中阶梯光栅光谱仪谱图信息处理方法研究

发布时间：2018-01-03 02:18

本文关键词：中阶梯光栅光谱仪谱图信息处理方法研究　出处：《中国科学院长春光学精密机械与物理研究所》2017年博士论文　论文类型：学位论文

【摘要】：中阶梯光栅光谱仪是一种利用交叉色散配合面阵探测器同时实现高光谱分辨率和全谱直读的设备。相比于同体积的其他光谱仪器,中阶梯光栅光谱仪不仅光谱分辨率极高,可以实现全谱瞬态直读,而且具有宽波段、高能量传输效率、无扫描部件等诸多优点。随着光栅制作技术和高水平探测器研制技术的发展,中阶梯光栅光谱仪已经从天文观测转向民用,化学分析、质量检测、地质勘探、环境监测等领域都有广泛应用。目前,国外多家公司已经推出了一系列的中阶梯光栅光谱仪商品,而国内仍处于原理样机的研发阶段。制约我国中阶梯光栅光谱仪向产品化发展的瓶颈是其谱图信息处理的精度和速度,精度影响仪器指标、速度决定仪器性能。基于此,本文对中阶梯光栅光谱仪谱图信息处理方法进行了深入研究。具体研究工作如下:第一,提出了基于几何光学追迹的谱图还原模型建立方法。在分析中阶梯光栅光谱仪结构特点的基础上,建立了表征波长与成像光斑坐标之间对应关系的谱图还原模型。通过棱镜谱线弯曲的校正实现了整个像面内模型误差小于一个像元,极大提高了谱图还原的精度。第二,提出了一种适用于中阶梯光栅光谱仪原始图像特点的背景去除算法。通过分析原始二维图像光斑灰度值的特征,采用形态学平滑加边缘提取的方法锁定有效光斑并将其余位置的灰度值置为零。该方法减小了谱图数据处理的数据量,不仅优化了信噪比而且提高了谱图信息处理的速度。第三,根据不同波长特性入射光源对应的不同成像特点,提出了不同的波长提取算法。针对分立谱光源,采用基于光斑质心计算的波长提取算法,针对连续谱和分立谱光源,采用基于谱图还原模型的波长提取算法。选择合适的波长提取算法可以进一步提升谱图信息处理的速度。第四,设计编写了中阶梯光栅光谱仪上位机操作软件,集成谱图信息处理算法和相机控制功能于一体,实现仪器的智能化操作。第五,提出了在线光谱定标的方法。利用汞灯谱图数据的采集与分析,在一定容错范围内,通过算法计算实现谱图还原模型的自动校正。自动波长标定完善了中阶梯光栅光谱仪操作软件的功能。第六,通过对各光学元件公差的分析,提出了一种便于装调的结构,实现了仪器的快速装调,为中阶梯光栅光谱仪的商品化、批量化提供了新思路。
[Abstract]:The middle step grating spectrometer is a kind of equipment which can simultaneously achieve high spectral resolution and full spectral direct reading by using cross dispersive array detector, compared with other spectral instruments of the same volume. The middle step grating spectrometer not only has very high spectral resolution, but also can realize the whole spectrum transient direct reading, and it also has wide band and high energy transmission efficiency. With the development of grating fabrication technology and high level detector, the medium step grating spectrometer has shifted from astronomical observation to civil, chemical analysis, quality detection and geological exploration. Environmental monitoring and other fields have been widely used. At present, many foreign companies have launched a series of intermediate ladder grating spectrometer products. However, China is still in the research and development stage of the principle prototype. The bottleneck that restricts the development of the middle step grating spectrometer is the precision and speed of spectrum information processing, and the precision affects the instrument index. Speed determines the performance of the instrument. Based on this, the spectrum information processing method of the middle step grating spectrometer is deeply studied in this paper. The specific research work is as follows: first. In this paper, a method to establish the spectral pattern reduction model based on geometric optical tracing is proposed, based on the analysis of the structural characteristics of the stepped grating spectrometer. A spectral reduction model was established to characterize the relationship between the wavelength and the coordinates of the imaging spot. The error of the whole image plane model was less than one pixel by correcting the curve of the spectral line of the prism. The precision of spectral restoration is greatly improved. Secondly, a background removal algorithm suitable for the original image characteristics of the middle step grating spectrometer is proposed. The features of the original two-dimensional image spot gray value are analyzed. The method of morphological smoothing and edge extraction is used to lock the effective spot and set the gray value of the rest position to zero. This method reduces the data amount of spectral data processing. It not only optimizes SNR but also improves the processing speed of spectral information. Thirdly, according to different imaging characteristics of incident light source with different wavelength characteristics, different wavelength extraction algorithms are proposed. Wavelength extraction algorithm based on speckle centroid calculation is used for continuous and discrete spectrum sources. The wavelength extraction algorithm based on spectral reduction model is adopted. Choosing the appropriate wavelength extraction algorithm can further improve the speed of spectrum information processing. 4th, designed and compiled the upper computer operating software of the middle step grating spectrometer. The integrated spectral image information processing algorithm and camera control function are integrated to realize the intelligent operation of the instrument. 5th, the on-line spectral calibration method is put forward, and the data acquisition and analysis of mercury lamp spectrum diagram is made use of. In a certain range of fault tolerance, automatic calibration of spectral pattern reduction model is realized by algorithm calculation. Automatic wavelength calibration improves the function of operating software of middle step grating spectrometer. 6th. Through the analysis of the tolerance of each optical element, a convenient structure is proposed, which realizes the fast adjustment of the instrument, and provides a new way of thinking for the commercialization and batch quantization of the middle step grating spectrometer.
【学位授予单位】：中国科学院长春光学精密机械与物理研究所
【学位级别】：博士
【学位授予年份】：2017
【分类号】：TP391.41;TH744.1

【相似文献】