-
2004年,Candes等提出了压缩感知理论,是求解欠定方程组稀疏解的一种方法[14-15],后来随着人们对信息需求量的增加,压缩感知被广泛地应用于信号处理领域。传统的信号采样方法是基于经典的香农采样定理,为了不失真地恢复信号,采样频率应该不小于奈奎斯特频率(即信号频谱中的最高频率)的2倍,采集的数据量非常大且存在大量冗余,需要在采集完毕后再进行有损压缩。压缩感知可以避免这个问题,其基本概念是,如果一个信号在某个变换域是稀疏的,便可用一个与变换基不相关的观测矩阵将变换所得的高维信号投影到一个低维空间上,然后通过求解一个最优化问题,就可以从少量的投影中以高概率重构原信号[16],理论框架如图1所示。因此基于压缩感知理论,只要信号是稀疏的,那么它就可以由远低于香农采样定理要求的采样点重建恢复。与传统压缩重建方法所依据信号的连续性和有限带宽相比,压缩感知重建方法利用的是信号在某个域上具有稀疏性(可压缩性)的先验信息。由于大部分信号是可稀疏性的,压缩感知理论具有广泛适用性。同时压缩感知对信号进行亚采样,再用算法消除亚采样导致的伪影,直接在采样过程中实现压缩,避免传统的先采样后压缩导致的资源浪费问题,降低了系统对数据存储和传输能力的要求。
压缩感知测量和重构的过程主要分为两个部分:信号的获取采样和信号的重构恢复。假设x是为长度N的一维原始信号,一般的自然信号x本身并不是稀疏的,需要在某种稀疏基上进行稀疏表示即x=
$ \phi $ s,其中$ \phi $ 为稀疏基矩阵,s为稀疏度是K的稀疏信号。假设$ \varnothing $ 为M×N(M<N)的观测矩阵,对应着亚采样这一过程,将高维信号x投影到低维空间,则长度M的一维测量信号y可表示为:$$ y=\varnothing x=\varnothing \varphi s=\varTheta s $$ (1) 其中令
$ \varTheta $ =$ \varnothing \varphi $ ,即把$ \varnothing \varphi $ 合并成一个矩阵,称之为传感矩阵。因此,基于压缩感知的信号计算问题就是在已知测量信号y和观测矩阵$ \varnothing $ 的基础上,结合优化求解的算法,精确重构出原始信号x。压缩感知理论的核心内容主要体现在信号的稀疏表示、观测矩阵和信号重构算法三个方面。 -
自然界中的大部分信号都能在变换域上进行稀疏表示,如果当原始信号映射到变换域上时,大部分信号都为零或者趋近于零,即可认为原始信号在该变换域上是稀疏的,因此对于变换域的选择至关重要,只有合适的变换域才能保证信号的稀疏度,从而保证信号的重构精度。常见的变换域的选择有离散傅里叶变换(Discrete Fourier Transform,DFT)、离散余弦变换(Discrete Cosine Transform,DCT)、短时间傅里叶变换(Short time Fourier transform,STFT)、离散小波变换(Discrete Wavelet Transform,DWT)等[17]。(1) DFT:傅里叶变换是一种全局变换,它是以两个方向上都无限伸展的三角信号波作为正交基函数,因此无法表述信号的局部性质,适用于确定性信号和平稳信号。(2) DCT:类似于DFT,但只使用实数。(3) STFT:又称窗口傅里叶变换,在DFT中引入了局部化的窗函数,可提取局部信息。当窗函数取为高斯窗时,一般称为Gabor变换,也是最优的窗口傅里叶变换。(4) DWT:为了更好地处理非平稳信号和分析局部信息,小波变换将傅里叶变换中无限长的三角函数基换成了有限长的会衰减的小波基,并且针对STFT中窗函数灵活性差的问题,小波函数窗口的大小、形状可随频率的变化而变化,其变化规律使得小波变换具有良好的局部化特性。常见的小波函数包括Haar系列、Daubechies系列、Coiflets系列、Symlets系列、Biorthogonal系列等[18],根据应用领域的不同,选择合适的小波基系列进行稀疏表示。目前,小波变换由于其优越的特性被广泛用于各种数据信号的处理,在模式识别、语音识别等方面都有较好的发展。
-
在压缩感知理论中,通过变换域得到信号的稀疏表示后,观测矩阵的设计也是一大要点,其目的在于将原始信号投影到低维的观测信号并且保证原始信息不遗漏不破坏,以便能重构出高精度的原始信号。由公式(1)可知,方程的个数远小于未知数的个数(M<N),压缩感知可以视为在一定条件下求解欠定问题。通常欠定问题是没有确定解的,为了能够得到精确的解,Candes等通过理论分析并推导得出确定解的约束条件,即为有限等距性质(Restricted Isometry Property,RIP)[19-20],其定义的公式如下:
$$ (1-{\delta }_{k}){\left|\left|x\right|\right|}_{2}^{2} \leqslant {\left|\right|\varnothing x\left|\right|}_{2}^{2}\leqslant (1+{\delta }_{k}){\left|\right|x\left|\right|}_{2}^{2} $$ (2) 式中:
$ \delta $ k为约束等距常数,范围为(0,1)。$ \delta $ k可以看作是观测矩阵$ \varnothing $ 中K列组成的子集与正交矩阵的相似程度。$ \delta $ k越小说明观测矩阵内的任意K阶列向量之间的正交性越好,利用其对稀疏信号进行感知采样时,传感矩阵各列采集到的信息差异性越大,越有利于进行信号的重构和恢复。然而利用RIP性质作为判据来构造和优化传感矩阵是极其复杂的,不具有实际操作性。据此,Baraniuk给出了RIP的等价条件[20]:如果观测矩阵$ \varnothing $ 和稀疏矩阵$ \phi $ 不相关(即$ \varnothing $ 的行和$ \phi $ 的列不能相互稀疏表示),那么传感矩阵在很大概率上满足RIP准则。由于稀疏矩阵是固定的,因此设计观测矩阵是关键,对矩阵的优化也主要集中在如何使其列间相关性变得更弱。观测矩阵的分类主要有两种:(1)随机测量矩阵,例如高斯矩阵、伯努利矩阵、部分傅里叶矩阵,部分哈达玛矩阵等;(2)确定性测量矩阵,例如托普利兹矩阵、循环矩阵、混沌序列矩阵、多项式矩阵等。压缩感知理论要求观测矩阵与信号尽可能的不相关,所以普遍采用的是性能较好的随机测量矩阵。这类测量矩阵虽然重构结果比较好,但是也存在很多不足:首先,随机矩阵具有不确定性,因此在仿真实验中需要通过大量的实验求均值的方法来降低不确定性对实验结果带来的影响;其次,在实际应用中,该类测量矩阵计算复杂度高,占用存储空间大,而且硬件难以实现。因此,确定性测量矩阵成为测量矩阵新的研究方向。确定性测量矩阵虽然能够很好的弥补硬件实现上的不足,但是该类测量矩阵同样也存在一些不足:和随机测量矩阵相比,重建效果上存在一定的差距;而且由于理论不完善,还存在应用方面的限制。因此观测矩阵的设计需要满足高性能、低复杂度、低存储量、高普适性等主要原则。 -
当原始信号为K阶稀疏且观测矩阵满足RIP时,就能够从低维测量信号中准确重构出高维原始信号,即求解欠定方程可得到一个最优解。压缩感知理论最后一个环节就是信号重构算法,在具体应用与实现的过程中,每一种算法针对于特定的指标来讲都不是最优的,因此选择信号重构算法时有必要做折中处理[21],需要考虑噪声鲁棒性、计算速度、测量信号的数量等因素。若原始信号的先验信息是稀疏表示,公式(1)的逆求解问题就可以转化成求解极小化L0范数问题:
$$ \mathit{min}{\left|\right|s\left|\right|}_{0}\;{\rm{s.t.}}{\left|\right|y-\varTheta s\left|\right|}_{2}\leqslant \varepsilon $$ (3) 式中:
