CN107205103B - Ultra-high speed compression photographic device based on compressed sensing and stripe camera principle - Google Patents

Abstract

Translated fromChinese

本发明提供了一种基于压缩感知和条纹相机原理的超高速压缩摄影装置，该装置包括：透镜、分束器、空间光调制器、平面镜、内置CCD的条纹相机及解码器；首先通过空间光调制器对捕获图像信息进行编码，然后采用条纹相机对不同时刻图像信息进行叠加并压缩采样，最后基于压缩感知原理对图像进行重构。本装置能够在条纹相机成像速度一定的情况下对超快过程进行二维成像，将图像信息输出到解码器，通过TwIST算法高质量地重构出二维图像动态过程（x‑y‑t），其成像速度可以达到10^12帧/秒，是一种可以测量不重复超快事件的单次拍照测量技术。

The invention provides an ultra-high-speed compression photography device based on the principle of compressed sensing and streak camera. The device includes: a lens, a beam splitter, a spatial light modulator, a plane mirror, a streak camera with a built-in CCD, and a decoder; The modulator encodes the captured image information, then uses the streak camera to superimpose and compress the image information at different times, and finally reconstructs the image based on the principle of compressed sensing. The device can perform two-dimensional imaging of ultra-fast processes under the condition that the imaging speed of the streak camera is constant, output the image information to the decoder, and reconstruct the dynamic process of two-dimensional images with high quality through the TwIST algorithm (x‑y‑t) , its imaging speed can reach 10^12 frames/second, and it is a single-shot photogrammetry technology that can measure non-repetitive ultrafast events.

Description

Translated fromChinese

基于压缩感知和条纹相机原理的超高速压缩摄影装置Ultra-high-speed compression photography device based on compressed sensing and streak camera principles

技术领域technical field

本发明属于超快成像技术领域，涉及一种可用于纳秒和皮秒量级的超快物理、化学、生物等过程的二维图像测量技术，可用来测量时频域任意形状皮秒激光脉冲，观测微纳加工动态过程，分析强光场下分子结构变化动力学过程等。The invention belongs to the technical field of ultrafast imaging, and relates to a two-dimensional image measurement technology that can be used for nanosecond and picosecond ultrafast physical, chemical, biological and other processes, and can be used to measure picosecond laser pulses of any shape in the time-frequency domain , observe the dynamic process of micro-nano processing, analyze the dynamic process of molecular structure change under strong light field, etc.

背景技术Background technique

高速度捕获瞬态场景图像一直是摄影师们追求的梦想和目标，最典型的早期例子就是1878年记录奔跑中的马和1887年超声速子弹摄影。然而，直到二十世纪末，这种超高速成像仍然没有被突破，这时成像速度是每秒10^5帧。后来，随着基于电荷耦合器件(CCD)和互补金属氧化物半导体(CMOS)的电子成像传感器出现彻底改变了人们对高速成像的认识，使得成像速度可以达到每秒10^7帧。尽管这种传感器具有非常广泛的影响和应用，但是这种基于CCD和CMOS技术的成像速度受限于芯片存储和电子读出速度，因此不可能进一步提高成像速度，限制了很多相关领域的实际应用，比如测量光速或接近光速运动物体。High-speed capture of transient scene images has always been the dream and goal of photographers. The most typical early examples are the recording of running horses in 1878 and the supersonic bullet photography in 1887. However, until the end of the twentieth century, this ultra-high-speed imaging was still not broken, when the imaging speed was 10^5 frames per second. Later, with the advent of electronic imaging sensors based on charge coupled devices (CCDs) and complementary metal oxide semiconductors (CMOS), people's understanding of high-speed imaging has been completely changed, enabling imaging speeds to reach 10^7 frames per second. Although this sensor has a very broad impact and application, the imaging speed based on CCD and CMOS technology is limited by the chip storage and electronic readout speed, so it is impossible to further improve the imaging speed, which limits the practical application in many related fields , such as measuring the speed of light or moving objects close to the speed of light.

