自动光谱测试分筛系统研制

Research and development of automatic spectral test and classification system

为实现对工件的自动光谱测试分筛,研制自动光谱测试分筛系统。该文采用光谱曲线重合度算法及颜色判定方法,根据采集光谱曲线参数研究光谱曲线重合度算法,包括重合度函数法、寻峰法和固定点特征值法等,同时研究利用色差值进行颜色判定的方法。实验表明:光谱测量精度达到纳米级,样品测试错误率小于0.5%,该方法高效、稳定可靠、精度高。

To realize automatic spectral test and classification for workpieces,an automatic system for spectral test and classification is researched and developed.The system uses spectral curve matching algorithm and color decision method to do the research.Firstly,mechanical motion and spectral acquisition subsystems were developed.The spectral curve matching algorithm was studied according to the parameters of captured spectral curve,which includes coincidence function,peak search method and the fixed-point characteristics value method.Then,the color difference value was used for studying the color determination method.Finally,through experimental tests,the spectral measurement accuracy reaches to the nanoscale,the test error is in the range of 0.5%.The results indicate this automatic spectral test and classification system is able to meet the requirements of efficient,stable and reliable,high precision.