$ {\left|\right|s\left|\right|}_{0} $ 表示s中非零元素个数,即稀疏度;$ \varepsilon $ 表示噪声。但它是一个非确定性多项式问题,即求解需要将所有的可能性逐一尝试,算法的复杂度更高。因此上述极小化L0范数问题并不实用。极小化L1范数是最接近L0的凸优化问题,具有比L0更好的优化求解特性:$$ \mathit{min}{\left|\right|s\left|\right|}_{1}\;{\rm{s.t.}}{\left|\right|y-\varTheta s\left|\right|}_{2}\leqslant \varepsilon $$ (4) 目前压缩感知的信号重构算法主要分为五类[22]:贪婪算法,凸优化算法,贝叶斯算法,非凸优化算法和暴力算法。其中,贪婪算法和凸优化算法的计算效率较高,是比较常用的算法。贪婪算法从测量矩阵中选择与信号最匹配的列向量来构建稀疏逼近,并求出信号残差,然后再选择与信号残差最为匹配的列向量,经过一定次数的迭代,求解出局部最优解。最基础的算法就是由Mallet提出的匹配追踪算法(Matching Pursuit,MP)[23],但MP无法保证每次迭代结果最优,导致计算效率低下,因此研究人员提出了例如正交匹配追踪算法(Orthogonal Matching Pursuit,OMP)[24]、分段OMP(StOMP)算法[25]和正则化OMP(ROMP)算法[26]等改进算法。凸优化算法是一种非线性优化技术,将压缩感知信号重构问题转化为极小化L1范数的凸优化问题,与贪婪算法相比,其优势在于局部最优解即为全局最优解,计算重构精度更高。主要的方法包括基追踪算法(Basis Pursuit,BP)[27]、梯度投影稀疏重建算法(Gradient Projection for Sparse Reconstruction,GPSR)[28]、迭代阈值收缩算法(Iterative Shrinkage-Thresholding,IST)[29]等。
Compressed spectral measurement technology based on coding of spectrum domain
-
摘要: 光谱测量技术在生物医药、国防、安检、生产监控、地质勘测、物质分析、环境保护和减灾防灾等方面有着广泛的应用。但受制于现有探测器件和应用的技术条件,传统类型光谱仪在上述领域的应用灵活性和适用性的限制较多,光谱系统微型化和可集成化是确定发展的趋势之一。光谱成像系统有着向微型化、芯片化和智能化发展的迫切需求,且伴随相关计算光谱成像理论的成熟完善,计算型光谱仪有望在减少器件或系统重量与尺寸的同时,大幅提升光谱分辨能力。基于压缩感知理论的计算型光谱仪具有实时性好、适用范围广、结构调整灵活、成本低廉等诸多优势。文中参考压缩感知理论的基础框架,详细对比多种分光结构的设计方法,分析光谱域直接编码的压缩光谱测量技术,归纳总结具有压缩感知功能的智能芯片化光谱仪的发展趋势和技术问题。Abstract: Spectral detection technology has a wide range of applications in biomedicine, national defense, security check, production monitoring, geological survey, material analysis, environmental protection, disaster reduction and so on. However, due to the existing detectors and application technology conditions, the flexibility and applicability of traditional spectral instruments is limited in above territories. The miniaturization and integration of spectral systems is one of the inevitable development trends. With the critical demand of miniaturized, chip and intelligent spectral imaging and the maturation of the computational spectral imaging theory, computational spectrometers have attracted much attention because of the ability to improve spectral resolution while reducing the mass and volume of devices or systems. The computational spectrometers based on compressed sensing theory have the advantages of short calculation time, wide application range, flexible structure, low cost and so on. This review compares the design methods of various spectral structure based on the framework of compressed sensing theory, analyzes the compression spectral measurement technology which realizes the direct coding of spectrum domain and reveals the development trend and bottleneck of miniaturized, intelligent chip spectrometers based on compressed sensing algorithm.
-
图 2 基于液晶电光效应的计算型光谱仪。(a)液晶可变相位延迟器的光谱测量系统示意图及重构光谱[80];(b)液晶可调谐滤光片结构示意图及重构光谱[81]
Figure 2. Computational spectrometer based on liquid crystal electro-optic effect. (a) Schematic diagram of liquid crystal variable phase retarder spectral measurement system and reconstructed spectra[80]; (b) Schematic diagram of liquid crystal tunable filter structure and reconstructed spectrum[81]
图 3 基于FP腔的计算型光谱仪。(a)压电FP腔光谱测量系统示意图及重构光谱[52];(b) FP微阵列滤光片结构示意图和FP腔两种不同的光谱响应及重构光谱[57]
Figure 3. Computational spectrometer based on FP cavity. (a) Schematic diagram of piezoelectric FP cavity spectral measurement system and reconstructed spectra[52]; (b) Schematic diagram of FP microarray filter and two different spectral responses of FP cavity and reconstructed spectra[57]
图 4 基于光学薄膜的计算型光谱仪。(a)薄膜滤波器阵列光谱仪原理图及产生的随机透过率[58];(b)多层介质滤波器结构示意图和两种不同的透射光谱及重构光谱[60]
Figure 4. Computational spectrometer based on optical thin film. (a) Schematic diagram of thin-film filter array spectrometer and its random transmittance[58]; (b) Schematic diagram of multilayer dielectric filter structure and two different transmission spectra and reconstructed spectra[60]
