融合背景图像信息和多特征压缩的相关滤波跟踪算法

针对相关滤波类跟踪算法目标背景图像信息利用率较低、目标特征表达能力较弱的问题，提出了一种融合背景图像信息的多特征压缩跟踪算法。首先，在上下文感知滤波器的基础上，将背景图像信息加入位置滤波器。其次，提取颜色名(Color Name, CN)特征与梯度直方图(Histogram of Oriented Gradient, HOG)特征，使用最大响应因子及平均峰相关能量(Average Peak-to-Correlation Energy, APCE)评估跟踪结果的可信度，实现两种特征的自适应融合。最后，利用特征降维简化模型的复杂度，实现算法运行速度的提升。实验结果表明，改进后的算法在遮挡、形变、尺度变化等复杂环境下均具有较高的鲁棒性，其跟踪精度和成功率指标均优于DSST及其他主流的跟踪算法，并且仍保持了实时性。

Correlation Filter Tracking Algorithm Based on Background Image Information Fusion and Multi-feature Compression

As the traditional tracking algorithm extracted few features and had low utilization of target background image information, a correlation filter tracking based on background image imformation fusion and multi-feature compression is proposed in this paper. Firstly, the surrounding background image information is integrated into the translation filter based on the context-aware filter. Secondly, the color names (CN) features and the histogram of oriented gradient (HOG) features are extracted, and the methods of maximum response factor and average peak-to-correlation energy (APCE) are used to evaluate the credibility of the tracking results and fuse the response maps adaptively. Finally, feature dimension reduction method is used to simplify the complexity of the model and improves the running speed of the algorithm. The experimental results show that the improved algorithm has a higher robustness under the complicated environment such as occlusion, deformation and scale variation, etc. The improved algorithm performs superiorly compared with DSST and other main tracking algorithm in terms of tracking accuracy and success rate, with the advantage of real-time tracking.