1) feature fusion selection
特征融合约减
2) feature reduction
特征约减
1.
Experimentalresults show that intrusion feature set selected by the feature reduction algorithm is better, an.
提出一种基于AdaBoost的入侵特征约减算法,利用该算法约减入侵特征中的冗余特征,构造Ada-加权和Ada-域值分类器,并与支持向量机分类器进行对比。
3) salient points reduction
特征点约减
4) features fusion
特征融合
1.
Comparing features combination with features fusion in Chinese named entity recognition;
中文名实体识别中的特征组合与特征融合的比较
2.
With the above improved particle filter algorithm,an adaptive multiple features fusion target tracking method is proposed.
该算法根据目标状态与粒子状态之间的距离,利用核函数产生权值对粒子进行二次加权,根据粒子的二次加权结果进行粒子重采样;以改进的粒子滤波算法为框架,提出了一种自适应多特征融合目标跟踪方法,利用相似性度量动态地评价特征对目标与背景的区分能力,并自适应地计算特征融合权重,以适应目标跟踪过程中目标与背景的变化,提高目标跟踪的鲁棒性。
5) feature fusion
特征融合
1.
Research on model of remote sensing image rebuilding based on image feature fusion;
基于影像特征融合的遥感图像重构模式研究
2.
Study of algorithm for gait recognition fased on feature fusion;
基于特征融合的步态识别算法研究
3.
Algorithm for gait recognition based on feature fusion;
基于特征融合的步态识别算法
6) fusion feature
融合特征
1.
This paper defines four features of a target based on IR radiation characteristics, The fusion feature was presened after the four features were normalized and weighted.
算法依据红外目标的辐射特性与背景之间的相关关系,得到目标的灰度和梯度特征量,通过融合处理,得到加权归一化的融合特征矢量。
2.
Methods Three features of wave feature,wavelet feature,fusion feature were selected as the feature vectors for 35 normal subject ECGs,and correlation coefficient discriminate analysis was used for human identification.
方法选取35位健康人ECG波形的波形特征、小波特征和融合特征作为特征向量,通过相关系数阈值法进行ECG身份识别。
补充资料：减约
1.俭省节约。
说明：补充资料仅用于学习参考，请勿用于其它任何用途。
参考词条