为了对目标进行快速的检测，提出了一种新的基于支持向量机的级联式分类器的构造方法。该级联分类器由若干个线性SVM弱分类器构成，结构简单，分类时间极快。针对级联结构中的每个节点的训练给出了一个新的SVM框架下的二次规划模型，这使得每个节点都有较高的正样本检测率和适当的负样本错检率。-In order to quickly detect the target, a new cascade based on support vector machine classifier constructor. The cascade is controlled by a number of weak linear SVM classifier structure, simple structure, classification of time fast. Cascade structure for the training of each node is given a new SVM quadratic programming model under the fr a mework, which allows each node has a higher detection rate of positive samples and negative samples of appropriate fault detection rate.