(2)Target of Data Pre-process ·Data selection数据选择 Select the most efficient data for analysis 选择最有效数据进行特定目的分析 ·Data discretization数据离散化 Process continuous data to discrete data 将连续数据进行离散处理 ·Data feature extraction数据特征提取 Abstract original features into a set of obvious physical significance (Gabor,geometric feature [angular point,invariant]. texture [LBP HOG])or statistical significance properties. 将原始特征转换为一组具有明显物理意义(Gabor、几何特征[角点、不 变量]、纹理[LBP HOG])或统计意义的特征 ATA 28 Copyright 2019 by Xiaoyu Li.Copyright © 2019 by Xiaoyu Li. 28 (2) Target of Data Pre-process  Data selection数据选择 Select the most efficient data for analysis 选择最有效数据进行特定目的分析  Data discretization 数据离散化 Process continuous data to discrete data 将连续数据进行离散处理  Data feature extraction 数据特征提取 Abstract original features into a set of obvious physical significance (Gabor, geometric feature [angular point, invariant], texture [LBP HOG]) or statistical significance properties. 将原始特征转换为一组具有明显物理意义（Gabor、几何特征[角点、不 变量]、纹理[LBP HOG]）或统计意义的特征