WO2012146156A1 - 一种单光子计数成像系统及其方法 - Google Patents

一种单光子计数成像系统及其方法 Download PDF

Info

Publication number
WO2012146156A1
WO2012146156A1 PCT/CN2012/074533 CN2012074533W WO2012146156A1 WO 2012146156 A1 WO2012146156 A1 WO 2012146156A1 CN 2012074533 W CN2012074533 W CN 2012074533W WO 2012146156 A1 WO2012146156 A1 WO 2012146156A1
Authority
WO
WIPO (PCT)
Prior art keywords
single photon
lens
control system
photon
dmd control
Prior art date
Application number
PCT/CN2012/074533
Other languages
English (en)
French (fr)
Inventor
翟光杰
杜克铭
王超
俞文凯
Original Assignee
中国科学院空间科学与应用研究中心
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by 中国科学院空间科学与应用研究中心 filed Critical 中国科学院空间科学与应用研究中心
Priority to US14/000,421 priority Critical patent/US8723130B2/en
Priority to JP2014506741A priority patent/JP6211512B2/ja
Priority to EP12776896.8A priority patent/EP2685227B1/en
Publication of WO2012146156A1 publication Critical patent/WO2012146156A1/zh

Links

Classifications

    • GPHYSICS
    • G01MEASURING; TESTING
    • G01JMEASUREMENT OF INTENSITY, VELOCITY, SPECTRAL CONTENT, POLARISATION, PHASE OR PULSE CHARACTERISTICS OF INFRARED, VISIBLE OR ULTRAVIOLET LIGHT; COLORIMETRY; RADIATION PYROMETRY
    • G01J1/00Photometry, e.g. photographic exposure meter
    • G01J1/42Photometry, e.g. photographic exposure meter using electric radiation detectors
    • G01J1/44Electric circuits
    • GPHYSICS
    • G01MEASURING; TESTING
    • G01JMEASUREMENT OF INTENSITY, VELOCITY, SPECTRAL CONTENT, POLARISATION, PHASE OR PULSE CHARACTERISTICS OF INFRARED, VISIBLE OR ULTRAVIOLET LIGHT; COLORIMETRY; RADIATION PYROMETRY
    • G01J1/00Photometry, e.g. photographic exposure meter
    • G01J1/02Details
    • G01J1/04Optical or mechanical part supplementary adjustable parts
    • G01J1/0407Optical elements not provided otherwise, e.g. manifolds, windows, holograms, gratings
    • G01J1/0414Optical elements not provided otherwise, e.g. manifolds, windows, holograms, gratings using plane or convex mirrors, parallel phase plates, or plane beam-splitters
    • GPHYSICS
    • G01MEASURING; TESTING
    • G01JMEASUREMENT OF INTENSITY, VELOCITY, SPECTRAL CONTENT, POLARISATION, PHASE OR PULSE CHARACTERISTICS OF INFRARED, VISIBLE OR ULTRAVIOLET LIGHT; COLORIMETRY; RADIATION PYROMETRY
    • G01J1/00Photometry, e.g. photographic exposure meter
    • G01J1/42Photometry, e.g. photographic exposure meter using electric radiation detectors
    • GPHYSICS
    • G01MEASURING; TESTING
    • G01JMEASUREMENT OF INTENSITY, VELOCITY, SPECTRAL CONTENT, POLARISATION, PHASE OR PULSE CHARACTERISTICS OF INFRARED, VISIBLE OR ULTRAVIOLET LIGHT; COLORIMETRY; RADIATION PYROMETRY
    • G01J1/00Photometry, e.g. photographic exposure meter
    • G01J1/42Photometry, e.g. photographic exposure meter using electric radiation detectors
    • G01J1/44Electric circuits
    • G01J2001/4413Type
    • G01J2001/442Single-photon detection or photon counting

