标题: | 基于影像处理之人体姿态辨识 Image-Based Human Pose Recognition |
作者: | 赖裕宏 Lai, Yu-Hung 宋开泰 陈福川 Song, Kai-Tai Chen, Fu-Chuang 电机学院电机产业专班 |
关键字: | 姿态辨识;感测资料融合;类神经网路;影像辨识;Body pose recognition;sensor data fusion;neural network;image recognition |
公开日期: | 2009 |
摘要: | 本论文之主要目的在于藉由摄影机撷取到的影像资讯,得以在环境中寻找人员的存在,并且完成六种人体姿态的辨识。本系统使用肤色与发色资讯,以及连通标记法(Connected component labeling)完成人头的侦测,再用椭圆模型与人体模型来辨识影像中存在的人体。利用所找到的人体资讯与相关特征,本论文完成一套可用来判断人体姿态的影像处理系统。此外,本论文使用类神经网路融合影像辨识资讯与实验室之基于三轴加速规之人体姿态估测资讯,实验结果发现,融合前的影像平均辨识率为79.23%,人体姿态估测模组为88%,融合后之平均辨识率可达93.5%。 Real-time body pose information is very useful for many human-robot interaction applications. However, due to the motion of both human and the robot, robust body pose recognition poses a challenge in such a system design. This thesis aims to locate a human body in the image plane and then recognize six body poses through image recognition. The color-space techniques and the method of connected component are used to detect a human. Ellipse models and body shape patterns are used to locate human body in the video stream. Furthermore, a neutral network has been designed to fuse data from image recognition and inertial sensors to improve the recognition rate under various environmental variations. Experimental results show that the average recognition rate of six body poses is 93.5%, an improvement from 79.23% and 90.67% of using only image recognition and inertial sensor respectively. |
URI: | http://140.113.39.130/cdrfb3/record/nctu/#GT009493507 http://hdl.handle.net/11536/37955 |
显示于类别: | Thesis |
文件中的档案:
If it is a zip file, please download the file and unzip it, then open index.html in a browser to view the full text content.