第三章主要讨论了分类器设计过程中的特征空间优化设计问题、分类器设计准则、基本方法和判别函数的知识。
The third chapter mainly introduces characteristic space optimization design, the classification's design criterion, the essential method and the knowledge about knowledge function.
特征空间、几何变换、相似性测度以及优化算法是设计医学图象配准方法时要考虑的四个主要因素。
Many investigator explore in the domain. Medical images registration methods have four main factors: features space, geometry transformation, similarity, and optimizing algorithm.
但是传统的算法未对数据对特征进行优化，亦未考虑图像的空间信息，对噪声图像分割不理想。
However, as the conventional FCM doesn't optimize data in feature space and doesn't involve any spatial information, it is sensitive to noise.
研究提出局部脱耦方法用以简化问题规模、缩小搜索空间，选择特征统计算法进行优化方案的搜索。
In this paper, local decoupling method is put forward to simplify the scale of optimization problem and to reduce its searching scope.
均值漂移算法是一种基于核密度梯度估计的特征空间分析算法，其实质是一种统计优化过程。
Mean shift based image segmentation algorithm is a kind of kernel density estimation based feature space analysis algorithm, and the nature of it is statistical optimization.
均值漂移算法是一种基于核密度梯度估计的特征空间分析算法，其实质是一种统计优化过程。
Mean shift based image segmentation algorithm is a kind of kernel density estimation based feature space analysis algorithm, and the nature of it is statistical optimization.
应用推荐