering and Technology(Natural Sciences Edition), 2003, 0(9) |
| |
| Authors: | LUO Hongyu Gao Peng |
| |
| Abstract: | In this paper, the method for the optimal gray threshold is given, which is based on the maximum class separation distance criterion of pattern recognitive decision theory.The chips that are lined with alumina, are taken for numeric example. By UMA measure the segmentation quality is contrasted between maximum class separation distance criterion and bi - modal histogram method. The result states that maximum class separation distance criterion is obviously superior to bi - modal histogram method in the segment quality and the rosbustness. |
| |
| Keywords: | digital image processing image segmentation maximum class separation distance criterion threshold value |
| 本文献已被 CNKI 等数据库收录！ |
