模式识别技术关于注意力判定的方法研究
|
摘要
随着网络技术的日益成熟,远程教育系统也逐渐成为研究的热点。远程教育系统的目的是实现个性化、因材施教的高效教学方式,是对传统教学模式的一种拓宽和延伸,远程教育系统突破了对求知者的时间、地域、资源共享的限制。然而,目前大多数的研究者多集中关注如何设计个性化的学习资源、提供导航机制,诊断学习者的知识技能、制定有效的学习方案等方面。事实上,他们都忽略考虑这样一个问题,即所有的研究都是基于学习者与远程教育系统的交互以及在系统中反映出来的数据进行的推理和判定。如果学习者在学习过程中有端正的学习态度,且注意力集中,那么这些数据是能充分反映出学习者自身的情况的,系统基于此得出的判断也是真实和可靠的。反之,得到的数据便是不可信的、无用的。所以在远程教育系统设计的过程中必须考虑学习者在学习过程中的注意力是否集中的问题。
本文主要提出了模式识别技术在远程教学系统中一种新的应用模式。通过对国内远程教育系统进行研究分析,提出了目前远程教学系统存在的弊端,进而提出了注意力判别系统在远程教学系统中的重要性。通过对模式识别技术研究现状的分析,提出了注意力分析系统在技术上的可行性。本文利用模式识别技术和数学中的模糊集和概率计算,把感性的注意力用理性的数学计算来表示。通过对学习者的人脸图像进行分析,跟踪来判断学习者的注意力情况,并给出相应的提示。本文给出了该系统的体系结构以及详细的流程,分别介绍了图像处理模块、模式识别模块、注意力分析模块三大模块的工作原理。通过摄像头获取人脸图像,依据图像二值化、高斯平滑、直方图均衡化等手段对获得的初始图像数据进行预处理;利用模式识别技术识别出人眼位置,最后Susan算子对识别出瞳孔在眼睛中的方位,进而分析出注意力。本文最后给出了系统核心算法的实现过程。
实验结果表明,这个研究的方向在应用中具有有效性和实用性。在一定程度上解决了学习者在远程教育中注意力不集中,配合程度不高等问题。
Along with the networking mature, the distance learning system also gradually becomes the research day by day the hot spot. The distance learning system's goal realizes the highly effective teaching way which the personalization, teaches students in accordance with their aptitude, is expands to the tradition educational model one kind and extends, the distance learning system broke through to the seek knowledge time, the region, the resource sharing limit. However, how do the majority researchers concentrate at present pay attention design personalized the study resources, diagnoses aspects and so on learner's knowledge skill, formulation effective study plan. In fact, they neglect considered that this kind of question, namely all research are alternately as well as reflected based on the learner and the distance learning system in the system the data carries on inference and determination. If the learner has the straight study manner in the learning process, and the attention is centralized, then these data are can reflect fully learner own situation, the system the judgment which obtains based on this is also real and reliable. Otherwise, obtains the data is incredible, useless. Therefore in the distance learning system design's process must consider whether learner's to concentrate in learning process attention question.
This article mainly proposed the pattern recognition technology in the multimedia teaching systems one kind of new application pattern. Through to the pattern recognition engineering research present situation's analysis, proposed the attention analysis system feasibility technically. Through carries on the analysis to learner's person face image, the track assesses learner's attention situation, and gives the corresponding prompt. This article has given this system's architecture as well as the detailed flow, introduced the imagery processing module, the pattern recognition module, the attention analysis module three big module principles of work separately. Gains the person face image through the camera, is smooth based on Gauss. Distinguishes the human eye position using the pattern recognition technology, finally the Susan operator to distinguishes pupil's in eye position, then analyzes the attention. This article gave the system core algorithm to realize the process finally.
