基于低分辨率的人脸识别方法研究

Survey on face recognition method based on low-resolution

摘要:

随着我国经济实力的提高，大多公共场所都配备有监控设备[1]，用以对出现行人及行为的监控，通过监控识别人脸是其发展的重要方面以后也将进一步发展。为减少视频数据的空间占比[2]，目前市场大部分采用低分辨率的存储方式来存放视频数据，这对人脸的识别率带来了极大的影响。人脸识别是完善和加深图像识别在视频监控领域应用的基础，如何能够准确、快速、低成本的完成人脸识别工作，无论在生活领域、信息安全领域和公共安全领域等都有着无限的应用空间。因此，本文在前人所做的基础之上，结合近些年该领域的发展情况，进行了比较有针对性的描述。对一些比较热点的方法进行了详细的理论介绍，希望相关研究工作者能够基于本文所作的相关研究，发挥其更大的价值。

Abstract:

With the development of China’s economic strength, more and more monitoring equipments are used to monitor predestrians in many public places. The Facial recognition is and will be further important aspect in this field. In order to reduce the spatial proportion of video data, most of the current market uses low-resolution storage to store video data, which has a great impact on the recognition rate of human faces. the method can improve the face recognition effect under the environment of low resolution.complete the number of accurate, fast, low cost of statistical work, no matter in the field of business and security areas, will bring infinite space.Therefore, based on the previous work, this paper makes a more targeted description based on the development of the field in recent years.In this paper, some hot methods are introduced in detail, and it is hoped that the relevant researchers can give play to their greater value on this basis.