基于OpenCV的距离估计的数据可视化研究

Research on Data visualization based on OpenCV distance Estimation

设计了一种基于OpenCV的人脸姿态与距离估计系统,该系统用MFC建立人脸识别系统框架,并结合OpenCV库,实现了通过摄像头实时获取人脸信息并进行姿态估计和距离估计的功能。其中,在图像特征处理中,通过结合使用PCA算法来降低由原LBP算子所获取的特征向量的维数,在获得人脸特征后,再通过SVM方式加以确定和分析,以此提升了距离估计的精确度。该系统最终将获取到的数据进行可视化,同时可将这项技术扩展到电视机、手机以及pad等电子设备,并在生活中得到更加实际的应用。

In this paper,a face pose and distance estimation system based on OpenCV is designed.The system uses MFC to establish the face recognition system frame,and combined with OpenCV library,realizes the real-time acquisition of face information through the camera and the function of attitude estimation and distance estimation.Among them,in the image feature processing,PCA algorithm is combined to reduce the dimension of the feature vector obtained by the original LBP operator.After the face features are obtained,SVM method is used to determine and analyze them,so as to improve the accuracy of distance prediction.The system will eventually visualize the data and extend the technology to TV sets,mobile phones and pads for more practical applications.