CKPCA-HOG-SVM

  为了准确地对监控场景中的运动目标进行语义上的分类，提出了一种基于聚类的核主成分分析梯度方向直方图和二又决策树支持向量机的运动目标分类算法。(In order to accurately monitor the movement of scene targets semantic classification, the clustering based on kernel principal component analysis of gradient direction histograms, and two and a support vector machine decision tree classification algorithm of moving objects.)