基于光照感知权重融合的多模态行人检测算法  

Multi-Modal Pedestrian Detection Algorithm Based on Illumination Perception Weight Fusion

作　　者：刘珂琪 董绵绵[1] 郜辉[1] 吕志刚[1] 郭宝亿[2] 庞敏 Liu Keqi;Dong Mianmian;Gao Hui;Zhigang Lü;Guo Baoyi;Pang Min(School of Electronic and Information Engineering,Xi’an Technological University,Xi’an 710021,Shaanxi,China;Undergraduate College,Xi’an Technological University,Xi’an 710021,Shaanxi,China;Beijing Institute of Microelectronics Technology,Beijing 100000,China)

机构地区：[1]西安工业大学电子信息工程学院,陕西西安710021 [2]西安工业大学本科生院,陕西西安710021 [3]北京微电子技术研究所,北京100000

出　　处：《激光与光电子学进展》2023年第16期137-147,共11页Laser & Optoelectronics Progress

基　　金：国家自然科学基金(62171360);陕西省科技厅重点研发计划(2022GY-110);西安工业大学校长基金面上培育项目(XGPY200217);西安市智能兵器重点实验室(2019220514SYS020CG042);2022年度陕西高校青年创新团队项目。

摘　　要：针对现有利用可见光与红外模态融合的行人目标检测算法在全天候环境下漏检率高的问题,提出一种基于光照感知权重融合的多模态行人目标检测算法。首先,使用引入高效通道注意力(ECA)机制模块的ResNet50作为特征提取网络,分别提取两个模态的特征;其次,对现有光照加权感知融合策略进行改进,通过设计一种新的光照感知加权融合机制获取可见光与红外模态的对应权重,并进行加权融合得到融合特征,从而降低算法的检测漏检率;最后,将从特征网络最后一层提取的多模态特征和生成的融合特征共同送入到检测网络,完成行人目标检测。实验结果表明,所提算法在KAIST数据集下具有良好的检测性能,在全天候下对行人目标的检测漏检率为11.16%。Existing pedestrian target detection algorithm based on visible light and infrared modal fusion has a high missed detection rate in all-weather environment.In this paper,we propose a novel multi-modal pedestrian target detection algorithm based on illumination perception weight fusion to solve this problem.First,ResNet50,incorporating an efficient channel attention(ECA)mechanism module,was used as a feature extraction network to extract the features of both visible light and infrared modes,respectively.Second,the existing illumination weighted sensing fusion strategy was improved.A new illumination weighted sensing fusion mechanism was designed to attain the corresponding weights of the visible light and infrared modes,and weighted fusion was performed to achieve fusion features to reduce the missed detection rate of the algorithm.Finally,the multi-modal features extracted from the last layer of the feature network and the generated fusion features were fed into the detection network to accomplish the detection of pedestrian targets.Experimental results show that the proposed algorithm has an excellent detection performance on the KAIST dataset,and the missed detection rate for pedestrian targets in all-weather is 11.16%.

关 键 词：多模态图像融合 注意力机制 光照感知权重融合 行人检测 

分 类 号：TP391.41[自动化与计算机技术—计算机应用技术]