Citation: | Chu Songnan, Liu Haitao, Hu Qiqi, et al. An accurate measurement method for the spatial resolution of area array spectral imaging equipment[J]. Opto-Electronic Engineering, 2019, 46(11): 180458. doi: 10.12086/oee.2019.180458 |
[1] | 张声荣.光电成象遥感器研制中一些基本参数的选择[J].中国空间科学技术, 1981(2): 18-21, 17. Zhang S R. Basic parameter selection in the development of electro-optical imaging remote sensor[J]. Chinese Space Science and Technology, 1981(2): 18-21, 17. |
[2] | 丁晓波.感光胶片与CCD的静态分辨力比较[J].影像技术, 1997(2): 50-51. Ding X B. Comparison of static resolving power of CCD and photographic film[J]. Image Technology, 1997(2): 50-51. |
[3] | Holst G C. Imaging system performance based upon Fλ/d[J]. Optical Engineering, 2007, 46(10): 103204. doi: 10.1117/1.2790066 |
[4] | 吴启海, 胡峻, 邬子刚, 等.数码照相机分辨率的测量: GB/T 19953-2005[S].北京: 中国标准出版社, 2006. Wu Q H, Hu J, Wu Z G, et al. Digital still cameras-Resolution measurements: GB/T 19953-2005[S]. Beijing: China Standard Press, 2006. |
[5] | Polder G, van der Heijden G W A M. Calibration and characterization of spectral imaging systems[J]. Proceedings of SPIE, 2001, 4548: 10-17. doi: 10.1117/12.441362 |
[6] | Wang W L, Li C Y, Tollner E W, et al. A liquid crystal tunable filter based shortwave infrared spectral imaging system: calibration and characterization[J]. Computers and Electronics in Agriculture, 2012, 80: 135-144. doi: 10.1016/j.compag.2011.09.003 |
[7] | 王凌, 张平, 冯华君, 等.用多次移位成像提高CCD成像分辨力的反演解析[J].光电工程, 2003, 30(3): 62-65. doi: 10.3969/j.issn.1003-501X.2003.03.019 Wang L, Zhang P, Feng H J, et al. An inversion analysis method based on multi-shift imaging for improvement of CCD imaging resolution[J]. Opto-electronic Engineering, 2003, 30(3): 62-65. doi: 10.3969/j.issn.1003-501X.2003.03.019 |
[8] | Guzhov V I, Il'inykh S P, Marchenko I O. Method of increasing the spatial resolution in digital holographic microscopy[J]. Optoelectronics, Instrumentation and Data Processing, 2018, 54(3): 301-306. doi: 10.3103/S8756699018030135 |
[9] | 杨桦, 焦文春, 朱永红, 等. CCD相机在系统奈奎斯特频率处的调制传递函数[J].光学学报, 2002, 22(3): 313-316. doi: 10.3321/j.issn:0253-2239.2002.03.013 Yang H, Jiao W C, Zhu Y H, et al. Modulation transfer function of CCD camera at Nyquist frequency[J]. Acta Optica Sinica, 2002, 22(3): 313-316. doi: 10.3321/j.issn:0253-2239.2002.03.013 |
Overview: Spatial resolution is an important performance parameter of digital imaging equipment. At present, the commonly used laboratory measurement method for human eyes is to interpret the image of the spatial resolution testing board. In the field of spectral imaging, a quantitative measurement method of spatial resolution based on data processing has been established. The spectral image of black-white line pair resolution testing board is taken in this method. If the ratio of the adjacent gray peak value and gray valley value (PVR) corresponding to the black-white line pair on the image is larger than $\sqrt 2 $, the line pair can be distinguished. However, when the detector is not capable of resolving the targets, the method has the phenomenon of different measurement results caused by different imaging locations, which will make the results obtained by different experiments difficult to be consistent. In order to solve the above problems, this paper proposes an accurate measurement method for the spatial resolution of the camera of array spectral imaging. By taking a series of spectral images of black-white line pairs under continuous precision displacement, the curve of variation of gray level with displacement of two adjacent pixels in the translation direction is drawn. If the difference in pixel sizes is ignored, the spacing between two curves in the translation direction corresponds to the imaging range value of a single pixel. With arbitrary points as the starting point and with the imaging range value of a single pixel as the space to segment the smooth gray level curve of a single pixel varying with displacement, the distribution results of gray level with pixel in various possible imaging locations can be obtained theoretically. Then, the minimum value of PVR value is found. If the minimum value is greater than $\sqrt 2 $, the black-white line pair can be distinguished at all imaging locations. If the minimum value is less than $\sqrt 2 $, the black-white line pair cannot be distinguished at the some imaging locations. As the gray level distribution of the black-white line pair image is close to sinusoidal distribution near Nyquist frequency, the moving step of imaging can be reduced by curve fitting, and the curve of pixel gray level changing with displacement is more smooth. Then, the paper proves that a simple curve segmentation method can directly obtain the minimum PVR value, which significantly reduces the data processing capacity of this method. At last, the feasibility and correctness of the method are verified by the experimental measurement for an area array spectral imaging equipment. This method can effectively avoid the defects of existing methods when the resolution of detector is insufficient, the data processing is relatively simple, and the results are reliable, which is conducive to maintaining the consistency of the measurement results of spatial resolution between different experiments. In addition, this method can also be applied to the accurate measurement of the modulation transfer function of the array imaging system, so as to quickly find the measurement target matching the Nyquist frequency of the system.
Three types of typical imging result at Nyquist frequency
Imaging results of blackwhite lines with accurate shift
Dimensional relevance
Example for curve division
Curve division for PVR minimum
Typical results
The results of division points in figure 6 moving to the right
The resolution board for testing
The imaging results and the correspondence of blackwhite lines
Gray level variation of adjacent pixels with shift. (a) 18 lp/mm; (b) 20 lp/mm