Detecting the moving vehicles in jittering traffic scenes is a very difficult problem because of the complex environment.Only by the color features of the pixel or only by the texture features of image cannot establis...Detecting the moving vehicles in jittering traffic scenes is a very difficult problem because of the complex environment.Only by the color features of the pixel or only by the texture features of image cannot establish a suitable background model for the moving vehicles. In order to solve this problem, the Gaussian pyramid layered algorithm is proposed, combining with the advantages of the Codebook algorithm and the Local binary patterns(LBP) algorithm. Firstly, the image pyramid is established to eliminate the noises generated by the camera shake. Then, codebook model and LBP model are constructed on the low-resolution level and the high-resolution level of Gaussian pyramid, respectively. At last, the final test results are obtained through a set of operations according to the spatial relations of pixels. The experimental results show that this algorithm can not only eliminate the noises effectively, but also save the calculating time with high detection sensitivity and high detection accuracy.展开更多
mean shift目标跟踪使用任一个单独特征都很难对大幅度的光照、背景变化和车辆大范围运动具有强鲁棒性,如单一的颜色特征对光照较为敏感,光照突变容易导致目标丢失。当背景颜色和目标颜色相近时也易造成目标丢失,因此利用目标的单一特...mean shift目标跟踪使用任一个单独特征都很难对大幅度的光照、背景变化和车辆大范围运动具有强鲁棒性,如单一的颜色特征对光照较为敏感,光照突变容易导致目标丢失。当背景颜色和目标颜色相近时也易造成目标丢失,因此利用目标的单一特征信息往往很难实现鲁棒的目标跟踪。文章提出基于颜色和LBP多特征mean shift跟踪方法,使跟踪结果不再过分依赖某一特征,增强了对背景变化、目标大范围运动的鲁棒性。展开更多
微表情是一个人试图隐藏内心真实情感却又不由自主流露出的不易被察觉的面部表情。与一般面部表情相比,微表情最显著的特点是持续时间短、强度弱,往往难以有效识别。文中提出了一种基于LBP-TOP(Local Binary Pattern from Three Orthogo...微表情是一个人试图隐藏内心真实情感却又不由自主流露出的不易被察觉的面部表情。与一般面部表情相比,微表情最显著的特点是持续时间短、强度弱,往往难以有效识别。文中提出了一种基于LBP-TOP(Local Binary Pattern from Three Orthogonal Planes)特征和支持向量机(Support Vector Machine,SVM)分类器的微表情识别方法。首先,采用LBP-TOP算子来提取微表情特征;然后,提出一种基于ReliefF与局部线性嵌入(Locally Linear Embedding,LLE)流形学习算法相结合的特征选择算法,对提取的LBP-TOP特征向量进行降维;最后,使用径向基函数(Radial Basis Function,RBF)核的SVM分类器进行分类,将测试样本图像序列的微表情分为5类:高兴、厌恶、压抑、惊讶、其他。在CASME Ⅱ微表情数据库上采用"留一人交叉验证"(Leave-One-Subject-Out Cross Validation,LOSO-CV)的方式进行了实验,可得到58.98%的分类准确率。实验结果表明了该算法的有效性。展开更多
基金Project(61172047)supported by the National Natural Science Foundation of China
文摘Detecting the moving vehicles in jittering traffic scenes is a very difficult problem because of the complex environment.Only by the color features of the pixel or only by the texture features of image cannot establish a suitable background model for the moving vehicles. In order to solve this problem, the Gaussian pyramid layered algorithm is proposed, combining with the advantages of the Codebook algorithm and the Local binary patterns(LBP) algorithm. Firstly, the image pyramid is established to eliminate the noises generated by the camera shake. Then, codebook model and LBP model are constructed on the low-resolution level and the high-resolution level of Gaussian pyramid, respectively. At last, the final test results are obtained through a set of operations according to the spatial relations of pixels. The experimental results show that this algorithm can not only eliminate the noises effectively, but also save the calculating time with high detection sensitivity and high detection accuracy.
