电力企业生产环境智能门禁管理系统的设计  

Design of intelligent access control system for production environment of electric power enterprises

作　　者：甘绪桐 GAN Xutong(National Energy Changyuan Jingmen Power Generation Co.,Ltd.,Jingmen 448000,China)

机构地区：[1]国能长源荆门发电有限公司,湖北荆门448000

出　　处：《中国高新科技》2025年第1期118-120,共3页

摘　　要：随着信息技术的飞速发展,电力企业对安全性和工作效率的要求日益提高。然而,传统的面部识别系统在光线不足、遮挡及高面部相似度等复杂环境中表现不佳。文章研究了基于遗传算法改进的ROI-KNN卷积神经网络,并在电力企业生产环境中进行测试。实验结果表明,改进后的系统在低光环境下识别准确率提高至98%,在遮挡条件下准确率为75.16%,在高面部相似度条件下准确率为76.8%。改进系统显著提升了智能门禁管理系统的识别性能和适应能力。With the rapid development of information technology,electric power enterprises are increasingly demanding safety and efficiency.However,traditional facial recognition systems do not perform well in complex environments such as poor lighting,occlusion,and high facial similarity.Based on this,the ROI-KNN convolutional neural network improved by genetic algorithm is studied in this paper,and tested in the production environment of electric power enterprises.The experimental results show that the recognition accuracy of the improved system is increased to 98%under low light,75.16%under occlusion and 76.8%under high facial similarity.The improved system significantly improves the identification performance and adaptability of the intelligent access control management system.

关 键 词：面部识别 遗传算法 ROI-KNN卷积神经网络 智能门禁 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]