图 5 基于光子晶体的计算型光谱仪。(a)光子晶体光谱仪工作原理示意图及各种LED光谱的重构[64];(b) Si孔光子晶体板结构示意图及透射光谱和多条洛伦兹线重构信号[65];(c)不同观测矩阵的光谱信号恢复率和相对误差及稀疏度为10和20的光谱重构[66]
Figure 5. Computational spectrometer based on photonic crystal. (a) Schematic diagram of photonic crystal spectrometer and reconstructed spectra of LED[64]; (b) Schematic diagram of Si-hole photonic crystal plate structure and its transmission spectrum and multiple Lorentz line reconstructed signals[65]; (c) Spectral signal recovery rate and relative error of different observation matrices and reconstructed spectra with sparsity of 10 and 20[66]
图 6 基于量子点的计算型光谱仪。(a)胶体量子点滤波器结构和透射光谱及重构的随机和单色光光谱[67];(b)近红外量子点计算型光谱仪工作过程示意图及重构光谱[68];(c)钙钛矿量子点光谱仪结构示意图及其重构光谱[69]
Figure 6. Computational spectrometer based on quantum dots. (a) Structure and transmission spectra of colloidal quantum dot filter and reconstructed spectra of random and monochromatic light[67]; (b) Schematic diagram of NIR quantum dot computational spectrometer and reconstructed spectra[68]; (c) Schematic diagram of perovskite quantum dot spectrometer and reconstructed spectra[69]
图 7 基于纳米线的计算型光谱仪。(a)成分梯度纳米线光谱仪结构示意图及重构光谱[78];(b)硅纳米线阵列结构示意图及重构光谱[79]
Figure 7. Computational spectrometer based on nanowire. (a) Schematic diagram of composition gradient nanowire spectrometer and reconstructed spectra[78]; (b) Schematic diagram of silicon nanowire array structure and reconstructed spectra[79]
图 8 基于等离子体的计算型光谱仪。(a)等离子体滤波阵列光谱传感器原理图和重构光谱[72];(b)铝纳米光栅阵列结构示意图和其光谱响应曲线及重构光谱[73];(c)片上中红外气体传感器阵列结构示意图及重构CO2分子吸收光谱[74];(d)金纳米正方形滤波器阵列结构示意图及重构聚乙烯光谱[75];(e)金纳米圆盘阵列结构示意图及其重构随机光谱
Figure 8. Computational spectrometer based on plasmon. (a) Schematic diagram of plasmon filter array spectral sensor and reconstructed spectra [72]; (b) Schematic diagram of aluminum nano-grating array structure and its spectral response curve and reconstructed spectrum[73]; (c) Schematic diagram of on-chip mid-infrared gas sensor array and reconstructed spectrum of CO2 molecular absorption[74]; (d) Schematic diagram of gold nano-square array and reconstructed spectrum of polyethylene[75]; (e) Schematic diagram of gold nano-disk array structure and reconstructed spectra
-
[1] Green R O, Eastwood M L, Sarture C M, et al. Imaging spectroscopy and the Airborne Visible Infrared Imaging Spectrometer (AVIRIS) [J]. Remote Sensing of Environment, 1998, 65(3): 227-248. doi: 10.1016/S0034-4257(98)00064-9 [2] Whitmore L, Wallace B A. Protein secondary structure analyses from circular dichroism spectroscopy: Methods and reference databases [J]. Biopolymers, 2008, 89(5): 392-400. doi: 10.1002/bip.20853 [3] Perr J M, Furton K G, Almirall J R. Solid phase microextraction ion mobility spectrometer interface for explosive and taggant detection [J]. Journal of Separation Science, 2005, 28(2): 177-183. doi: 10.1002/jssc.200401893 [4] Elmore A J, Shi X, Gorence N J, et al. Spatial distribution of agricultural residue from rice for potential biofuel production in China [J]. Biomass & Bioenergy, 2008, 32(1): 22-27. [5] Bacon J R, Butler O T, Cairns W R L, et al. Atomic spectrometry update - a review of advances in environmental analysis [J]. Journal of Analytical Atomic Spectrometry, 2020, 35(1): 49-53. doi: 10.1039/C9JA90060H [6] Shang J, Meng Q L, Huang R S, et al. Nondestructive testing of kiwifruit quality and maturity by fiber optic spectroscopy [J]. Optics and Precision Engineering, 2021, 29(5): 1190-1198. (in Chinese) doi: 10.37188/OPE.20212905.1190 [7] Su Y, Liao Y-X, Guo Y-L. Extraction of individual component spectra in gas chromatograph coupled with ion trap mass spectrometer by principle component analysis [J]. Acta Chimica Sinica, 