直到2014年，由美国圣路易斯华盛顿大学生物医学工程系Lihong Wang教授的研究团队发展了一种基于压缩传感理论和条纹相机相结合的技术，即第一代压缩超快成像(Compressed ultrafast photography,CUP)技术，这种CUP技术可以捕获不重复变化事件。条纹相机用空间坐标的形式记录了每个时刻发生的事件，它的灵敏度决定了CUP的性能。目前，该装置已经使得成像速度提高到每秒10^11帧，这是迄今为止世界上最快只接受式的摄像机。Until 2014, the research team of Professor Lihong Wang from the Department of Biomedical Engineering of Washington University in St. Louis developed a technology based on the combination of compressed sensing theory and streak cameras, namely the first generation of compressed ultrafast photography (CUP). ) technology, this CUP technology can capture non-repeated change events. The streak camera records events at each moment in the form of spatial coordinates, and its sensitivity determines the performance of the CUP. At present, the device has increased the imaging speed to 10^11 frames per second, which is by far the fastest reception-only camera in the world.

发明内容SUMMARY OF THE INVENTION

本发明的目的在于提供一种基于压缩感知和条纹相机原理的超高速压缩摄影装置，该装置可以高质量地重构出纳秒甚至皮秒量级的超快物理、化学、生物等过程的动态图像信息。The object of the present invention is to provide an ultra-high-speed compression photography device based on the principle of compressed sensing and streak camera, which can reconstruct dynamic images of ultra-fast physical, chemical, biological and other processes in nanoseconds or even picoseconds with high quality information.

实现本发明目的的具体技术方案是：The concrete technical scheme that realizes the object of the present invention is:

一种基于压缩感知和条纹相机原理的超高速压缩摄影装置，特点是该装置包括第一透镜、第二透镜、第三透镜、分束器、第一空间光调制器、第二空间光调制器、第一平面镜、第二平面镜、第三平面镜、第四平面镜、内置CCD的条纹相机及解码器，所述第一透镜与分束器光路连接；分束器一路与第一空间光调制器、第二透镜、第一平面镜依次光路连接；分束器另一路与第二空间光调制器、第三透镜、第二平面镜依次光路连接；第一平面镜与第三平面镜光路连接；第二平面镜与第四平面镜光路连接；第三平面镜、第四平面镜分别连接内置CCD的条纹相机；内置CCD的条纹相机连接解码器。An ultra-high-speed compression photography device based on compressed sensing and streak camera principles, characterized in that the device includes a first lens, a second lens, a third lens, a beam splitter, a first spatial light modulator, and a second spatial light modulator , a first plane mirror, a second plane mirror, a third plane mirror, a fourth plane mirror, a streak camera with built-in CCD and a decoder, the first lens is connected with the beam splitter optical path; the beam splitter is connected with the first spatial light modulator, The second lens and the first plane mirror are optically connected in sequence; the other path of the beam splitter is optically connected to the second spatial light modulator, the third lens, and the second plane mirror in sequence; the first plane mirror is optically connected to the third plane mirror; the second plane mirror is connected to the third plane mirror. The four plane mirrors are connected by optical paths; the third plane mirror and the fourth plane mirror are respectively connected to the streak camera with built-in CCD; the streak camera with built-in CCD is connected to the decoder.

所述分束器分束后的两条光路通过各光学器件的总光程相等。The total optical paths of the two optical paths after the beam splitter pass through the optical devices are equal.

所述分束器满足反射和透射的两路光辐射的能量相同，即50％的光被反射，50％的光被透射；分束器反射的一路光经由第一空间光调制器的调制后到达第二透镜、透射的一路光经由第二空间光调制器的调制后到达第三透镜。The beam splitter satisfies that the energy of the reflected and transmitted light radiations is the same, that is, 50% of the light is reflected, and 50% of the light is transmitted; the beam splitter reflects one light after being modulated by the first spatial light modulator. After reaching the second lens, the transmitted light reaches the third lens after being modulated by the second spatial light modulator.