Definitions

  • the invention relates to the technical field of extremely weak light detection, in particular to a single photon counting imaging system and a method thereof, which adopts a compression sensing theory and a DLP technology, and can realize a high quality two-dimensional image of a very weak light object by using a point detector.
  • Imaging. Background technique
  • a general imaging apparatus obtains an image by recording the light intensity and position of a point on the observation object.
  • the light intensity of the observed object is attenuated to a certain extent and reaches the single photon level, it becomes a discrete pulse signal.
  • a single photon is a very weak light that is considered to be the smallest unit of energy that is indivisible to light and is the limit that can be detected.
  • Single photon detection technology plays an important role in the fields of bio-luminescence, medical diagnosis, non-destructive substance analysis, astronomical observation, spectrometry, and quantum optics. It is very meaningful to study the application of extremely low light imaging detection technology in these fields.
  • Photon counting imaging is a kind of extremely weak light detection technology. Usually, it records and accumulates the photon count at the imaging position and accumulates the photon at the data processing end to obtain an image.
  • the core is the surface detector.
  • the size of the detector (array size), the sensitivity range, and the response band directly affect the quality of image acquisition at a single photon level.
  • the bin detector used for single photon detection level is not only expensive, but can be realized in a few bands, and the sensitivity of the bin detector is low, and there is a strong demand for two-dimensional imaging with less mature technology and extremely weak light objects. Contradictions between.
  • CS theory Compressed sensing theory
  • the CS theory consists of two parts: compressed sampling and sparse reconstruction.
  • Compressed sampling is the process of mapping a measured signal from a high dimension to a low dimension.
  • the hypothesis is the measured data, ⁇ is the observed data, eR ix "is the random projection matrix ( ⁇ "), and ee ⁇ is the measurement noise.
  • the compressed sampling process can be described as (1):
  • the random projection matrix ⁇ also called the measurement matrix, needs to satisfy RIP (Restricted Isometry Property):
  • is not related to ⁇ , the smaller the number of measurements required for sampling, so ⁇ is generally designed as a random matrix.
  • the first term is the least squares constraint, denoted as ( ⁇ ; the second term is a constraint on sparsity; the sum of the two is the objective function, denoted as ⁇ .
  • DLP technology is a technology proposed by Texas Instruments (TI). It combines with digital video or graphic signals. Its micromirror and lens system can reflect digital images onto a screen or other surface.
  • the core of the DLP chip is the DLP chip.
  • Digital Micro-mirror Device (DMD Control System), which is currently the world's most sophisticated optical switch. It contains a matrix of up to 2 million micromirrors mounted on hinges. Each micromirror is less than one-fifth the size of a human hair, and each micromirror can be within a certain angular range (usually -12 ° and +12°) wobble.
  • the pulse width modulated wave (PWM) is used to drive the micromirror to make high-speed jitter between 0 and 1, which can be achieved in the middle. status.
  • PWM pulse width modulated wave
  • the DMD control system and its associated sophisticated electronic components are the so-called DLP technology, which has mature products and is widely used in projection instruments and other products.
  • the object of the present invention is to provide a single photon counting imaging system and method thereof, in order to solve the contradiction between the low sensitivity of the panel detector, the immature technology and the strong demand for two-dimensional imaging of extremely weak light objects.
  • DLP technology is used to randomly change image signals into random light intensity signals, and then a single photon counter is used as a detecting component to obtain a counting signal.
  • the present invention provides a single photon counting imaging system, wherein the single photon counting imaging system adopts a compression sensing theory and a DLP technique, and uses a single photon counter as a detecting component to realize a single photon.
  • Two-dimensional imaging of a level of very weak light object the single photon counting imaging system comprising: a filter, a first lens 1, a DMD control system, a second lens 2, a single photon counter, and a data processing unit;
  • the DMD combines the first lens 1 and the second lens 2 for converting the two-dimensional image data into a one-dimensional data sequence to complete the compressed sampling of the measured signal, and the extremely weak light filtering the stray light through the filter, through the first lens 1 imaging at the DMD control system, and controlling the probability that photons are reflected to the second lens 2 by the DMD control system, controlling photon focusing via the second lens 2;
  • the data processing unit combines the single photon counter to perform the sparse reconstruction, and the data processing unit converts the photon according to the probability that the photon is detected into a measured value according to the single photon counter for a certain period of time, and the measurement matrix of the DMD control system is optimized. Reconstruct the photon density image and solve the 2D image.
  • the second lens 2 is focused to an optical path of the single photon counter and is provided with an optical attenuator for attenuating the light to the working range of the single photon detector.
  • the optical attenuator is designed to prevent saturation of the measured photon density and excessive gate time of the single photon counter.
  • the present invention also provides a single photon counting imaging method, which adopts a compression sensing theory and a DLP technique, and uses a single photon counter as a detecting component to realize a single photon level.
  • Two-dimensional imaging of extremely weak light objects the steps of which include:
  • the compressed sampling is performed by the DMD control system combining the first lens 1 and the second lens 2 for converting the two-dimensional image data into a one-dimensional data sequence to complete the compressed sampling of the measured signal.
  • the very weak light filters out the stray light through the filter is imaged by the first lens 1 at the DMD control system, and is controlled by the DMD control system to control the probability that the photon is reflected to the second lens 2, and the photon is controlled by the second lens 2;
  • the sparse reconstruction is performed by the data processing unit according to the probability that the photon is counted into a photon number according to a single photon counter for a certain period of time, and the measurement matrix on the DMD control system is used to reconstruct the photon density image by an optimization algorithm. , solve the two-dimensional image.
  • the method includes the following steps:
  • the compressed sampling is a process in which the measured signal is mapped from a high dimension to a low dimension:
  • is a wavelet transform matrix and ⁇ is a Gaussian random matrix
  • the inner product value of the array corresponding to an elemental expression of the observation vector y in (1), ", ⁇ ' are the 'th' elements of ⁇ ' and respectively; change the DMD control system according to the measurement matrix, repeat the measurement, The entire observation data y can be obtained;
  • the method attenuates the light to the working range of the single photon detector by the optical attenuator after the second lens 2 is focused to the optical path of the single photon counter.
  • the invention has the advantages that: the invention adopts the Compressive Sensing (CS) and the Digital Light Processing (DLP) technology to solve the imaging problem of using the point detector to achieve high detection sensitivity. Its sensitivity can reach single photon level, the resolution is directly related to the DMD control system, and the DMD control system can achieve high resolution.
  • the invention can be widely applied in the fields of biological self-luminescence detection, medical diagnosis, non-destructive substance analysis, astronomical observation, defense military, spectrometry, quantum electronics and the like.
  • the invention is based on the Compressive Sensing (CS) theory, adopts a single photon point detector as a detecting component, and realizes two-dimensional imaging of extremely weak light by a single photon counter, and has a simple structure and a sensitivity to a single photon level.
  • the resolution is directly related to the DMD control system, and the DMD control system can achieve high resolution at present, which solves the current low sensitivity of the focal plane sensor in the field, the array size is small, the detection wavelength range is relatively single and the extremely weak light object is two-dimensional.
  • FIG. 1 is a schematic view showing the structure of a single photon counting imaging system of the present invention.
  • Figure 2 is a simulation experiment result of the present invention
  • Figure 2 (a) is the original photon density image
  • Figure 2 (b) is a random matrix on the DMD control system in one measurement, black dots represent 0, white dots represent 1, gray dots Represents the intermediate value
  • Figure 2 (d) is the residual image of the IWT algorithm.
  • the extremely weak light emitted by the observation object is filtered by the filter to remove the stray light, and is imaged by the first lens 1 at the DMD control system, and the DMD control system controls the probability that the photon is reflected to the second lens 2, Passing the second lens 2 controls the photon convergence point.
  • the function of the optical attenuator is to attenuate the light to the working range of the single photon detector when the light is too strong.
  • the photon is counted by the single photon counter for a certain period of time, and the value can be converted into the detected photon. The probability of the number is taken as the measured value.
  • the photon density image is reconstructed by the data processing unit according to the measured value and the measurement matrix on the DMD control system through an optimization algorithm.
  • the optical attenuator is designed to prevent saturation of the measured photon density and the single gate time of the single photon counter.
  • the columns of the two-dimensional image are connected end to end, and are converted into "one-dimensional column vector of xl , where each element in (1) represents the photon density at the corresponding position; the DMD control system has the same resolution, its The columns are connected end to end and form a one-dimensional row vector of lx ", corresponding to a row in the measurement matrix ⁇ , wherein each element represents the probability of photons being transmitted to the second lens 2 at the corresponding position.
  • the probability p(r> is proportional to the intensity of light at a point in a meta area ⁇ at any time.
  • Biochips are typically very weak light sources and are currently easily observed by fluorescent labeling. In fact, organisms have self-luminous properties, and the self-luminous spectrum contains a lot of important information. It can be directly observed using photon counting imaging technology. In the experiment, a biochip image with a resolution of 64x64 was selected, and the gray scale was 256 levels, and the highest gray level corresponds to the number of photons being O x 102 - 1 . Under the assumption that the original image is not known, the Gaussian matrix is used for compression sampling, and the IWT sparse reconstruction algorithm is used for image reconstruction, and the results shown in Fig.

