东北大学学报(自然科学版) ›› 0, Vol. ›› Issue (): 0-0.

• 信息科学与工程 •    下一篇

渐晕对面阵CCD高温计精度影响及校正方法研究

张育中,胡振伟,孟红记,谢植   

  1. 东北大学信息科学与工程学院
  • 收稿日期:2013-07-25 修回日期:2013-10-02 出版日期:2014-05-15 发布日期:2013-11-22
  • 通讯作者: 张育中
  • 基金资助:

    高碳钢连铸内部质量多目标协调优化与控制研究

Research on Vignetting Influence on Accuracy of Temperature Measurement and Correction Method for CCD-based Pyrometer

  • Received:2013-07-25 Revised:2013-10-02 Online:2014-05-15 Published:2013-11-22
  • Supported by:

    Research on multi-objective coordinated Optimization and Control for internal quality of high-carbon steel in continuous casting

摘要:

基于面阵CCD(charge coupled device)的高温计可获取被测目标整个温度场信息,成为近年来高温计研究的热点。但由于存在光学系统渐晕,直接测量得到的温度场会发生严重畸变,大大降低了测温精度。为提高温度场测量精度,根据辐射测温及几何光学理论,建立了基于面阵CCD温度场测量的温度畸变数学模型,分析了光学参数对温度场测量精度的影响,并提出了一种基于图像邻域灰度梯度分布稀疏特性的渐晕系数估计方法。该方法与利用积分球标定方法相比,其渐晕系数估计的最大绝对误差为0.052,且其有效性在铸坯表面温度场校正实验中得到了进一步验证。

关键词: 渐晕, 面阵CCD, 温度校正, 邻域灰度梯度, 稀疏性

Abstract:

Since the CCD-based pyrometer can obtain temperature field of objects, it has attracted more and more attention in academic studies recently. However, due to the existence of vignetting, there is serious photometric distortion in temperature field, and so the accuracy of temperature measurement is degraded badly. In order to improve the accuracy of temperature field measurement, a mathematical model for temperature distortion based on theories of radiation thermometry and geometrical optics is built in this paper, and the influence of optical system parameters on the accuracy of temperature field measurement is analyzed. Based on the sparsity of neighborhood gray gradient distribution (NGD), a method for the calibration of vignetting coefficient is also proposed in this paper. Compared with the calibration method using a integrating sphere, its maximum absolute error for vignetting estimation is only 0.052, and the experiment of temperature field correction for casting billets further validates the effectiveness of this calibration method.

Key words: vignetting, CCD-area, temperature correction, neighborhood gray gradient, sparsity