BP神经网络人脸检测研究

BP Neural Network of Human Face Examination Testing

使用了反向传播神经网络和支撑向量机两种技术分别完成了人脸检测算法,对比了两种方法在人脸检测上的优劣.实验证明,在本文描述的环境下,支撑向量机技术能够更有效地处理多人脸、复杂背景的定位,算法高效、稳定,具有较强鲁棒性.

Using BP neural network and SVM to complete human face testing separately, this article compares the advantage and disadvantage of the two methods. According to the experiments, under the circumstances as described in this article, SVM technique can handle the positioning of more than one human faces and complex background more effectively. Moreover, this method is efficient and stable, with a rather strong robust.