提出了一种新型的基于纳米颗粒上转换发光的防伪编码方法。该方法将MATLAB图像处理算法、支持向量机分类和径向基函数神经网络分类相结合, 通过分析纳米材料在980 nm抽运激光激发下产生的绿光和红光的上转换发光光谱信息, 及该纳米材料在防伪基底上的空间分布特征, 实现了一种具有高防伪度及建库与识别自动化的防伪方式。实验验证了所制作出的防伪标识难复制且其特征图样具有高稳定性, 鉴定准确率可达95.4%。

A novel anti-counterfeiting encoding method based on the nanoparticle upconversion luminescence is proposed. This method combines the image processing algorithms based on MATLAB, support vector machine (SVM) classification and radial basis function neural network classification. With the analysis of the upconversion luminescence spectral characteristics of the green and red lights emitted from the nano-materials which are pumped by the 980 nm laser, and the spatial distribution characteristics of these nano-materials on the anti-counterfeiting substrate, the anti-counterfeiting method with a high-security and an automatic capability of database building and pattern recognizing is realized. The experiment confirms that it is difficult to replicate the prepared anti-counterfeiting markings whose characteristic patterns have a high stability, and the identification accuracy is up to 95.4%.