基于KCCA的煤矿人员特征融合识别

煤矿井下险情时有发生,为险情下方便救援,需要监控各区域人员签到,准确掌握人员身份及分布情况。在煤矿井下由于光线不足、黑尘干扰等原因,影响人员的识别和管理。传统的识别由于单独依赖人脸识别来辨别井下人员数量和身份,易受到矿井下恶劣的环境影响而导致出现无法识别和识别效率低等问题,所以可靠程度较低。为解决上述问题,提出了一种基于KCCA算法的人脸特征和虹膜特征融合的煤矿井下人员签到识别方法。首先提取出人脸特征和虹膜特征,然后采用KCCA算法对采集到的人脸特征和虹膜特征进行融合,去除图片中无效的信息,降低算法复杂度,最后利用TAN分类完成人员认证,准确识别人员身份。实验表明,该算法降低了计算复杂度,提高了身份识别的准确度,增强了工作人员的安全监控。

Feature Fusion Recognition of Personnel in Coal Mine Based on KCCA

Dangerous situations occur in underground coal mines from time to time.To facilitate rescue in dangerous situations,it is necessary to monitor personnel check-in in various regions and accurately grasp the identity and distribution of personnel.In an underground coal mine,due to weak light and coal dust,the identification and management of personnel can be affected.What’s more,the traditional identification system relies solely on face recognition to identify the number and identity of personnel,which is easily affected by the harsh environment under the mine and leads to problems such as failure to identify and low recognition efficiency,so the reliability is low.To solve these problems,we propose a sign-in identification method for underground coal mine personnel fusing their face and iris feature based on KCCA.At first,the face features and iris features are extracted and then fused by KCCA to remove the useless information in the image and lower the algorithm complexity.In the end,TAN classifier can finish the personnel identification,which is correct and accurate.The experiment shows that the proposed algorithm can cut the complexity of calculation,improve the accuracy of identity recognition and strengthen the coal mine safety surveillance.