摘要
针对体全息图像库模式识别存在一些难题,为了提高全息图像库的模式识别准确性,提出一种新型的体全息图像库模式识别方法。首先对当前体全息图像库模式识别现状进行分析,指出了当前方法存在的不足,然后提取体全息图像的特征向量,并采用相关向量机对体全息图像库中的图像进行模式识别,最后采用具体实验对本文方法的可行性进行分析。结果表明,本文方法可以对体全息图像进行准确分类,提高了体全息图像模式识别的正确率,解决传统体全息图像识别方法准确率低,识别速度慢的缺陷。
In view of the problem of pattern recognition of volume holographic image library, a new pattern recog- nition method of volume holographic image library is proposed in order to improve the accuracy of pattern recognition. Firstly, the current status of pattern recognition of the current volume holographic image database is analyzed, and the shortcomings of the current method are pointed out, then the feature vector of the volume holographic image is extracted, and the pattern recognition of the image in the volume holographic image database is carried out by using the correlation vector machine. Finally, the feasibility of this method is analyzed by the experiment. Results show that, this method can be of volume holographic image of classification accuracy, to improve the accuracy of volume holographic image pattern recognition and to solve the low accuracy in traditional volume holographic image identification method and slow speed of recognition of defects.
出处
《激光杂志》
北大核心
2016年第3期68-70,共3页
Laser Journal
基金
湖北省科技厅项目(201400823)
关键词
体全息
图像数据库
模式识别
相似性度量
volume holography
image database
pattern recognition
similarity measure