Landscapes

  • Physics & Mathematics (AREA)
  • General Physics & Mathematics (AREA)
  • Spectroscopy & Molecular Physics (AREA)
  • Investigating Or Analysing Materials By Optical Means (AREA)
  • Investigating, Analyzing Materials By Fluorescence Or Luminescence (AREA)
  • Photometry And Measurement Of Optical Pulse Characteristics (AREA)

Description

一种单光子计数成像系统及其方法
技术领域
本发明涉及极弱光探测的技术领域, 特别涉及一种单光子计数成像系统及其方法, 它采用压縮传感理论和 DLP技术, 利用点探测器可以实现极弱光对象的高质量二维成 像。 背景技术
普通成像设备通过记录观察对象上某点的光强和位置获得图像。当观察对象的光强 衰减到一定程度, 达到单光子水平, 就变成了离散脉冲信号。 单光子是一种极弱光, 被 认为是光不可分割的最小能量单位, 是可以探测的极限。单光子探测技术应用于生物自 发光、 医疗诊断、 非破坏性物质分析、 天文观测、 光谱测量、 量子光学等领域, 并在其 中扮演着重要角色。 研究极弱光成像探测技术在这些领域应用的发展非常有意义。
光子计数成像就是一种极弱光探测技术,通常它通过记录成像位置的光子计数以及 探测到光子的概率,在数据处理端进行累计和融合获得一幅图像,其核心是面元探测器, 面元探测器规模 (阵列大小)、 灵敏度范围、 以及响应波段直接影响能否获得单光子水 平的图象获取质量。但是, 用于单光子探测水平的面元探测器不但价格昂贵, 只能够在 少数波段可以实现, 且面元探测器灵敏度低,存在技术不甚成熟与极弱光对象二维成像 的强烈需求之间的矛盾。
压縮传感理论 (CS理论) 是由 E.J.Canc½S等人提出的, 它打破了传统的线性采样 模式, 表明可压縮信号的少量线性随机投影中包含足够的信息来重建原信号。
CS理论包括两部分: 压縮采样和稀疏重建。
压縮采样是被测信号由高维向低维映射的过程。假设 是被测数据, ^是 观测数据, eRix "是随机投影矩阵( <<"), e e ^是测量噪声, 那么, 压縮采样过 程可以描述为 (1) 式:
γ = χ + β (χ)
如果 是变换域稀疏的, 即 ·9 = Ψ Ψ是稀疏变换矩阵, 那么 (1) 式变化为 (2) 式:
y (2)
随机投影矩阵 Φ, 也叫测量矩阵, 需要满足 RIP (Restricted Isometry Property):
(l- ≤|K (i J χ 其中, 定义为使所有 s稀疏的向量; c都满足此不等式的最小常数, 并且 < 1。
另外, Φ与 Ψ越不相关, 采样所需的测量数 越小, 所以, 一般情况下 Φ设计为 随机矩阵。
稀疏重建实际上是在已知观测数据 y和测量矩阵 φ的条件下求解 a) 式中的 , 这是一个 ill-po 一般用最优化方法求解, 可描述为 (3) 式:
Figure imgf000004_0001
如果 是 题可以描述为 (4) 式:
Figure imgf000004_0002
(3)式和 (4)式中, 第一项是最小二乘约束, 记为 (^ ; 第二项是对 稀疏度的 一种约束; 两项之和是目标函数, 记为^ ^。
DLP技术是美国德州仪器公司(TI)提出的一项技术, 它与数字视频或图形信号结 合起来, 其微镜和透镜系统可以将数字图像反射到屏幕或其它表面, 其核心是 DLP芯 片——数字微镜器件 (Digital Micro-mirror Device, 简称 DMD控制系统), 这目前是世 界上最精密的光开关。 它包含一个多达 200 万个安装在铰链上的微镜的矩阵, 每个微 镜的大小小于人的头发丝的五分之一, 每一个微镜都可以在一定角度范围内 (通常 -12° 和 +12°)摆动, 如果把这两种状态记为 0和 1, 那么, 用脉宽调制波(PWM)来驱动微 镜, 使其在 0和 1之间高速抖动, 就可以实现中间状态。 DMD控制系统及其相关的精 密的电子元件就是所谓的 DLP技术,该技术有成熟产品,广泛用于投影仪器等产品中。
这种"先采样, 后重建"的思想使得将二维信号转换为随时间分布的一维信号, 并且 由单个探测器采样成为可能。而点探测器无论在探测灵敏度、波长范围具有更宽的选择 范围, 成本优势明显,利用点探测器实现单光子记数成像成为未来单光子水平成像的重 要发展趋势。 发明内容
