双特征模型核相关滤波目标跟踪算法  

Kernel correlation filtering algorithm based on a dual-feature model

作　　者：孟琭[1] 李诚新 Meng Lu;Li Chengxin(College of Information Science and Engineering,Northeastern University,Shenyang 110004,China)

机构地区：[1]东北大学信息科学与工程学院

出　　处：《中国图象图形学报》2019年第12期2183-2199,共17页Journal of Image and Graphics

基　　金：国家自然科学基金项目(61973058)~~

摘　　要：目的基于深度学习的目标跟踪算法,利用卷积深层作为特征,虽然精度高但无法做到实时跟踪;基于相关滤波的跟踪算法,利用HOG(histogram of oriented gridients)、CN(color name)和颜色直方图作为特征,速度快但精度较差。为兼顾目标跟踪算法的实时性与准确性,提出了一种基于双模型核相关滤波算法。方法提出了自适应的双特征模型选择机制,主特征模型采用浅层纹理特征HOG,辅助特征模型采用包含深层语意信息的CNN(convolutional neural networks)特征,二者协同作用,产生更加稳定的相关滤波器。为了提高算法的速度,采用主成分分析(PCA)技术对高维的CNN特征进行降维,并通过尺度优化、最优解求解方式优化等方法提高跟踪算法的准确性。结果在公开数据集OTB-2013上,本文算法与目前先进且速度能达到实时的SiamFC (fully-convolutional Siamese networks)、MEEM (multiple experts using entropy minimization)、SAMF (scale adaptive multiple features)、DSST (discriminative scale space tracking)等跟踪算法进行比较,一次成功率(OPE)结果显示,本文算法在距离精准度指标中综合排名第一,与KCF(kernel correlation filter)算法相比,本文算法的距离精准度提高了25.2%,重叠成功率提高了25.6%,平均速度达到38帧/s。结论本文提出的双模型自适应机制,针对主特征模型的置信响应自适应调用最优模型策略,并且实时更新模型,在综合考虑跟踪准确性和跟踪实时性的情况下,本文提出的目标跟踪算法的性能优于目前的跟踪算法。Objective The target tracking algorithm that is based on deep learning and uses deep convolution features is highly accurate, but it cannot be tracked in real time nor applied to actual situations. The deep convolutional features of convolutional neural networks(CNNs) contain advanced semantic information. Even when the target appearance model has serious interference, such as illumination variation, deformation, and other interference factors, the deep convolution features still exhibit an accurate discriminative performance in the target. Although the tracking algorithm based on correlation filtering is fast(up to several hundred frame/s), it is inaccurate. The algorithm uses the histogram of oriented gradient(HOG), color name(CN), and color histogram as statistical features to calculate the correlation of two image blocks. The position with the highest correlation is the predicted position. To balance the real-time tracking capability and accuracy of the target tracking algorithm, this study proposes a dual-model kernel correlation filtering algorithm based on the combination of the accuracy of the deep convolution feature algorithm and the speed of the correlation filtering algorithm. Method An adaptive dual-feature model selection mechanism is proposed. The dual model consists of main-and auxiliary-feature models. The main-feature model adopts a shallow texture feature. The dimension of the HOG feature is relatively low. Thus, it has a high calculation speed. The main-feature model is used for the real-time tracking of video sequences with clear texture contour features, and the kernel correlation function of the correlation filter of the main-feature model uses the Gaussian kernel function. The auxiliary-feature model employs CNN features containing deep semantic information. When serious interference factors, such as illumination variation, occlusion, and deformation, occur in video sequences, the auxiliary-feature model with deep CNN features is used to determine and correct the target position because su

关 键 词：目标跟踪 自适应特征 卷积神经网络 相关滤波 主成分分析 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]