廖清君, 胡晓宁, 黄爱波, 陈洪雷, 叶振华, 丁瑞军. 拼接型短波红外探测器的光谱响应特性[J]. 红外与激光工程, 2023, 52(9): 20220890. DOI: 10.3788/IRLA20220890
引用本文: 廖清君, 胡晓宁, 黄爱波, 陈洪雷, 叶振华, 丁瑞军. 拼接型短波红外探测器的光谱响应特性[J]. 红外与激光工程, 2023, 52(9): 20220890. DOI: 10.3788/IRLA20220890
Liao Qingjun, Hu Xiaoning, Huang Aibo, Chen Honglei, Ye Zhenhua, Ding Ruijun. Spectral responsivity of mosaic SWIR detectors[J]. Infrared and Laser Engineering, 2023, 52(9): 20220890. DOI: 10.3788/IRLA20220890
Citation: Liao Qingjun, Hu Xiaoning, Huang Aibo, Chen Honglei, Ye Zhenhua, Ding Ruijun. Spectral responsivity of mosaic SWIR detectors[J]. Infrared and Laser Engineering, 2023, 52(9): 20220890. DOI: 10.3788/IRLA20220890

拼接型短波红外探测器的光谱响应特性

Spectral responsivity of mosaic SWIR detectors

  • 摘要: 红外探测器的光谱响应一致性影响高光谱成像仪器的动态范围,研究高光谱成像用拼接型短波红外探测器在同一光谱维的响应均匀性对提高高光谱成像性能有重要意义。通过测量相对光谱响应和窄带响应,对响应波段为1.0~2.5 μm、规格为2000×256的碲镉汞短波红外探测器光谱响应率进行测量和分析,提出用光谱响应非均匀性定量化分析光谱响应一致性。分析了在80 ℃和140 ℃不同的黑体温度下,窄带滤光片的中心波长和半带宽不同时,带外截止深度为OD3时,带外信号对窄带性能测试误差的影响。通过测量探测器模块的光谱响应率,计算拼接的2000×256探测器在1 μm、1.9 μm和2.5 μm处的响应非均匀性分别为6.23%、6.06%和4.07%。光谱响应率的准确测量实现了拼接型短波红外探测器的光谱响应一致性的定量化评价,有利于探测器在高光谱成像中的合理应用。

     

    Abstract:
      Objective  Hyperspectral imaging can not only get the two-dimensional geometric spatial information of the observed objects, but also obtain the continuous high-resolution spectral information which can reflect the physical and chemical characteristics of the target. It is a very important method for target detection and recognition based on hyperspectral remote sensing information. Spectral range of typical imaging spectrometer is 0.4-2.5 µm due to the ground objects' reflection of solar radiation. Mercury Cadmium Telluride (Hg1-xCdxTe) detectors cover a bandwidth of 0.8-30 µm as the alloy composition of Hg1-xCdxTe material is tuned in terms of cut-off wavelength. Hg1-xCdxTe detectors are the major part of the imaging spectrometer for detection in short waveband. As the swath width of the imaging spectrometer increased, larger scale infrared focal plane array (IRFPA) is needed. Mosaic ultra-large scale shortwave infrared (SWIR) detectors can meet the demand for wide field of view detection in space application. The detector modules for butting have their own spectral responsivity. Hyperspectral imaging demands that the mosaic IRFPA has high uniformity of the spectral response. Therefore, it is necessary to measure and analyse the spectral responsivity specification of the mosaic IRFPA accurately and quantitatively for the hyperspectral imaging application. For this purpose, a method for evaluating the absolute spectral responsivity of the mosaic SWIR detectors is proposed in this paper.
      Methods   This paper presents a method for measuring the absolute spectral responsivity accurately and quantitative analysis of the spectral responsivity specification of the mosaic 2 000×256 SWIR detector for imaging spectrometer. The relative response spectrum is measured by a precisely calibrated grating monochromator system. Five optical filters with different center wavelength (CW) and full width at half maximum (FWHM) were chosen to analyze and measure the narrow band responsivity (Tab.1). The center wavelength of the filter is 1225 nm, 1670 nm, 2062 nm, 2420 nm and 2470 nm respectively. The bandwidth is 10 nm and 50 nm, and the cut-off depth is OD3 (optical density). Spectral responsivity is calculated by relative response and narrow-band responsivity.
      Results and Discussions   The cut-off wavelength of detector to be tested is 2.6 μm, and its pitch size is 30 μm×60 μm. The integration time of the read-out integrated circuit (ROIC) is 4.4 ms and integration capacity is 65 fF. F number of the Dewar is 0.9. The results of output signal analysis with filter of different CW at different black body temperature show that narrow-band responsivity is much lower than out-of-band response (Tab.2, Fig.3) with 1# filter and much higher (Tab.2, Fig.4) with 5# filter. The possibility of narrow-band signal's accurate measurement at 1200 nm is discussed if the bandwidth is widened to 200 nm and the cut-off depth is adapted to OD4 and OD5 (Tab.3). It shows that narrow band responsivity can be measured precisely only when cut-off depth is smaller than OD5 and FWHM is wider than 200 nm. Based on the result of the analysis, for HgCdTe SWIR detector the measurement error is smallest when the filter's center wavelength is 2470 nm, FWHM is 50 nm, and cut-off depth is OD3 at 80 ℃ black body temperature. The absolute spectral responsivity of four HgCdTe detectors is measured by the relative response curve and narrow-band responsivity (Fig.7). According to the spectral responsivity curve, the responsivity non-uniformity of four detectors can be calculated to be 6.23%, 6.06%, 4.07% at 1 μm, 1.9 μm and 2.5 μm respectively (Fig.8).
      Conclusions  In this study, a quantitative method for measuring the spectral responsivity accurately and analyzing the spectral responsivity specification of the mosaic 2000×256 SWIR detector for imaging spectrometer is proposed. The results of this study demonstrated that spectral responsivity of Hg1-xCdxTe SWIR can be measured accurately when the filter's center wavelength is 2470 nm, FWHM is 50 nm, and cut-off depth is OD3 at 80 ℃ black body temperature. Narrow-band spectral response output signal is much larger than signal caused by out-of-band response. The spectral responsivity non-uniformity of the four detectors helps to evaluate the response uniformity of spectral dimension response of 2000×256 SWIR detector quantitatively. The results have demonstrated that the use of this measuring method promotes appropriate application of IRFPA detectors in hyperspectral imaging.

     

/

返回文章
返回