本发明的目的在于, 为解决目前面元探测器灵敏度低、技术不甚成熟与极弱光对象 二维成像的强烈需求之间的矛盾, 从而提供一种单光子计数成像系统及其方法, 以压縮 传感 (Compressive Sensing, CS) 理论为基础, 采用 DLP技术将图象信号随机变化变 成随机的光强信号, 再以单光子计数器为探测元件获得记数信号, 实现了用点探测器对 极弱光对象进行二维成像。 为实现上述目的, 本发明提供了一种单光子计数成像系统, 其特征在于, 该单光子 计数成像系统采用压縮传感理论和 DLP技术, 并以单光子计数器为探测元件, 实现了 单光子级别的极弱光对象的二维成像, 所述的单光子计数成像系统包括: 滤光片、第一 透镜 1、 DMD控制系统、 第二透镜 2、 单光子计数器和数据处理单元; 其中,
DMD结合第一透镜 1和第二透镜 2, 用于将二维图像数据转化为一维数据序列完 成被测信号的压縮采样, 极弱光通过滤光片滤除杂光, 经第一透镜 1在 DMD控制系统 处成像, 并由 DMD控制系统控制光子被反射到第二透镜 2的概率, 经过第二透镜 2控 制光子聚焦;
数据处理单元结合单光子计数器完成稀疏重建,数据处理单元根据单光子计数器在 一定时间内对光子进行计数折算成探测到光子数的概率作为测量值,和 DMD控制系统 上的测量矩阵经过最优化算法重建光子密度图像, 解算出二维图像。 作为上述技术方案的一种改进,所述的第二透镜 2聚焦后至单光子计数器的光路上 还设有光衰减器, 用于将光衰减到单光子探测器的工作范围。该光衰减器的设计是为了 防止被测光子密度过大和单光子计数器的门控时间过长引起的饱和。 为实现上述的另一发明目的, 本发明还提供了一种单光子计数成像方法, 该方法采 用了压縮传感理论和 DLP技术, 并以单光子计数器为探测元件, 实现了单光子级别的 极弱光对象的二维成像, 其步骤包括:
1 ) 压縮采样的步骤;
所述的压縮采样是由 DMD控制系统结合第一透镜 1和第二透镜 2, 用于将二维图 像数据转化为一维数据序列完成被测信号的压縮采样,
极弱光通过滤光片滤除杂光, 经第一透镜 1在 DMD控制系统处成像, 并由 DMD 控制系统控制光子被反射到第二透镜 2的概率, 经过第二透镜 2控制光子聚焦;
2) 稀疏重建的步骤;
所述的稀疏重建是由数据处理单元根据单光子计数器在一定时间内对光子进行计 数折算成探测到光子数的概率作为测量值,和 DMD控制系统上的测量矩阵经过最优化 算法重建光子密度图像, 解算出二维图像。 作为上述技术方案的一种改进, 所述的方法包括步骤如下:
所述的压縮采样, 是被测信号由高维向低维映射的过程:
11 ) 假设 是被测数据, y e ^是观测数据, e Rix "是随机投影矩阵
( k « n -) > 是测量噪声, 那么, 压縮采样的过程可以描述为 (1 ) 式: 如果 是变换域稀疏的, 即 >9 = Ψ , Ψ是稀疏变换矩阵, 那么 (1) 式变化为 (2) 式:
y = 1 19 + e (2)
式中, Ψ为小波变换矩阵, Φ为 Gaussian随机矩阵;
12) 假设测量数为 , 二维图像的像素个数为 n, 则 (1) 式中的测量矩阵则为
Φ = {Φ1,···,Φ!,···,Φ,}; ^是0>的第 行, 把 ^x ^的二维图像的列首尾相连, 化成
"xl的一维列向量, 对应 (1) 式中的 , 其中的每一个元素代表相应位置处的光子密 度; DMD控制系统具有同样的分辨率, 它的列首尾相连, 化成 lx"的一维行向量, 对 应测量矩阵 Φ中的一行, 其中的每个元素代表相应位置处光子透射到第二透镜 2 的概 率;
13) 假设测量周期为: Γ, 在这段时间内, DMD控制系统保持不变, 单光子计数器 探测到的光子数为 W, 那么 就相当于光子密度图像与 DMD 系统上的随机数
Figure imgf000006_0001
阵列的内积值, 对应于(1)式中观察向量 y的一个元素 式中, "、 ^'分 别是 Φ '和 的第 '个元素; 根据测量矩阵改变 DMD控制系统, 重复 次测量, 就可以 得到整个观测数据 y ;
14)所述的稀疏重建是在已知观测数据 y和测量矩阵 Φ的条件下求解( 1 )式中的 一般用最优化方法求解, 可描述为 (3) 式:
. ( 1
mini
2 - - (3)
如果 是变换域稀疏的, 对应于 (2) 式的重建问题可以描述为 (4) 式:
1
mini
(3)式和 (4)式中, 第一项是最小二乘约束, 记为 ^(^; 第二项是对 稀疏度的 一种约束; 两项之和是目标函数, 记为 ^^)。 作为上述技术方案的进一步的改进, 所述的最优化方法是采用 IWT (Iterative Wavelets Thresholding)算法, 把本次迭代的估计值做 DWT变换, 对变换系数做阈值处 理, 再做 DWT反变换得到下一次迭代的估计值; 如果阈值处理函数描述为1 ^'v)=^Omax^l_v'Q} , 那么该算法可以描述为:
4
Figure imgf000007_0001
其中, 《= α^Φ 。
