基于改进型深度LLE和随机森林的人脸检测算法

针对人脸检测问题的特点,提出一种基于改进型深度LLE(Locally Linear Embedding)算法和随机森林相结合的人脸检测算法。首先,通过采集图像的深度信息,结合图像的颜色信息,构建三维图像信息数据库,再通过改进的LLE算法得到最优降维结果,按一定比例选取训练集,输入随机森林算法建立数据分类器;最后,将测试集输入到训练完成的分类器中,实现人脸图像的检测。选取Yale,JAFFE 2类数据集与传统算法进行对比实验,验证算法的优越性和可行性。实验结果表明:所提出的算法可以有效地完成人脸检测,检测率高于传统算法7%左右。

Face Detection Algorithm Based on Modified Depth LLE and Random Forest

For the characteristics of face detection, a novel face detection algorithm based on modified depth Locally Linear Embedding(LLE) and Random Forest is proposed. Firstly, the depth information of images are collected by Kinect, and the three-dimensional image data base can be established by the depth information and colour information. Secondly, the dimension of data sets are reduced by modified LLE, and the optimal results of data dimension reduction can be gotten. The training sets are gotten by the proportion of data sets, and data classifier can be gotten by Random Forest. Finally, the test sets are inputted, and the face detection can be achieved. The two classes of data sets are selected as the experimental data, which consist of Yale and JAFFE. The experiment results show that the proposed method not only has a great effect to achieve face detection, but the detection rate is higher than the traditional algorithms about 7%.