人脸图像远距离识别技术研究

发布时间：2017-12-31 13:03

本文关键词：人脸图像远距离识别技术研究　出处：《湖北工业大学》2017年硕士论文　论文类型：学位论文

【摘要】：人脸识别技术自从上世纪60年代被提出以来,一直是研究热点,也是人工智能领域的重点,尤其是近年来随着安保图像联网、智能监控和安保图像云端计算等技术的快速推进,安保系统的人脸识别技术在嫌疑犯身份确认、在逃人员通缉、重点场所布控、特殊人群身份确认、户籍管理等方面正逐步向实用过渡。当然,随着监控设备在城市的安装,在实际应用中很多时候需要在一定距离下对目标物的身份确认,所以远距离的人脸图像识别是一项具有重大理论意义和实际应用价值的课题,如何利用计算机克服自然场景中的人脸的姿态变化和光线影响,进而精确的检测识别出人脸图像,成为了目前图像领域的热门问题。本文介绍了目前人脸远距离识别的研究背景,剖析了人脸远距离识别的国内外研究现状,研究了人脸识别的工作原理和技术,分析了系统的需求分析,提出了远距离的人脸图像重构以及遮挡下的人脸检测和识别的课题和可行性方法。首先介绍了人脸识别的意义和技术发展过程,以及概括说明了人脸识别的通用架,总结了人脸识别的难点问题。例如:光照的变化、人脸姿态的变化、表情的变化、有遮挡物、天气原因等外界因素的干扰等问题,这些因素都会影响到人脸识别的准确度,因此,本文旨在研究远距离人脸识别在实际应用识别中的应用。为的就是能够随着技术进步可以快速的、有效地发展远距离识别技术。最后测试了原型系统的性能。实验证明,该原型系统在人脸识别率和识别速度方面均表现良好,达到了系统预期的设计目标。
[Abstract]:Since it was proposed in -40s, face recognition technology has been a hot topic in the field of artificial intelligence, especially with the security image networking in recent years. Intelligent surveillance and security image cloud computing technology advances rapidly, security system face recognition technology in the suspect identification, wanted by fugitives, key location control, special crowd identification. Household registration management and other aspects of the gradual transition to practical. Of course, with the installation of monitoring equipment in the city, many times in practical applications need to identify the object at a certain distance. So long-distance face image recognition is a topic of great theoretical significance and practical application value. How to use computer to overcome the change of human face attitude and the influence of light in natural scene. Then accurate detection and recognition of face image has become a hot issue in the field of image. This paper introduces the research background of face remote recognition and analyzes the research status of face remote recognition at home and abroad. The working principle and technology of face recognition are studied, and the requirement analysis of the system is analyzed. The task and feasibility method of face detection and recognition under occlusion are put forward. Firstly, the significance and technical development of face recognition are introduced. And generalize the general frame of face recognition, summarize the difficult problems of face recognition, such as: the change of illumination, the change of face pose, the change of expression, and the occlusion. Weather factors and other external factors such as interference, these factors will affect the accuracy of face recognition, so. The purpose of this paper is to study the application of long-distance face recognition in practical applications. Finally, the performance of the prototype system is tested. Experiments show that the prototype system performs well in face recognition rate and recognition speed, and achieves the desired design goal.
【学位授予单位】：湖北工业大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：TP391.41

【参考文献】