作为上述技术方案的另一种改进,所述的方法在所述的第二透镜 2聚焦后至单光子 计数器的光路上, 还通过光衰减器将光衰减到单光子探测器的工作范围。 本发明的优点在于: 本发明采用压縮传感理论(Compressive Sensing, 简称 CS )结 合数字光处理 (Digital Light Processing, 简称 DLP) 技术的方案来解决利用点探测器实 现高探测灵敏度的成像问题, 其灵敏度可以达到单光子水平, 分辨率与 DMD控制系统 直接相关, 而 DMD控制系统目前可以达到很高的分辨率。 本发明可广泛应用于生物自 发光检测、 医疗诊断、 非破坏性物质分析、 天文观测、 国防军事、 光谱测量、 量子电子 学等领域。
本发明以压縮传感(Compressive Sensing, CS )理论为基础, 以单光子点探测器为 探测元件, 用单光子计数器实现极弱光的二维成像, 结构简单, 灵敏度可以达到单光子 水平, 分辨率与 DMD控制系统直接相关, 而 DMD控制系统目前可以达到很高的分辨 率, 解决了目前该领域中焦平面传感器灵敏度低、 阵列规模小、探测波长范围相对单一 与极弱光对象二维成像的强烈需求之间的矛盾。
附图说明
图 1是本发明的单光子计数成像系统的结构示意图。
图 2是对本发明的模拟实验结果, 图 2 ( a) 是原始光子密度图像; 图 2 (b ) 是一 次测量中 DMD控制系统上的随机矩阵, 黑点代表 0, 白点代表 1, 灰点代表中间 值; 图 2 ( c) 是 IWT算法的重建图像, 与原始图像的相关系数 Cov = 0.9783, 信 噪比 PSNR=23.95dB ; 图 2 ( d) 是 IWT算法的残差图像。
具体实施方式
以下结合附图对本发明作进一步的详细说明。
如图 1所示, 观察对象发出的极弱光被滤光片滤除杂光, 经第一透镜 1在 DMD控 制系统处成像, 由 DMD控制系统控制光子被反射到第二透镜 2的概率, 经过第二透镜 2的控制光子会聚点,光衰减器作用是当光过强时将光衰减到单光子探测器的工作范围, 由单光子计数器在一定时间内对光子进行计数,该数值可以折算成探测到光子数的概率 作为测量值。最后由数据处理单元根据测量值和 DMD控制系统上的测量矩阵经过最优 化算法重建光子密度图像。其中,光衰减器的设计是为了防止被测光子密度过大和单光 子计数器的门控时间过长引起的饱和。
为了便于数学上的理解, 假设测量数为 , ( 1 ) 式中的测量矩阵写为
Φ = {Φ1 , · · · , Φ! , · · ·, Φ, } ; ^是 的第 行。 我们把
Figure imgf000008_0001
^的二维图像的列首尾相连, 化成" x l的一维列向量, 对应 (1 ) 式中的 其中的每一个元素代表相应位置处的光 子密度; DMD控制系统具有同样的分辨率, 它的列首尾相连, 化成 l x "的一维行向量, 对应测量矩阵 Φ中的一行,其中的每个元素代表相应位置处光子透射到第二透镜 2的概 率。 假设测量周期为 , 在这段时间内, DMD控制系统保持不变, 单光子计数器探测 到的光子数为 W, 那么 就相当于光子密度图像与 DMD控制系统上的随机数阵列 的内积值, 对应于 (1 ) 式中观察向量 y的一个元素 ^ ' ( φ"、 ^'分别是 φ'' 和 的第 ^ '个元素)。 根据测量矩阵改变 DMD控制系统, 重复 次测量, 就可以得到整 个观测数据 , 在物理上实现 (1 ) 式的过程。
根据光子学的知识, 在一个元面积 ^内, 任意时刻在 点观察到一个光子的概率 p(r> 正比于该处光强。 我们用生物芯片的灰度图像模拟光子密度图像。
生物芯片是典型的极弱光源, 目前主要通过荧光标记的方法使其便于观察。实际上 生物都有自发光的特性, 并且自发光光谱包含很多重要的信息。采用光子计数成像技术 就可以直接观测。 实验中选择分辨率为 64x64的生物芯片图像, 灰度为 256级, 最高灰 度级对应光子数为 O x 1021。在假设不知道原图像的情况下,采用 Gaussian矩阵进行 压縮采样, IWT稀疏重建算法进行图像重建, 得到图 2所示的结果, 其中, (a)是原始 光子密度图像; (b) 是一次测量中 DMD控制系统上的随机矩阵, 黑点代表 0, 白点代 表 1, 灰点代表中间值; (c) 是 IWT算法的重建图像; (d) 是 IWT算法的残差图像。
图 2 (c) 所示的重建图像与原始图像图 2 (a) 的相关系数 Cov = 0.9783, 信噪比 PSNR=23.95dB, 满足对生物芯片图像的观测需求。 实验结果表明, 本发明硬件可以实 现压縮采样, 软件算法可以实现图像重建。
最后所应说明的是, 以上实施例仅用以说明本发明的技术方案而非限制。尽管参照 实施例对本发明进行了详细说明, 本领域的普通技术人员应当理解,对本发明的技术方 案进行修改或者等同替换, 都不脱离本发明技术方案的精神和范围,其均应涵盖在本发 明的权利要求范围当中。