2007, 65(14): 1377-1380. [8] Shi Z, Wang Q L, Peng J, et al. Development of a national VNIR soil-spectral library for soil classification and prediction of organic matter concentrations [J]. Science China-Earth Sciences, 2014, 57(7): 1671-1680. doi: 10.1007/s11430-013-4808-x [9] Huang C M, Bi Q, Stiles G S, et al. Fast full search equivalent encoding algorithms for image compression using vector quantization [J]. IEEE Transactions on Image Processing, 1992, 1(3): 413-416. doi: 10.1109/83.148613 [10] Srivastava M, Freed J H. Singular Value Decomposition Method To Determine Distance Distributions in Pulsed Dipolar Electron Spin Resonance: II. Estimating Uncertainty [J]. Journal of Physical Chemistry A, 2019, 123(1): 359-370. doi: 10.1021/acs.jpca.8b07673 [11] Wang P, Menon R. Computational spectroscopy via singular-value decomposition and regularization [J]. Optics Express, 2014, 22(18): 21541-2150. doi: 10.1364/OE.22.021541 [12] Chang C-C, Lee H-N. On the estimation of target spectrum for filter-array based spectrometers [J]. Optics Express, 2008, 16(2): 1056-1061. doi: 10.1364/OE.16.001056 [13] Donoho D L. Compressed sensing [J]. IEEE Transactions on Information Theory, 2006, 52(4): 1289-306. doi: 10.1109/TIT.2006.871582 [14] Candes E J, Wakin M B. An introduction to compressive sampling [J]. IEEE Signal Processing Magazine, 2008, 25(2): 21-30. doi: 10.1109/MSP.2007.914731 [15] Candes E J, Romberg J, Tao T. Robust uncertainty principles: Exact signal reconstruction from highly incomplete frequency information [J]. IEEE Transactions on Information Theory, 2006, 52(2): 489-509. doi: 10.1109/TIT.2005.862083 [16] 蓝若明. 基于压缩感知的单光子计数光谱测量技术研究[D].北京: 中国科学院国家空间科学中心, 2016. Lan R M. Single-photon counting spectral measurement based on compressed sensing [D]. Beijing: National Space Science Center, Chinese Academy of Sciences, 2016. (in Chinese) [17] 王志强. 基于C++的纸病图像数据的压缩感知的研究 [D].西安: 陕西科技大学, 2016. Wang Z Q. Research on compressed sensing of diseased image data based on C++ [D]. Xi'an: University of Science and Technology, 2016. (in Chinese) [18] Jiao L C, Tan S. Multiscale geometric analysis of image: Review and prospect [J]. Acta Electronica Sinica, 2003(S1): 1975-1981. (in Chinese) [19] Baraniuk R, Davenport M, Devore R, et al. A simple proof of the restricted isometry property for random matrices [J]. Constructive Approximation, 2008, 28(3): 253-263. [20] Baraniuk R G. Compressive sensing [J]. IEEE Signal Processing Magazine, 2007, 24(4): 118-121. doi: 10.1109/MSP.2007.4286571 [21] 王爽. 基于电光调控透过率编码的压缩感知光谱测量研究[D].太原: 中北大学, 2020. Wang S. Research on compressed sensing spectral measurement based on electro-optic regulated transmissivity coding [D]. Taiyuan: North University of China, 2020. (in Chinese) [22] Tropp J A, Wright S J. Computational methods for sparse solution of linear inverse problems [J]. Proceedings of the IEEE, 2010, 98(6): 948-958. doi: 10.1109/JPROC.2010.2044010 [23] Mallat S G, Zhang Z F. Matching pursuits with time-frequency dictionaries [J]. IEEE Transactions on Signal Processing, 1993, 41(12): 3397-3415. doi: 10.1109/78.258082 [24] Tropp J A, Gilbert A C. Signal recovery from random measurements via orthogonal matching pursuit [J]. IEEE Transactions on Information Theory, 2007, 53(12): 4655-4666. doi: 10.1109/TIT.2007.909108 [25] Donoho D L, Tsaig Y, Drori I, et al. Sparse solution of underdetermined systems of linear equations by stagewise orthogonal matching pursuit [J]. IEEE Transactions on Information Theory, 2012, 58(2): 1094-1121. doi: 10.1109/TIT.2011.2173241 [26] Needell D, Vershynin R. Signal recovery