本文主要提出了模式识别技术在远程教学系统中一种新的应用模式。通过对国内远程教育系统进行研究分析,提出了目前远程教学系统存在的弊端,进而提出了注意力判别系统在远程教学系统中的重要性。通过对模式识别技术研究现状的分析,提出了注意力分析系统在技术上的可行性。本文利用模式识别技术和数学中的模糊集和概率计算,把感性的注意力用理性的数学计算来表示。通过对学习者的人脸图像进行分析,跟踪来判断学习者的注意力情况,并给出相应的提示。本文给出了该系统的体系结构以及详细的流程,分别介绍了图像处理模块、模式识别模块、注意力分析模块三大模块的工作原理。通过摄像头获取人脸图像,依据图像二值化、高斯平滑、直方图均衡化等手段对获得的初始图像数据进行预处理;利用模式识别技术识别出人眼位置,最后Susan算子对识别出瞳孔在眼睛中的方位,进而分析出注意力。本文最后给出了系统核心算法的实现过程。
实验结果表明,这个研究的方向在应用中具有有效性和实用性。在一定程度上解决了学习者在远程教育中注意力不集中,配合程度不高等问题。
Along with the networking mature, the distance learning system also gradually becomes the research day by day the hot spot. The distance learning system's goal realizes the highly effective teaching way which the personalization, teaches students in accordance with their aptitude, is expands to the tradition educational model one kind and extends, the distance learning system broke through to the seek knowledge time, the region, the resource sharing limit. However, how do the majority researchers concentrate at present pay attention design personalized the study resources, diagnoses aspects and so on learner's knowledge skill, formulation effective study plan. In fact, they neglect considered that this kind of question, namely all research are alternately as well as reflected based on the learner and the distance learning system in the system the data carries on inference and determination. If the learner has the straight study manner in the learning process, and the attention is centralized, then these data are can reflect fully learner own situation, the system the judgment which obtains based on this is also real and reliable. Otherwise, obtains the data is incredible, useless. Therefore in the distance learning system design's process must consider whether learner's to concentrate in learning process attention question.
This article mainly proposed the pattern recognition technology in the multimedia teaching systems one kind of new application pattern. Through to the pattern recognition engineering research present situation's analysis, proposed the attention analysis system feasibility technically. Through carries on the analysis to learner's person face image, the track assesses learner's attention situation, and gives the corresponding prompt. This article has given this system's architecture as well as the detailed flow, introduced the imagery processing module, the pattern recognition module, the attention analysis module three big module principles of work separately. Gains the person face image through the camera, is smooth based on Gauss. Distinguishes the human eye position using the pattern recognition technology, finally the Susan operator to distinguishes pupil's in eye position, then analyzes the attention. This article gave the system core algorithm to realize the process finally.
引文
[1]刘航.远程教学优势及其对教育改革的意义.广东教育学院学报.1999,05.
[2]李克东.数字化学习—信息技术与课程整合的核心.电化教育研究.2001,8.
[3]邹志煌,孙鑫,程武山.人脸识别技术产品的发展概况. video application & project. 01(23):91-93
[4]詹曙著.人脸图象识别技术的研究.中国科学技术大学博士论文.
[5]钟玉琢,王琪,贺玉文.基于对象的远程数据压缩编码国际标准.科学出版社. 2000.
[6] http://wwwam.hhi.de/mpeg-video/#MPEG4,2002-09-01.
[7]张洪明.人脸检测技术的研究与实现.哈尔滨工业大学工学硕士学位论文.图书分类号:TP391.
[8]张晶,李志敏,黄凡.一种改进的自适应模板匹配法.微计算机信息.2008,27.
[9] P S F van Keulen.The Meaning of the Phrase WNùSPTùL-QBRTYK B?LWM in 2 Kings XXII 20.1568-5330 (Online):256-260
[10]王玲,龚健雅.一种基于自适应谐振理论的TM遥感影像神经网络分类器.测绘科学.2003, 28(3):39-42
[11]李云峰,杨益,田俊香.人脸识别的研究进展与发展方向.科技资讯.2008,O5:4-6.
