基于视觉测量的脉搏图像三维重构  

3D Reconstruction of Pulse Image Based on Visual Measurement

作　　者：张爱华[1,2] 李金平[1,2] 林冬梅[1,2] 

机构地区：[1]兰州理工大学电气工程与信息工程学院,兰州730050 [2]甘肃省工业过程先进控制重点实验室,兰州730050

出　　处：《数据采集与处理》2012年第5期570-575,共6页Journal of Data Acquisition and Processing

基　　金：国家自然科学基金(30670529)资助项目;甘肃省自然科学研究基金计划(1014RJZA013)资助项目

摘　　要：动脉搏动情况是人体循环系统动态过程中一个可检测的生物讯号。在单目视觉脉搏图像传感器的基础上,依据脉搏图像传感器中薄膜及其网格变化约束的特点,建立了空间测量原理的数学模型,运用Harris角点检测算法,对连续各帧脉搏图像中的角点进行跟踪匹配,从而重构出脉搏搏动的三维信息。通过实验获得了平脉、滑脉、迟脉和濡脉4种脉象的三维信息,实现了脉搏图像的三维重构,为进一步在时空域中提取多种表征脉搏状态的特征奠定了基础。The pulse situation is a detectable biological signal during the dynamic process of human body circulation.To obtain more pulse information,the mathematical model of space measurement principle is established according to the characteristics of the thin film grid in a pulse image sensor,and it is built on the self-developed pulse image sensor with monocular vision.The corners of pulse image in each successive frame are tracked and matched by the Harris corner detection algorithm,and the three-dimensional information of the pulse situation is reconstructed.Then,the pulse images of the normal pulse,the slippery pulse,the slow pulse,and the soft pulse are obtained and analyzed by experiment.Therefore,the three-dimensional reconstruction of pulse images is realized.It is helpful for the further research in feature extration for characterizing the pulse states in space-time domain.

关 键 词：图像传感器 空间位移测量 匹配与跟踪 三维重构 脉搏 

分 类 号：R318[医药卫生—生物医学工程]