SIFT等局部特征的词袋模型实现。包括K-means聚类，直方图特征的形成，以及KNN分类。-SIFT local features such as word bag model implementation. Including K-means clustering to form histogram features, and KNN classification.