[12]http://zh.wikipedia.org/wiki/%E9%9A%90%E9%A9%AC%E5%B0%94%E5%8F%AF%E5%A4%AB%E6%A8%A1%E5%9E%8B,2010-02-25.
[13]http://www.eefocus.com/article/07-12/26224251212269636.html,2007-12-25.
[14]S.Birchfield . Elliptical Head tracking using intensity gradients and color histograms.Computer Vision and Pattern Recognition,1998:232-237.
[15]王路群.远程与远程教学.武汉冶金管理干部学院学报.2000,04.
[16]胡为芹,杨晓宏.论现代远程教学在成人教育发展中的优势.西北成人教育学报,2003, 02.
[17]吴蓉瑾.引导学生注意力在一张一弛中“保鲜”.文汇报.2010-01-26,007
[18]模糊逻辑简介.[EB/OL].http://210.40.32.30/gzzy/ArticleShow.asp?ArticleID=1322
[19]熊超.模式识别理论及其应用综述.中国科技信息.2006,06.
[20]V. I. Donskoi.Partially defined optimization problems: An approach to a solution that is based on pattern recognition theory.Journal of Soviet Mathematics.1993: 1664-1668.
[21]施光燕,董加礼.最优化方法.高等教育出版社,1999-09.
[22]王晓东.计算机算法设计与分析.电子工业出版社,2001-01.
[23]陈松灿,伍艳莲.图像的模糊识别方法研究与实现.电子学报.2002,280(11):50-54.
[24]J Dowdall, I Pavlidis, G Bebis. Face Detection in the Near-IR Spectrum Image and Vision Computing [Z]. 2003, 21: 565-578.
[25]H M El-Bakry, H Stoyan. Fast neural networks for sub-matrix (object/face) detection,Circuits and Systems [A]. ISCAS '04. Proceedings of the 2004 International Symposium [C]. 2004, 5: 764-767.
[26]Xiaodan Zhang,Hai Zhao, Shouzhi Wei, Weiji SU. Application of Fuzzy Pattern Recognition in Hydroelectricity Simulation Examining System.Computer Engineering,2005-03.
[27]吴顺祥,吉国力.数据库系统与知识系统的对比分析.计算机工程与应用.1999-09: 83-85.
[28]程录庆.关系数据理论教学探讨.南京人口管理干部学院学报.2003-01,19(1):54-56
[29]宗华丽.模式识别在玉米叶部病害诊断中的应用. http://www.paper.edu.cn
[30]张中如.网络远程教学对传统职业教育的影响.现代商贸工业.2007,01.
[31]王路群.远程与远程教学.武汉冶金管理干部学院学报.2000,04.
[32]严超,苏光大.人脸特征的定位与提取.中国图像图形学报.1998,3(5):375-380.
[33]李刚,宋文静.基于图像直方图的车牌图像二值化方法研究.交通运输系统工程与信息. 2009,01.
[34]张翠平,苏光大.人脸识别技术综述.中国图像图形学报(A版).2000,5 (11):885-894.
[35]顾华,苏光大,杜成.人脸识别中的眼角定位和预处理.2003年中国智能自动化会议(CIAC 2003):429-433.
[36]Tianyang Zheng,Wei pan,Bing Jia.Study on the Application of Pattern Recognition Technology in Distance Education System.ICCEE2009:427-429
[2]李克东.数字化学习—信息技术与课程整合的核心.电化教育研究.2001,8.
[3]邹志煌,孙鑫,程武山.人脸识别技术产品的发展概况. video application & project. 01(23):91-93
[4]詹曙著.人脸图象识别技术的研究.中国科学技术大学博士论文.
[5]钟玉琢,王琪,贺玉文.基于对象的远程数据压缩编码国际标准.科学出版社. 2000.
[6] http://wwwam.hhi.de/mpeg-video/#MPEG4,2002-09-01.
[7]张洪明.人脸检测技术的研究与实现.哈尔滨工业大学工学硕士学位论文.图书分类号:TP391.
