融合人脸识别与数据分析的智能签到系统设计

Design of Intelligent Conference Attendance System Based on Face Recognition and Data Analysis

人工智能带动计算机视觉算法迅速发展。人脸识别具有特征明显、不易伪造、安全性高等特点,成为计算机视觉的一个重要分支。基于Hog算法对人脸图像进行特征提取,基于改进的kNN算法进行分类识别。对收集的数据进行科学分析和数据挖掘,将数据信息进行图表可视化。不仅识别迅速、准确度高,还增强了数据分析的趣味性和交互性。应用在会议签到场合,只需采集8张左右的图片,经过预处理和算法优化,就能实现高达99%的准确率,有效解决了传统会议人工签到速度慢、数据可利用率低、组织效率差等弊端。

With the outbreak of artificial intelligence in recent years,the direction of computer vision algorithm has developed rapid ly,and has become the focus of domestic and foreign experts and scholars.Face recognition has the characteristics of obvious features,difficulty to fake,high security and so on.As an important branch of computer vision,it also emerged at the historic moment.This pa per extracts features from face images based on HOG algorithm,and classifies and recognizes face images based on improved kNN algo rithm.Through collecting data,connecting to the database,scientific analysis and data mining,finally the data information is visual ized in charts.It not only recognizes quickly and accurately,but also enhances the interesting and interactive nature of data analysis.The most prominent feature is that only by collecting about eight pictures for each person,after preprocessing and algorithm optimiza tion,the accuracy can reach as high as 99%.In practical applications,data mining can help people make better judgments,so as to take the most appropriate action in the most appropriate time.It effectively solves the drawbacks of slow manual check-in speed,low data availability and poor organizational efficiency of traditional meetings.