from Incomplete and inaccurate measurements via regularized orthogonal matching pursuit [J]. IEEE Journal of Selected Topics in Signal Processing, 2010, 4(2): 310-316. doi: 10.1109/JSTSP.2010.2042412 [27] Kim S-J, Koh K, Lustig M, et al. An interior-point method for large-scale l(1)-regularized least squares [J]. IEEE Journal of Selected Topics in Signal Processing, 2007, 1(4): 606-617. doi: 10.1109/JSTSP.2007.910971 [28] Figueiredo Ma T, Nowak R D, Wright S J. Gradient projection for sparse reconstruction: Application to compressed sensing and other inverse problems [J]. IEEE Journal of Selected Topics in Signal Processing, 2007, 1(4): 586-597. doi: 10.1109/JSTSP.2007.910281 [29] Wright S J, Nowak R D, Figueiredo M a T. Sparse reconstruction by separable approximation [J]. IEEE Transactions on Signal Processing, 2009, 57(7): 2479-2493. doi: 10.1109/TSP.2009.2016892 [30] Yang Z, Albrow-Owen T, Cai W, et al. Miniaturization of optical spectrometers [J]. Science, 2021, 371(6528): 480. [31] Zhao Y S, He W J, Liu Z Y, et al. Development of convex flaring grating in coded aperture spectral Imager [J]. Infrared and Laser Engineering, 2022,51(03): 20220007. (in Chinese) [32] Gao Z D, Gao H X, Zhu Y Y, et al. Summary of snapshot spectral imaging technology [J]. Optics and Precision Engineering, 2020, 28(6): 1323-1343. (in Chinese) doi: 10.3788/OPE.20202806.1323 [33] Zheng Q, Wen L, Chen Q. Research progress of computational microspectrometer based on speckle inspection [J]. Opto-Electronic Engineering, 2021, 48(3): 4-8. [34] Redding B, Cao H. Using a multimode fiber as a high-resolution, low-loss spectrometer [J]. Optics Letters, 2012, 37(16): 3384-3386. doi: 10.1364/OL.37.003384 [35] Liew S F, Redding B, Choma M A, et al. Broadband multimode fiber spectrometer [J]. Optics Letters, 2016, 41(9): 2029-2032. doi: 10.1364/OL.41.002029 [36] Meng Z, Li J, Yin C, et al. Multimode fiber spectrometer with scalable bandwidth using space-division multiplexing [J]. Aip Advances, 2019, 9(1): 015004. doi: DOI:10.1063/1.5052276 [37] Wan N H, Meng F, Schroeder T, et al. High-resolution optical spectroscopy using multimode interference in a compact tapered fibre [J]. Nature Communications, 2015, 6(1): 7762. doi: DOI:10.1038/ncomms8762 [38] Sefler G A, Shaw T J, Valley G C. Demonstration of speckle-based compressive sensing system for recovering RF signals [J]. Optics Express, 2018, 26(17): 21390-21402. doi: 10.1364/OE.26.021390 [39] Redding B, Liew S F, Bromberg Y, et al. Evanescently coupled multimode spiral spectrometer [J]. Optica, 2016, 3(9): 956-962. doi: 10.1364/OPTICA.3.000956 [40] Piels M, Zibar D. Compact silicon multimode waveguide spectrometer with enhanced bandwidth [J]. Scientific Reports, 2017, 7(1): 43454. doi: DOI:10.1038/srep43454 [41] Redding B, Liew S F, Sarma R, et al. Compact spectrometer based on a disordered photonic chip [J]. Nature Photonics, 2013, 7(9): 746-751. doi: 10.1038/nphoton.2013.190 [42] Yang T, Xu C, Ho H-P, et al. Miniature spectrometer based on diffraction in a dispersive hole array [J]. Optics Letters, 2015, 40(13): 3217-3220. doi: 10.1364/OL.40.003217 [43] Gat N. Imaging spectroscopy using tunable filters: A review. Conference on Wavelet Applications VII[C]//Proc SPIE, 2000, 4056: 50-64. [44] Carmo J P, Rocha R P, Bartek M, et al. A review of visible-range Fabry-Perot microspectrometers in silicon for the industry [J]. Optics and Laser Technology, 2012, 44(7): 2312-2320. doi: 10.1016/j.optlastec.2012.03.036 [45] Zhang H Y, Wang X L, Soos J, et al. Design of a miniature solid-state NIR spectrometer[C]// Conference