所述第二空间光调制器放置在第一透镜的像平面上，第一空间光调制器根据反射光路的光程，放置在等效的像平面上；所述第一空间光调制器及第二空间光调制器对不同时刻图像信息进行编码。The second spatial light modulator is placed on the image plane of the first lens, and the first spatial light modulator is placed on an equivalent image plane according to the optical path of the reflected light path; the first spatial light modulator and the second spatial light modulator are placed on the equivalent image plane. Two spatial light modulators encode image information at different times.

所述第一平面镜、第三平面镜与第二平面镜、第四平面镜分别改变光路方向，将光反射进入内置CCD的条纹相机中。The first plane mirror, the third plane mirror, the second plane mirror, and the fourth plane mirror change the direction of the light path respectively, and reflect the light into the streak camera with built-in CCD.

所述内置CCD的条纹相机接收来自第三平面镜、第四平面镜反射的光，在使用中将条纹相机的狭缝打开到最大以在CCD相机获得一个二维时间偏移图像，即为单次曝光时间内的若干幅图像的叠加像。在应用中，CCD将九个像素(3*3)合并在一起使用以提高探测灵敏度。The streak camera with built-in CCD receives the light reflected from the third plane mirror and the fourth plane mirror, and in use, the slit of the streak camera is opened to the maximum to obtain a two-dimensional time offset image on the CCD camera, which is a single exposure. An overlay of several images over time. In application, CCD combines nine pixels (3*3) to improve detection sensitivity.

所述解码器存储CCD相机图像，对由条纹相机获得的时间偏移的二维图像进行三维解码重构，并且输出解码后的动态图像(x-y-t)。The decoder stores the CCD camera image, performs three-dimensional decoding and reconstruction on the time-shifted two-dimensional image obtained by the streak camera, and outputs the decoded dynamic image (x-y-t).

本发明通过空间光调制器对不同时刻图像信息进行编码，然后利用解码器对条纹相机中得到的二维时间偏移图像进行解码就可以重构出二维动态图像信息(x-y-t)。使用空间光调制器可以获得的光的辐射的信息更加多元化，而不再是由数字微镜器件(DigitalMicrowaveDevice，简称DMD)所产生的0-1二元化编码。The present invention encodes image information at different times through a spatial light modulator, and then uses a decoder to decode the two-dimensional time-shifted image obtained in the streak camera to reconstruct two-dimensional dynamic image information (x-y-t). The information of the radiation of light that can be obtained by using the spatial light modulator is more diversified, instead of the 0-1 binary code generated by a digital micro-mirror device (Digital Microwave Device, DMD for short).

本发明的优点是：The advantages of the present invention are:

(1)本发明使用Hamamatsu C7700条纹相机，成像速度可以达到每秒10^12帧，比普通的摄像装置提高了5个数量级；与现有CUP系统相比，也提高了一个数量级。(1) The present invention uses the Hamamatsu C7700 streak camera, and the imaging speed can reach 10^12 frames per second, which is 5 orders of magnitude higher than the ordinary camera device; compared with the existing CUP system, it is also improved by an order of magnitude.

(2)本发明可以实现空间时间三维信息(x-y-t)重构，可以直接观察超快过程的动态变化。(2) The present invention can realize the reconstruction of space-time three-dimensional information (x-y-t), and can directly observe the dynamic changes of the ultra-fast process.

(3)本发明是单次拍照测量，可以测量不重复或者不可逆发生事件。(3) The present invention is a single-shot measurement, which can measure non-repetitive or irreversible events.

(4)本发明通过空间光调制器对动态图像信息进行编码，可以产生高斯矩阵密码，比现有CUP系统的重构图像质量明显提升。(4) The present invention encodes the dynamic image information through the spatial light modulator, and can generate a Gaussian matrix cipher, which is significantly improved compared with the reconstructed image quality of the existing CUP system.

