摘要
对基于深度学习的红外图像步态识别方法进行研究,利用卷积神经网络相关技术搭建一个深度学习模型,以此对红外图像中人体步态轮廓特征进行学习,对红外图像中人体步态身份做出识别。使用图像形态学中的闭运算对图片进行数据预处理,实验结果表明,经过预处理后的红外图像能够有效减少因红外相机成像原理导致的图像中产生冗余信息,有效提高模型的泛化能力。经过对实验数据的对比与分析,该模型及数据预处理的方法对红外图像中的人体步态识别有着比较显著的效果。
The gait recognition method based on deep learning in infrared image was studied,and a deep learning model was established using convolutional neural network to learn the contour features of human gait in infrared image and the identity of human gait was recognized in infrared image.The closed operation in image morphology was used to preprocess the image data.From the experimental results,it can be concluded that the infrared image after preprocessing can effectively reduce the redundant information in the image caused by the imaging principle of infrared camera,thus effectively improving the generalization ability of the model.Through the comparison and analysis of the experimental data,the model and the method of data preprocessing have significant effects on the recognition of human gait in the infrared image.
作者
朱小鹏
云利军
张春节
王坤
ZHU Xiao-peng;YUN Li-jun;ZHANG Chun-jie;WANG Kun(School of Information Science and Technology,Yunnan Normal University,Kunming 650500,China;Yunnan Key Laboratory of Opto-Electronic Information Technology,Yunnan Normal University,Kunming 650500,China)
出处
《计算机工程与设计》
北大核心
2022年第3期851-857,共7页
Computer Engineering and Design
基金
云南省应用基础研究计划重点基金项目(2018FA033)
云南师范大学研究生科研创新基金项目(ysdyjs2020156)。
关键词
卷积神经网络
红外图像
人体步态识别
数据预处理
冗余信息
convolutional neural network
infrared image
human gait recognition
data preprocessing
redundant information