[8]张晶,李志敏,黄凡.一种改进的自适应模板匹配法.微计算机信息.2008,27.
[9] P S F van Keulen.The Meaning of the Phrase WNùSPTùL-QBRTYK B?LWM in 2 Kings XXII 20.1568-5330 (Online):256-260
[10]王玲,龚健雅.一种基于自适应谐振理论的TM遥感影像神经网络分类器.测绘科学.2003, 28(3):39-42
[11]李云峰,杨益,田俊香.人脸识别的研究进展与发展方向.科技资讯.2008,O5:4-6.
[12]http://zh.wikipedia.org/wiki/%E9%9A%90%E9%A9%AC%E5%B0%94%E5%8F%AF%E5%A4%AB%E6%A8%A1%E5%9E%8B,2010-02-25.
[13]http://www.eefocus.com/article/07-12/26224251212269636.html,2007-12-25.
[14]S.Birchfield . Elliptical Head tracking using intensity gradients and color histograms.Computer Vision and Pattern Recognition,1998:232-237.
[15]王路群.远程与远程教学.武汉冶金管理干部学院学报.2000,04.
[16]胡为芹,杨晓宏.论现代远程教学在成人教育发展中的优势.西北成人教育学报,2003, 02.
[17]吴蓉瑾.引导学生注意力在一张一弛中“保鲜”.文汇报.2010-01-26,007
[18]模糊逻辑简介.[EB/OL].http://210.40.32.30/gzzy/ArticleShow.asp?ArticleID=1322
[19]熊超.模式识别理论及其应用综述.中国科技信息.2006,06.
[20]V. I. Donskoi.Partially defined optimization problems: An approach to a solution that is based on pattern recognition theory.Journal of Soviet Mathematics.1993: 1664-1668.
[21]施光燕,董加礼.最优化方法.高等教育出版社,1999-09.
[22]王晓东.计算机算法设计与分析.电子工业出版社,2001-01.
[23]陈松灿,伍艳莲.图像的模糊识别方法研究与实现.电子学报.2002,280(11):50-54.
[24]J Dowdall, I Pavlidis, G Bebis. Face Detection in the Near-IR Spectrum Image and Vision Computing [Z]. 2003, 21: 565-578.
[25]H M El-Bakry, H Stoyan. Fast neural networks for sub-matrix (object/face) detection,Circuits and Systems [A]. ISCAS '04. Proceedings of the 2004 International Symposium [C]. 2004, 5: 764-767.
[26]Xiaodan Zhang,Hai Zhao, Shouzhi Wei, Weiji SU. Application of Fuzzy Pattern Recognition in Hydroelectricity Simulation Examining System.Computer Engineering,2005-03.
[27]吴顺祥,吉国力.数据库系统与知识系统的对比分析.计算机工程与应用.1999-09: 83-85.
[28]程录庆.关系数据理论教学探讨.南京人口管理干部学院学报.2003-01,19(1):54-56
[29]宗华丽.模式识别在玉米叶部病害诊断中的应用. http://www.paper.edu.cn
[30]张中如.网络远程教学对传统职业教育的影响.现代商贸工业.2007,01.
[31]王路群.远程与远程教学.武汉冶金管理干部学院学报.2000,04.
[32]严超,苏光大.人脸特征的定位与提取.中国图像图形学报.1998,3(5):375-380.
[33]李刚,宋文静.基于图像直方图的车牌图像二值化方法研究.交通运输系统工程与信息. 2009,01.
[34]张翠平,苏光大.人脸识别技术综述.中国图像图形学报(A版).2000,5 (11):885-894.
[35]顾华,苏光大,杜成.人脸识别中的眼角定位和预处理.2003年中国智能自动化会议(CIAC 2003):429-433.
[36]Tianyang Zheng,Wei pan,Bing Jia.Study on the Application of Pattern Recognition Technology in Distance Education System.ICCEE2009:427-429