Claims

权 利 要 求
1、 一种单光子计数成像系统, 其特征在于, 该单光子计数成像系统采用压縮传感 理论和 DLP技术, 并以单光子计数器为探测元件, 实现了单光子级别的极弱光对象的 二维成像, 所述的单光子计数成像系统包括: 滤光片、 第一透镜(1 )、 DMD控制系统、 第二透镜 (2)、 单光子计数器和数据处理单元; 其中,
DMD结合第一透镜(1 )和第二透镜(2), 用于将二维图像数据转化为一维数据序 列完成被测信号的压縮采样, 极弱光通过滤光片滤除杂光, 经第一透镜 (1 ) 在 DMD 控制系统处成像, 并由 DMD控制系统控制光子被反射到第二透镜 (2) 的概率, 经过 第二透镜 (2) 控制光子聚焦;
数据处理单元结合单光子计数器完成稀疏重建,数据处理单元根据单光子计数器在 一定时间内对光子进行计数折算成探测到光子数的概率作为测量值,和 DMD控制系统 上的测量矩阵经过最优化算法重建光子密度图像, 解算出二维图像。
2、根据权利要求 1所述的单光子计数成像系统,其特征在于,所述的第二透镜(2) 聚焦后至单光子计数器的光路上还设有光衰减器,用于将光衰减到单光子探测器的工作 范围。
3、 一种单光子计数成像方法, 该方法采用了压縮传感理论和 DLP技术, 并以单光 子计数器为探测元件, 实现了单光子级别的极弱光对象的二维成像, 其步骤包括:
1 ) 压縮采样的步骤;
所述的压縮采样是由 DMD控制系统结合第一透镜 (1 ) 和第二透镜 (2), 用于将 二维图像数据转化为一维数据序列完成被测信号的压縮采样,
极弱光通过滤光片滤除杂光,经第一透镜(1 )在 DMD控制系统处成像,并由 DMD 控制系统控制光子被反射到第二透镜 (2) 的概率, 经过第二透镜 (2) 控制光子聚焦;
2) 稀疏重建的步骤;
所述的稀疏重建是由数据处理单元根据单光子计数器在一定时间内对光子进行计 数折算成探测到光子数的概率作为测量值,和 DMD控制系统上的测量矩阵经过最优化 算法重建光子密度图像, 解算出二维图像。
4、 根据权利要求 3所述的单光子计数成像方法, 其特征在于, 所述的方法包括步 骤如下: 所述的压縮采样, 是被测信号由高维向低维映射的过程:
H ) 假设 e T是被测数据, 是观测数据, 0)eRix"是随机投影矩阵 (k«n), ee ^是测量噪声, 那么, 压縮采样的过程可以描述为 (1) 式: 如果 是变换域稀疏的, 即 >9 = Ψ , Ψ是稀疏变换矩阵, 那么 (1) 式变化为 (2) 式:
y = 1 19 + e (2)
式中, Ψ为小波变换矩阵, Φ为 Gaussian随机矩阵;
12) 假设测量数为 , 二维图像的像素个数为 n, 则 (1) 式中的测量矩阵则为
Φ = {Φ1,···,Φ!,···,Φ,}) ^是 的第 行, 把 ^x ^的二维图像的列首尾相连, 化成
"xl的一维列向量, 对应 (1) 式中的 , 其中的每一个元素代表相应位置处的光子密 度; DMD控制系统具有同样的分辨率, 它的列首尾相连, 化成 lx"的一维行向量, 对 应测量矩阵 Φ中的一行, 其中的每个元素代表相应位置处光子透射到第二透镜 (2) 的 概率;
13) 假设测量周期为: Γ, 在这段时间内, DMD控制系统保持不变, 单光子计数器 探测到的光子数为 W, 那么 就相当于光子密度图像与 DMD 系统上的随机数 阵列的内积值, 对应于(1)式中观察向量 y的一个元素
Figure imgf000011_0001
式中, "、 ^'分 别是 Φ '和 的第 '个元素; 根据测量矩阵改变 DMD控制系统, 重复 次测量, 就可以 得到整个观测数据 y;
14)所述的稀疏重建是在已知观测数据 y和测量矩阵 Φ的条件下求解( 1 )式中的 一般用最优化方法求解, 可
Figure imgf000011_0002
如果 是变换域稀疏的, 2) 式的重建问题可以描述为 (4) 式:
Figure imgf000011_0003
(3)式和 (4)式中, 第一项是最小二乘约束, 记为 ^(^; 第二项是对 稀疏度的 一种约束; 两项之和是目标函数, 记为^ ^)。
5、 根据权利要求 4所述的单光子计数成像方法, 其特征在于, 所述的最优化方法 是采用 IWT算法, 把本次迭代的估计值做 DWT变换, 对变换系数做阈值处理, 再做 DWT反变换得到下一次迭代的估计值; 如果阈值处理函数描述为1 ^'v)=^"(")max^l_v'Q} ,那么,该算法可以描述为:
= ψ- 1. s _丄 ν/( )),丄
a ) a
其中, α
6、 根据权利要求 3或 4所述的单光子计数成像方法, 其特征在于, 所述的方法在 所述的第二透镜 (2) 聚焦后至单光子计数器的光路上, 还通过光衰减器将光衰减到单 光子探测器的工作范围。
10
PCT/CN2012/074533 2011-04-25 2012-04-23 一种单光子计数成像系统及其方法 WO2012146156A1 (zh)

Priority Applications (3)

Application Number Priority Date Filing Date Title
US14/000,421 US8723130B2 (en) 2011-04-25 2012-04-23 Single photon-counting imaging system and method thereof
JP2014506741A JP6211512B2 (ja) 2011-04-25 2012-04-23 単一光子計数イメージングシステム及び方法
EP12776896.8A EP2685227B1 (en) 2011-04-25 2012-04-23 Single photon-counting imaging system and method thereof