on Infrared Detectors and Instrumentation for Astronomy, 1995, 2475: 376-383. [46] Herrmann H, Schafer K, Schmidt C. Low-loss tunable integrated acoustooptical wavelength filter in LiNbO3 with strong sidelobe suppression [J]. IEEE Photonics Technology Letters, 1998, 10(1): 120-122. doi: 10.1109/68.651131 [47] Guarino A, Poberaj G, Rezzonico D, et al. Electro-optically tunable microring resonators in lithium niobate [J]. Nature Photonics, 2007, 1(7): 407-410. doi: 10.1038/nphoton.2007.93 [48] Yao Y, Hou J, Liu H, et al. Design of programmable multi-wavelength tunable filter on lithium niobate [J]. Results in Physics, 2019, 15: 102741. [49] Miao Y P, Wu J X, Lin W, et al. Magnetic field tunability of optical microfiber taper integrated with ferrofluid [J]. Optics Express, 2013, 21(24): 29914-29920. doi: 10.1364/OE.21.029914 [50] Mallinson S R, Jerman J H. Miniature micromachined Fabry-Perot interferometers in silicon [J]. Electronics Letters, 1987, 23(20): 1041-1043. doi: 10.1049/el:19870728 [51] Stone J, Stulz L W, Marcuse D, et al. Narrow-band FiEnd etalon filters using expanded-core fibers [J]. Journal of Lightwave Technology, 1992, 10(12): 1851-1854. doi: 10.1109/50.202838 [52] Oiknine Y, August I, Blumberg D G, et al. Compressive sensing resonator spectroscopy [J]. Optics Letters, 2017, 42(1): 25-28. doi: 10.1364/OL.42.000025 [53] Yu X C, Xu Y Q, Cai J C, et al. Research progress of tunable micro-nano filter structure [J]. Chinese Optics, 2021, 14(5): 1069-1088. (in Chinese) doi: 10.37188/CO.2021-0044 [54] Correia J H, De Graaf G, Kong S H, et al. Single-chip CMOS optical microspectrometer [J]. Sensors and Actuators a-Physical, 2000, 82(1-3): 191-197. doi: 10.1016/S0924-4247(99)00369-6 [55] Kong S H, Correia J H, De Graaf G, et al. Integrated silicon microspectrometers [J]. IEEE Instrumentation & Measurement Magazine, 2001, 4(3): 34-38. [56] Emadi A, Wu H, De Graaf G, et al. Design and implementation of a sub-nm resolution microspectrometer based on a linear-variable optical filter [J]. Optics Express, 2012, 20(1): 489-507. doi: 10.1364/OE.20.000489 [57] 吴笑男. 基于压缩感知的改进FP微阵列光谱调制重建研究 [硕士]: 中北大学; 2021. Wu X N. Research on Spectral Modulation Reconstruction of Improved FP Microarray Based on Compressed Sensing [M]. Taiyuan: North University of China, 2021. (in Chinese) [58] Oliver J, Lee W-B, Lee H-N. Filters with random transmittance for improving resolution in filter-array-based spectrometers [J]. Optics Express, 2013, 21(4): 3969-3689. doi: 10.1364/OE.21.003969 [59] Wang S-W, Xia C, Chen X, et al. Concept of a high-resolution miniature spectrometer using an integrated filter array [J]. Optics Letters, 2007, 32(6): 632-640. doi: 10.1364/OL.32.000632 [60] Lee W-B, Kim C, Ju G W, et al. Design of thin-film filters for resolution improvements in filter-array based spectrometers using DSP [J]. Next-Generation Spectroscopic Technologies Ix, 2016, 9855: 98550. [61] Kim S H, Park H S, Choi J H, et al. Integration of colloidal photonic crystals toward miniaturized spectrometers [J]. Advanced Materials, 2010, 22(9): 946. doi: 10.1002/adma.200902456 [62] Momeni B, Hosseini E S, Adibi A. Planar photonic crystal microspectrometers in silicon-nitride for the visible range [J]. Optics Express, 2009, 17(19): 17060-17069. doi: 10.1364/OE.17.017060 [63] Wang Z, Yi S, Chen A, et al. Single-shot on-chip spectral sensors based on photonic crystal slabs [J]. Nature Communications, 2019, 10(1): 1020. doi: DOI:10.1038/s41467-019-08994-5 [64] Bryan K M, Jia Z, Pervez N K, et al. Inexpensive photonic crystal spectrometer for colorimetric