航空兵器2024，Vol.31Issue(1)：1-12,12.DOI:10.12132/ISSN.1673-5048.2022.0226
基于深度神经网络的目标跟踪算法综述
A Review of Target Tracking Algorithm Based on Deep Neural Network
摘要
Abstract
Target tracking is to predict the state of the current target according to the target context information of video sequence.Deep learning is gradually widely used in the field of target tracking.This paper elaborates on the de-velopment background of target tracking algorithms and deep learning,and reviews traditional target tracking.Based on different network task functions,target tracking algorithms based deep learning is divided into:deep learning target tracking algorithm based classification,deep learning target tracking algorithm based regression,and target tracking al-gorithm based on the combination of regression and classification.This paper selects representative target tracking algo-rithm for experiments,compares the characteristics of different algorithms.Finally,this paper analyzes the problems of current target tracking algorithms based on deep learning,and prospects the future development direction.The experi-mental results show that deep Siamese tracking networks are superior in accuracy and speed,which is becoming the ma-instream tracking algorithm framework at present.关键词
目标跟踪/深度学习/神经网络/卷积神经网络/孪生神经网络/生成对抗网络Key words
target tracking/deep learning/neural network/convolutional neural network/Siamese network/generative adversarial network分类
军事科技引用本文复制引用
郭凡,卢铉宇,李嘉怡,王红梅..基于深度神经网络的目标跟踪算法综述[J].航空兵器,2024,31(1):1-12,12.基金项目
国家自然科学基金项目(62376224) （62376224）
陕西省重点研发计划(2023-YBGY-232) （2023-YBGY-232）
