人体屈体角度图像识别关键技术研究  

Research on the key Technology of RecognizingImage AboutHuman pike Angle

作　　者：卢旭[1] 张钰婧 卓胜达 LU Xu;ZHANG Yu-jing;ZHUO Sheng-da(School of Automation,Guangdong Polytechnical Normal University;School of Computer Science,Guangdong Polytechnical Normal University,Guangzhou 510665,China)

机构地区：[1]广东技术师范学院自动化学院 [2]广东技术师范学院计算机科学学院,广东广州510665

出　　处：《软件导刊》2018年第6期210-212,共3页Software Guide

基　　金：广东省科技计划项目(2017A040405059);广州市科技计划项目产学研协同创新重大专项项目(201704020113)

摘　　要：随着社会的发展,人体图像识别的应用也越来越广泛,并出现了很多关于人体图像识别的新方法。研究人体屈体角度识别的实现方法,结合人体图像处理的一般流程,对人体屈体角度图像识别算法的预处理、人体图像分割、特征提取和特征判断匹配等步骤进行阐述,介绍了其中的关键技术——人体图像分割和特征点提取,并对现有算法的优缺点进行分析,以期为相关研究提供参考。As the progress of society,human body image recognition has increasingly wider application.Many new ideas about the implementation of image recognition have been emerging.Based on this background,the method of realizing human body pike angle is studied in this paper.According to the general process of human image processing,several steps of body image recognition algorithm are expounded in this paper,including pretreatment,body image segmentation,feature extraction and feature matching of judgement.And the key technologies of human body image segmentation and feature point extraction are studied.The advantages and disadvantages of these algorithms are analyzed,which can provide important reference for related research.

关 键 词：人体图像识别 屈体角度 人体图像分割 特征点提取 

分 类 号：TP317.4[自动化与计算机技术—计算机软件与理论]