sensing applications [J]. Optics Express, 2013, 21(4): 4411-4423. doi: 10.1364/OE.21.004411 [65] Wang Z, Yu Z. Spectral analysis based on compressive sensing in nanophotonic structures [J]. Optics Express, 2014, 22(21): 25608-25614. doi: 10.1364/OE.22.025608 [66] Guo P, Wang Z, Shi B, et al. Compressive sensing based on mesoscopic chaos of silicon optomechanical photonic crystal [J]. IEEE Photonics Journal, 2020, 12(5): 3022801. [67] Bao J, Bawendi M G. A colloidal quantum dot spectrometer [J]. Nature, 2015, 523(7558): 67. doi: 10.1038/nature14576 [68] Li H Y, Bian L H, Gu K, et al. A near-infrared miniature quantum dot spectrometer [J]. Advanced Optical Materials, 2021, 9(15): 0376. [69] Zhu X, Bian L, Fu H, et al. Broadband perovskite quantum dot spectrometer beyond human visual resolution [J]. Light-Science Applications, 2020, 9(1): 73. [70] Kurokawa U, Choi B I, Chang C-C. Filter-based miniature spectrometers: Spectrum reconstruction using adaptive regularization [J]. IEEE Sensors Journal, 2011, 11(7): 1556-1563. doi: 10.1109/JSEN.2010.2103054 [71] Tittl A, Leitis A, Liu M K, et al. Imaging-based molecular barcoding with pixelated dielectric metasurfaces [J]. Science, 2018, 360(6393): 1105. doi: 10.1126/science.aas9768 [72] Chang C-C, Lin N-T, Kurokawa U, et al. Spectrum reconstruction for filter-array spectrum sensor from sparse template selection [J]. Optical Engineering, 2011, 50(11): 4402. [73] Cerjan B, Halas N J. Toward a nanophotonic nose: A compressive sensing-enhanced, optoelectronic mid-infrared spectrometer [J]. Acs Photonics, 2019, 6(1): 79-86. doi: 10.1021/acsphotonics.8b01503 [74] Chen Q, Liang L, Zheng Q, et al. On-chip readout plasmonic mid-IR gas sensor [J]. Opto-Electronic Advances, 2020, 3(7): 190040. [75] Craig B, Shrestha V R, Meng J, et al. Experimental demonstration of infrared spectral reconstruction using plasmonic metasurfaces [J]. Optics Letters, 2018, 43(18): 4481-4484. doi: 10.1364/OL.43.004481 [76] Zheng B, Li L, Wang J, et al. On-chip measurement of photoluminescence with high sensitivity monolithic spectrometer [J]. Advanced Optical Materials, 2020, 8(11): 2000191. [77] Dong J J. The smallest nanowire spectrometers [J]. Frontiers of Optoelectronics, 2019, 12(4): 341. doi: 10.1007/s12200-019-0983-5 [78] Yang Z, Albrow-Owen T, Cui H, et al. Single-nanowire spectrometers [J]. Science, 2019, 365(6457): 1017. doi: 10.1126/science.aax8814 [79] Meng J, Cadusch J J, Crozier K B. Detector-only spectrometer based on structurally colored silicon nanowires and a reconstruction algorithm [J]. Nano Letters, 2020, 20(1): 320-328. doi: 10.1021/acs.nanolett.9b03862 [80] August Y, Stern A. Compressive sensing spectrometry based on liquid crystal devices [J]. Optics Letters, 2013, 38(23): 4996-4999. doi: 10.1364/OL.38.004996 [81] Wang Q, Ma L, Li C, et al. A spectral super-resolution method of LCTF based on compressive sensing [J]. Transactions of Beijing Institute of Technology, 2018, 38(1): 40-45, 72. [82] Jewell S A, Vukusic P, Roberts N W. Circularly polarized colour reflection from helicoidal structures in the beetle Plusiotis boucardi [J]. New Journal of Physics, 2007, 9(4): 99. [83] Lowrey S, De Silva L, Hodgkinson I, et al. Observation and modeling of polarized light from scarab beetles [J]. Journal of the Optical Society of America a-Optics Image Science and Vision, 2007, 24(8): 2418-2425. doi: 10.1364/JOSAA.24.002418 [84] Guo B. Photonic band gap structures of obliquely incident electromagnetic wave propagation in a one-dimension absorptive plasma photonic crystal [J]. Physics of Plasmas, 2009, 