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
CN201110103559.3 2011-04-25
CN201110103559.3A CN102759408B (zh) 2011-04-25 2011-04-25 一种单光子计数成像系统及其方法

Publications (1)

Publication Number Publication Date
WO2012146156A1 true WO2012146156A1 (zh) 2012-11-01

Family

ID=47053946

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2012/074533 WO2012146156A1 (zh) 2011-04-25 2012-04-23 一种单光子计数成像系统及其方法

Country Status (5)

Country Link
US (1) US8723130B2 (zh)
EP (1) EP2685227B1 (zh)
JP (1) JP6211512B2 (zh)
CN (1) CN102759408B (zh)
WO (1) WO2012146156A1 (zh)

Cited By (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN112957011A (zh) * 2021-02-01 2021-06-15 西安电子科技大学 高灵敏度微弱荧光信号探测系统、方法、存储介质及应用
CN114486746A (zh) * 2021-11-25 2022-05-13 中国科学院西安光学精密机械研究所 基于压缩感知的高分辨率光子集成成像系统及成像方法
CN115442505A (zh) * 2022-08-30 2022-12-06 山西大学 一种单光子压缩感知成像系统及其方法

Families Citing this family (25)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
EP2755327A4 (en) 2011-10-25 2015-11-04 Ct Space Sci & Applied Res Cas MULTI-DIMENSIONAL MULTI-DIMENSIONAL IMAGING SYSTEM AND METHOD WITH TEMPORARY RESOLUTION OR ULTRA-LIGHT LIGHT
CN103207015A (zh) * 2013-04-16 2013-07-17 华东师范大学 一种光谱重构方法及其光谱仪装置
CN103645491B (zh) * 2013-11-25 2016-12-07 中国科学院高能物理研究所 放射源定位方法、装置及系统
JP6775421B2 (ja) * 2014-03-20 2020-10-28 ケーエルエー コーポレイション 照明パターン形成を用いた計測光学系及び方法
CN103969693A (zh) * 2014-04-30 2014-08-06 中国科学院长春光学精密机械与物理研究所 光电探测成像系统及其成像方法
CN104238996B (zh) * 2014-09-04 2017-08-11 清华大学 源无关量子随机数的产生方法及装置
CN104375158B (zh) * 2014-09-10 2017-09-05 北京交通大学 一种单原子的探测方法及装置
JP6289339B2 (ja) * 2014-10-28 2018-03-07 株式会社日立ハイテクノロジーズ 荷電粒子線装置及び情報処理装置
CN104796674A (zh) * 2015-04-17 2015-07-22 南京理工大学 基于压缩感知的彩色成像装置及方法
TWI546533B (zh) * 2015-07-14 2016-08-21 龍彩科技股份有限公司 即時空間與時間光譜量測系統及其量測模組
CN105915868A (zh) * 2016-04-22 2016-08-31 南京理工大学 基于扩展小波树的彩色成像系统及方法
CN105915869A (zh) * 2016-04-22 2016-08-31 南京理工大学 一种彩色自适应压缩计算鬼成像系统及方法
US10520434B2 (en) * 2016-05-25 2019-12-31 Leica Microsystems Cms Gmbh Fluorescence lifetime imaging microscopy method having time-correlated single-photon counting, which method permits higher light intensities
CN106054038B (zh) * 2016-06-29 2019-01-15 南京理工大学 一种连通域标记的紫外单光子计数的方法
CN106772430B (zh) * 2016-12-30 2019-05-07 南京理工大学 基于多分辨率小波逼近的单像素光子计数三维成像系统及方法
CN106911893B (zh) * 2017-02-23 2020-04-03 北京建筑大学 一种单像素计算成像方法
CN109361833B (zh) * 2018-10-08 2020-08-11 南昌大学 一种单光子压缩视频传输装置的传输方法
CN109816603B (zh) * 2018-12-30 2023-04-11 天津大学 单光子计数成像的图像传感器图像还原方法
CN110187498B (zh) * 2019-05-27 2021-08-17 中国科学院国家空间科学中心 一种真热光关联成像系统
CN110443883B (zh) * 2019-07-08 2023-04-07 杭州电子科技大学 一种基于dropblock的单张彩色图片平面三维重建方法
CN110852981B (zh) * 2019-11-19 2023-08-22 中智科仪(北京)科技有限公司 一种单光子成像图像处理方法
CN111103062B (zh) * 2019-12-06 2021-02-12 太原理工大学 一种基于单光子计数二维成像的装置及方法
CN111208531A (zh) * 2020-01-19 2020-05-29 中国科学技术大学 一种基于宽谱光源的单光子成像激光雷达系统
CN111896125B (zh) * 2020-07-09 2021-05-18 武汉大学 一种单光子计数成像的偏振去噪方法
CN113534094B (zh) * 2021-05-28 2023-04-18 西安电子科技大学 一种基于纠缠态的量子探测恒虚警检测系统及检测方法

Citations (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US6700108B2 (en) * 2001-11-01 2004-03-02 Northrop Grumman Corporation Means for protecting optical focal plane sensor arrays against excessive irradiation
US20060239336A1 (en) * 2005-04-21 2006-10-26 Baraniuk Richard G Method and Apparatus for Compressive Imaging Device
CN1961568A (zh) * 2004-06-01 2007-05-09 帝国学院创新有限公司 成像系统

Family Cites Families (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
WO2002091147A1 (en) * 2001-05-09 2002-11-14 Magiq Technologies, Inc. Efficient use of detectors for random number generation
JP2005250714A (ja) * 2004-03-03 2005-09-15 Univ Nihon 光子乱数発生器
JP2011075491A (ja) * 2009-10-01 2011-04-14 Panasonic Corp 光量測定装置

