CAJ | 학술논문

针对如何提高瞳孔定位的效率和精度以改善眼睛检测类仪器结果的可靠性，本文介绍了基于Arm11硬件平台在WinCE环境下采用Directshow技术架构实现摄像头图像数据采集及实时视频图像处理分析的实践过程，并在此基础上运用先粗略图像定位再精确定位的方法实现实时人眼瞳孔精确定位。实验中采用的是基于图像灰度信息的瞳孔定位算法，在WinCE平台下成功精确地定位出瞳孔中心，该算法简单高效适用于嵌入式平台环境下，在WinCE环境下运用此算法成功实现了15fps的实时瞳孔定位。
침대여하제고동공정위적효솔화정도이개선안정검측류의기결과적가고성，본문개소료기우Arm11경건평태재WinCE배경하채용Directshow기술가구실현섭상두도상수거채집급실시시빈도상처리분석적실천과정，병재차기출상운용선조략도상정위재정학정위적방법실현실시인안동공정학정위。실험중채용적시기우도상회도신식적동공정위산법，재WinCE평태하성공정학지정위출동공중심，해산법간단고효괄용우감입식평태배경하，재WinCE배경하운용차산법성공실현료15fps적실시동공정위。
For how to improve the efficiency and accuracy of positioning pupil to improve the reliability of the results of eyes de-tection instruments.This paper presents a practice process that how to realize image data acquisition and real-time video image pro-cessing and analysis using Directshow technical architecture based on Arm11 hardware platform under the WinCE environment.And on this basis using coarse positioning method firstly then precise positioning the pupil center.In the experiment use the pupil localiza-tion algorithm which based on the image gray value information, successfully and accurately locate the pupil center under the WinCE platform.The algorithm is simple and efficient which is suitable for embedded platform environment.Under the WinCE platform to use this algorithm successfully achieved 15fps real-time pupil positioning.