16(4): 043508. [85] Hermann D, Diem M, Mingaleev S F, et al. Photonic crystals with anomalous dispersion: Unconventional propagating modes in the photonic band gap [J]. Physical Review B, 2008, 77(3): 035112. [86] Argyros A, Birks T A, Leon-Saval S G, et al. Photonic bandgap with an index step of one percent [J]. Optics Express, 2005, 13(1): 309-314. doi: 10.1364/OPEX.13.000309 [87] Meng Q, Ouyang Z, Wang Jong C, et al. Mode types and their related properties of one-dimensional photonic crystal resonant cavity [J]. Acta Optica Sinica, 2007, 27(7): 1290-1294. [88] Zou M Q, Yang R, Li J F, et al. Optical properties of quantum dots and their applications in life science [J]. Journal of Analytical Measurement, 2005(6): 133-137. (in Chinese) doi: 10.3969/j.issn.1004-4957.2005.06.036 [89] Zhang D, Wang P, Jiao X, et al. Progress in surface plasmon subwavelength optics [J]. Physics, 2005, 34(7): 508-512. [90] Zhen-Lin W. A review on research progress in surface plasmons [J]. Progress in Physics, 2009, 29(3): 287-324. [91] Ebbesen T W, Lezec H J, Ghaemi H F, et al. Extraordinary optical transmission through sub-wavelength hole arrays [J]. Nature, 1998, 391(6668): 667-669. doi: 10.1038/35570 [92] Liu H, Lalanne P. Microscopic theory of the extraordinary optical transmission [J]. Nature, 2008, 452(7188): 728-731. doi: 10.1038/nature06762 [93] Li M X, Wang D Y, Zhang C. Principle and application of metamaterial surface structure color [J]. Chinese Optics, 2021, 14(4): 900-926. (in Chinese) doi: 10.37188/CO.2021-0108 [94] Sun H, Tian W, Wang X, et al. In situ formed gradient bandgap-tunable perovskite for ultrahigh-speed color/spectrum-sensitive photodetectors via electron-donor control [J]. Advanced Materials, 2020, 32(14): e1908108. [95] Zhang M N, Wu X, Riaud A, et al. Spectrum projection with a bandgap-gradient perovskite cell for colour perception[J]. Light-Science & Applications. 2020, 9(1):324-327. [96] Crocombe R A. Portable spectroscopy [J]. Applied Spectroscopy, 2018, 72(12): 1701-1751. doi: 10.1177/0003702818809719 [97] Malinen J, Rissanen A, Saari H, et al. Advances in miniature spectrometer and sensor development[C]//Sensing Technologies + Applications, 2014, 9101: 91010C. [98] Herrero-Bermello A, Velasco A V, Podmore H, et al. Temperature dependence mitigation in stationary Fourier-transform on-chip spectrometers [J]. Optics Letters, 2017, 42(11): 2239-2242. doi: 10.1364/OL.42.002239 [99] Ng W, Minasny B, Montazerolghaem M, et al. Convolutional neural network for simultaneous prediction of several soil properties using visible/near-infrared, mid-infrared, and their combined spectra [J]. Geoderma, 2019, 352: 251-267. doi: 10.1016/j.geoderma.2019.06.016 [100] Xiao D, Ba Tuan L, Thai Thuy Lam H. Iron ore identification method using reflectance spectrometer and a deep neural network framework [J]. Spectrochimica Act a Part A: Molecular and Biomolecular Spectroscopy, 2020, 248: 119168. doi: DOI:10.1016/J.SAA.2020.119168 [101] Li D L, Lu B. Optical fiber sensor recognition algorithm based on deep neural network [J]. Infrared and Laser Engineering, 2022, 51(9): 20210971. (in Chinese) [102] Jha D, Choudhary K, Tavazza F, et al. Enhancing materials property prediction by leveraging computational and experimental data using deep transfer learning [J]. Nature Communications, 2019, 10(1): 5316. doi: DOI:10.1038/s41467-019-13297-w [103] Smith J S, Nebgen B T, Zubatyuk R, et al. Approaching coupled cluster accuracy with a general-purpose neural network potential through transfer learning [J]. Nature Communications, 2019, 10(1): 2903.