Patent Citations (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US6700108B2 (en) * 2001-11-01 2004-03-02 Northrop Grumman Corporation Means for protecting optical focal plane sensor arrays against excessive irradiation
CN1961568A (zh) * 2004-06-01 2007-05-09 帝国学院创新有限公司 成像系统
US20060239336A1 (en) * 2005-04-21 2006-10-26 Baraniuk Richard G Method and Apparatus for Compressive Imaging Device

Non-Patent Citations (5)

* Cited by examiner, † Cited by third party
Title
CANDES, E. ET AL.: "An introduction to compressive sampling", IEEE SIGNAL PROCESSING MAGZINE, vol. 25, no. 2, March 2008 (2008-03-01), pages 21 - 30, XP002589050 *
DUARTE, M. F. ET AL.: "Single-Pixel Imaging via Compressive Sampling", IEEE SIGNAL PROCESSING MAGAZINE, vol. 25, no. 2, March 2008 (2008-03-01), pages 83 - 91, XP002632914 *
FIGUEIREDO, M. ET AL.: "Gradient projection for sparse reconstruction: Application to compressed sensing and other inverse problems", IEEE JOURNAL OF SELECTED TOPICS IN SIGNAL PROCESSING, vol. 1, no. 4, December 2007 (2007-12-01), pages 586 - 597, XP011199163 *
MAGALHAES, F. ET AL.: "Active illumination single-pixel camera based on compressive sensing", APPLIED OPTICS, vol. 50, no. 4, 1 February 2011 (2011-02-01), pages 405 - 414, XP001560283 *
WU, QINGLIN ET AL.: "Single photon detection technology", PROGRESS IN PHYSICS, vol. 30, no. 3, September 2010 (2010-09-01), pages 296 - 306 *

Cited By (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN112957011A (zh) * 2021-02-01 2021-06-15 西安电子科技大学 高灵敏度微弱荧光信号探测系统、方法、存储介质及应用
CN112957011B (zh) * 2021-02-01 2022-06-24 西安电子科技大学 高灵敏度微弱荧光信号探测系统、方法、存储介质及应用
CN114486746A (zh) * 2021-11-25 2022-05-13 中国科学院西安光学精密机械研究所 基于压缩感知的高分辨率光子集成成像系统及成像方法
CN114486746B (zh) * 2021-11-25 2023-12-08 中国科学院西安光学精密机械研究所 基于压缩感知的高分辨率光子集成成像系统及成像方法
CN115442505A (zh) * 2022-08-30 2022-12-06 山西大学 一种单光子压缩感知成像系统及其方法
CN115442505B (zh) * 2022-08-30 2023-07-21 山西大学 一种单光子压缩感知成像系统及其方法

Also Published As

Publication number Publication date
CN102759408A (zh) 2012-10-31
EP2685227A1 (en) 2014-01-15
US8723130B2 (en) 2014-05-13
EP2685227B1 (en) 2018-10-10
EP2685227A4 (en) 2014-10-01
JP2014512540A (ja) 2014-05-22
CN102759408B (zh) 2015-04-15
JP6211512B2 (ja) 2017-10-11
US20130341487A1 (en) 2013-12-26

Similar Documents

Publication Publication Date Title
WO2012146156A1 (zh) 一种单光子计数成像系统及其方法
JP6952176B2 (ja) 時間的圧縮感知システム
AU2019208156B2 (en) Comprehensive fixed pattern noise cancellation
WO2012174940A1 (zh) 一种极弱光多光谱成像方法及其系统
US11187657B2 (en) Hyperspectral imaging with fixed pattern noise cancellation
US20110109773A1 (en) System and method for adaptive nonlinear compressed visual sensing
US20170102671A1 (en) Holographic light field imaging device and method of using the same
JP6975897B2 (ja) 画像生成装置及び撮像装置
Yu et al. Single-photon compressive imaging with some performance benefits over raster scanning
US11561134B2 (en) Compressed-sensing ultrafast spectral photography systems and methods
US20220100094A1 (en) Quantum-limited Extreme Ultraviolet Coherent Diffraction Imaging
WO2020069193A1 (en) Apparatus and methods for all-cmos compressive sensing
Mochizuki et al. Separation of multi-path components in sweep-less time-of-flight depth imaging with a temporally-compressive multi-aperture image sensor
Wang et al. Iterative adaptive photon-counting compressive imaging based on wavelet entropy automatic threshold acquisition
EP4198876A1 (en) Reducing image artefacts in electron microscopy
Saragadam et al. Wavelet tree parsing with freeform lensing
Joany et al. Design and analysis of single pixel imaging on magnetic resonance imaging
Alsolami et al. Imaging with SPADs and DMDs: Seeing through diffraction-photons
Wu Compressive optical imaging systems
WO2023111772A1 (en) Reducing image artefacts in electron microscopy
WO2022094695A1 (en) A method and a system for compressed ultrafast tomographic imaging
Oakham et al. The optimal optical readout for the x‐ray light valve—Document scanners
EP4367500A1 (en) Temporal super-resolution
Huang et al. Enhancing Compressive Single-Pixel Imaging with Zig-Zag-Ordered Walsh-Hadamard Light Modulation
Hardie et al. Impact of Detector-element Active-area Shape and Fill Factor on Image Sampling, Restoration, and Super-Resolution

Legal Events

Date Code Title Description
121 Ep: the epo has been informed by wipo that ep was designated in this application

Ref document number: 12776896

Country of ref document: EP

Kind code of ref document: A1

WWE Wipo information: entry into national phase

Ref document number: 2012776896

Country of ref document: EP

WWE Wipo information: entry into national phase

Ref document number: 14000421

Country of ref document: US

ENP Entry into the national phase

Ref document number: 2014506741

Country of ref document: JP

Kind code of ref document: A

NENP Non-entry into the national phase

Ref country code: DE