Journal of Frontiers of Computer Science and Technology ›› 2022, Vol. 16 ›› Issue (11): 2587-2595.DOI: 10.3778/j.issn.1673-9418.2103044

• Graphics and Image • Previous Articles     Next Articles

Improved Siamese Adaptive Network and Multi-feature Fusion Tracking Algorithm

LI Rui, LIAN Jirong+()   

  1. College of Computer and Communication, Lanzhou University of Technology, Lanzhou 730050, China
  • Received:2021-03-15 Revised:2021-06-08 Online:2022-11-01 Published:2021-06-16
  • About author:LI Rui, born in 1971, M.S., professor. Her research interests include intelligent information processing, pattern recognition and artificial intelligence.
    LIAN Jirong, born in 1995, M.S. candidate. His research interests include pattern recognition and artificial intelligence.
  • Supported by:
    National Natural Science Foundation of China(61761028);Key Research and Development Program of Gansu Province-Industrial(18YF1GA060)

改进的Siamese自适应网络和多特征融合跟踪算法

李睿, 连继荣+()   

  1. 兰州理工大学 计算机与通信学院,兰州 730050
  • 通讯作者: + E-mail: 649094574@qq.com
  • 作者简介:李睿(1971—),女,甘肃秦安人,硕士,教授,主要研究方向为智能信息处理、模式识别、人工智能。
    连继荣(1995—),男,甘肃定西人,硕士研究生,主要研究方向为模式识别、人工智能。
  • 基金资助:
    国家自然科学基金(61761028);甘肃省重点研发计划-工业类(18YF1GA060)

Abstract:

Aiming at the problem that tracking accuracy and tracking speed are difficult to balance in the current target tracking field. For example, a tracker based on correlation filtering can run at a very high speed, but the tracking accuracy is extremely low; a tracker based on deep learning can achieve high tracking accuracy, but the tracking speed is extremely low. On this basis, an improved Siamese adaptive network and multi-feature fusion target tracking algorithm are proposed. Firstly, the AlexNet network and the improved ResNet network are constructed on each branch of the Siamese network at the same time for feature extraction. Secondly, through end-to-end training at the same time, the tracking problem is decomposed into sub-problems of classifying each position label and returning to the bounding box. Finally, the shallow features and deep features are selected adaptively, and the target recognition and location are carried out based on multi-feature fusion. The proposed algorithm and some existing trackers are tested on the target tracking standard dataset. Experimental results show that the proposed algorithm can achieve high target tracking accuracy and success rate while ensuring tracking speed. At the same time, the algorithm has strong robustness in complex situations such as illumination changes, deformations, and background clutter.

Key words: object tracking, Siamese network, feature fusion, scale adaptation, ResNet network

摘要:

针对当前目标跟踪领域中跟踪精确度和跟踪速度难以平衡的问题,例如基于相关滤波实现的跟踪器能够以很高的速度运行,但跟踪准确性极低;基于深度学习实现的跟踪器能够实现较高的跟踪准确性,但跟踪速度较低。在此基础上,提出一种改进的Siamese自适应网络和多特征融合目标跟踪算法。首先在Siamese网络每个分支上同时构建AlexNet网络和改进的ResNet网络,用于特征提取。其次通过端到端的方式同时进行训练,将跟踪问题分解为分类每个位置标签和回归边界框子问题。最后对浅层特征和深层特征进行自适应选择,并基于多特征融合进行目标识别和定位。将提出的算法与现有的一些跟踪器在目标跟踪标准数据集上进行测试。实验结果表明, 提出的算法能够在确保跟踪速度的同时实现较高的跟踪精确度和成功率。同时,在光照变化、形变、背景杂波等复杂情况下,算法具有较强的鲁棒性。

关键词: 目标跟踪, Siamese网络, 特征融合, 尺度自适应, ResNet网络

CLC Number: