摘要
随着以MPEG-4,H.264为代表的高性能视频压缩格式和多片机控制系统的应用,现有的图像特征数据检测、识别能力却明显不足以应对关照极差、干扰因素极复杂的动态作业环境的需求。针对此提出将特征点定位支持向量机分类这两项人脸识别关键技术算法改进,并基于此改进重新设计人脸识别系统的结构体系和功能模块,以应对复杂的识别环境。
With the application of high - performance video compression format MPEG - 4, H. 264, and multi -chip control system, the image feature data detection, the ability to identify, but obviously not e-nough to respond to take care of poor, confounding factors is very complexdynamic operating environment requirements. Positioning the feature points in these two key technologies of face recognition technology algorithm improvements, and improve the re -design the structure of the face recognition system based on this system and functional modules support vector machine classifier to cope with the complex environment of recognition for this.
出处
《煤炭技术》
CAS
北大核心
2012年第8期1-3,共3页
Coal Technology
基金
国家自然科学基金项目(70962013)
关键词
特征点定位
决策树
支持向量机
Image feature point positioning
decision tree
support vector machine
