基于人脸识别的矿井人员管理技术研究

发布时间：2018-11-05 12:58

【摘要】：我国是煤炭生产大国,煤矿人员管理在煤矿安全生产中占有十分重要的地位。因此,引入生物识别的方法,对矿井人员的身份进行精确识别以及矿井人员安全管理系统的研究有重要意义。本文以人脸识别技术在矿井人员管理系统的应用为研究背景,分析了现有矿井人员管理系统、人员身份精确管理技术、人脸识别技术以及人脸识别技术在矿井人员管理中应用的可行性。介绍了常用的人脸检测和识别算法,通过对矿井人脸图像的特点分析,初步选择人脸检测和识别算法。针对图像传输中存在的噪声,采用中值和同态滤波的图像去噪方法,对图像进行预处理,消除了图像传输中的孤立噪声,加强了图像的对比度。采用Canny边缘检测的办法对矿帽及人脸轮廓特征提取,根据轮廓特征进行人脸检测。实验证明,该人脸检测的算法检测人脸准确,适合矿井人员的人脸检测。介绍了主成分分析法的原理及实现步骤,用主成分分析法对人脸图像进行特征提取,降低了图像处理的复杂度,便于进行人脸识别。建立深度信念网络的人脸识别模型,并且结合主成分分析法和深度信念网络算法建立改进的人脸识别模型。并在ORL,FLW人脸数据库中进行仿真实验。对比分析表明,主成分分析深度信念网络人脸识别模型具有较高的识别速度和识别率。并设计矿井人员人脸识别的人机界面,实现了人脸图像采集、人脸检测、人脸识别和人员信息显示功能。经过测试,本文主成分分析深度信念网络人脸识别算法对于测试的人脸图像检测和识别准确,对矿井人员考勤管理技术作了补充,在实际应用有重要的参考价值。对于矿井人员安全管理系统的进一步的完善具有一定的指导意义。
[Abstract]:Our country is a big country of coal production, the management of coal mine personnel occupies a very important position in coal mine safety production. Therefore, it is of great significance to introduce the biometric method to identify the identity of mine personnel and to study the safety management system of mine personnel. Based on the application of face recognition technology in mine personnel management system, this paper analyzes the existing mine personnel management system, personnel identity accurate management technology, The feasibility of the application of face recognition technology and face recognition technology in mine personnel management. This paper introduces the commonly used face detection and recognition algorithms, and selects the face detection and recognition algorithm by analyzing the characteristics of mine face image. Aiming at the noise existing in image transmission, the median and homomorphic filtering method is used to pre-process the image, eliminate the isolated noise in image transmission, and enhance the contrast of image. The method of Canny edge detection is used to extract the contour features of mining cap and face, and face detection is carried out according to the contour features. Experiments show that the face detection algorithm is accurate and suitable for face detection of mine personnel. The principle and implementation of principal component analysis (PCA) are introduced in this paper. The feature extraction of face image is carried out by principal component analysis (PCA), which reduces the complexity of image processing and facilitates face recognition. The face recognition model of depth belief network is established, and the improved face recognition model is established by combining principal component analysis and depth belief network algorithm. The simulation experiment is carried out in ORL,FLW face database. The comparative analysis shows that the face recognition model based on PCA depth belief network has high recognition speed and recognition rate. The man-machine interface of mine personnel face recognition is designed, and the functions of face image collection, face detection, face recognition and human information display are realized. After testing, the principal component analysis (PCA) depth belief network face recognition algorithm is accurate for the tested face image detection and recognition, and complements the attendance management technology of mine personnel. It has important reference value in practical application. It has certain guiding significance for the further improvement of mine personnel safety management system.
【学位授予单位】：西安科技大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：TD76;TP391.41

【参考文献】