A bundle adjustment algorithm based on the minimization of the distances of feature matches is also introduced to remove the accumulation of errors.
同时，本文设计出基于特征匹配对距离最小化的捆绑调整算法。
Investigates energy function minimization process for eye feature extraction based on deformable template.
研究眼睛特征提取过程中可变形模板中能量函数优化方法。
Because the proposed angular feature is invariant to rotation, translation and scale transformation, the registration result will converge to global minimization rather than local one.
由于本文提出的角特征不随旋转、平移和尺度的变化而变化，所以配准最终收敛到全局最小值，而不是局部最小值。
The SVM method is based on seeking on the Structural Risk Minimization by few learning samples supporting, and it has important feature such as good generalization and classification performance, etc.
支持向量机方法基于小学习样本条件下，通过寻求结构风险最小，以期获得良好的分类效果和泛化能力。
The SVM method is based on seeking on the Structural Risk Minimization by few learning samples supporting, and it has important feature such as good generalization and classification performance, etc.
支持向量机方法基于小学习样本条件下，通过寻求结构风险最小，以期获得良好的分类效果和泛化能力。
应用推荐