本装置具有接收式、多色再生、多帧成像和高像素的优点，可以一次测量像素为1000×1000×300(x-y-t)的不可逆超快发生事件，时间分辨率达到皮秒量级。此外，本发明中储存的数据是被压缩编码的二维数据，具有占内存小和高度安全的特点，具有很广阔的应用前景，比如大数据信息安全，卫星安全通讯等领域。The device has the advantages of receiving type, multi-color regeneration, multi-frame imaging and high pixel. It can measure irreversible ultrafast events with a pixel of 1000×1000×300 (x-y-t) at one time, and the time resolution reaches the order of picoseconds. In addition, the data stored in the present invention is compressed and encoded two-dimensional data, which has the characteristics of small memory occupation and high security, and has broad application prospects, such as big data information security, satellite security communication and other fields.

附图说明Description of drawings

图1为本发明结构示意图；Fig. 1 is the structural representation of the present invention;

图2为高斯光束先增强后减弱的过程图；Figure 2 is a process diagram of the Gaussian beam being enhanced first and then weakened;

图3为本发明模拟的高斯光束先增强后减弱的过程图；Fig. 3 is the process diagram that the Gaussian beam simulated by the present invention is enhanced first and then weakened;

图4为基于现有CUP系统模拟的高斯光束先增强后减弱的过程图。FIG. 4 is a process diagram of a Gaussian beam that is first enhanced and then weakened based on the simulation of the existing CUP system.

具体实施方式Detailed ways

以下结合附图和实施例进一步详细阐述本发明。The present invention will be described in further detail below with reference to the accompanying drawings and embodiments.

本发明包括第一透镜1、第二透镜2、第三透镜3、分束器4、第一空间光调制器5、第二空间光调制器6、第一平面镜7、第二平面镜8、第三平面镜9、第四平面镜10、内置CCD的条纹相机11及解码器12，所述第一透镜1与分束器4光路连接；分束器4一路与第二空间光调制器6、第三透镜3、第二平面镜8依次光路连接；分束器4另一路与第一空间光调制器5、第二透镜2、第一平面镜7依次光路连接；第二平面镜8与第四平面镜10光路连接；第一平面镜7与第三平面镜9光路连接；第四平面镜10、第三平面镜9分别连接内置CCD的条纹相机11；内置CCD的条纹相机11连接解码器12。The present invention includes afirst lens 1, asecond lens 2, athird lens 3, a beam splitter 4, a firstspatial light modulator 5, a second spatial light modulator 6, a first plane mirror 7, asecond plane mirror 8, a Threeplane mirrors 9, afourth plane mirror 10, afringe camera 11 with built-in CCD, and adecoder 12, thefirst lens 1 is optically connected to the beam splitter 4; the beam splitter 4 is connected to the second spatial light modulator 6, the third Thelens 3 and thesecond plane mirror 8 are connected in an optical path in sequence; the other path of the beam splitter 4 is connected with the firstspatial light modulator 5, thesecond lens 2 and the first plane mirror 7 in order in an optical path; thesecond plane mirror 8 is connected with thefourth plane mirror 10 in the optical path The first plane mirror 7 is connected to the optical path of thethird plane mirror 9; thefourth plane mirror 10 and thethird plane mirror 9 are respectively connected to thestreak camera 11 with built-in CCD;

本发明各器件的功能及要求：分束器4与第一透镜1相连，应该严格满足反射和透射的两路光辐射的能量相同，即50％的光被反射，50％的光被透射；第一、第二空间光调制器5、6(Spacial Light Modulator，简称SLM)，分别接收经由分束器4分束的两路光，对其进行空间强度调制后将光分别送入第二、第三透镜2、3，所述第一、第二空间光调制器5、6的位置要求，第二空间光调制器6应该直接放置在第一透镜1的像平面上，第一空间光调制器5也应该根据两条光路的光程必须严格相等，放置在等效的像平面上；第一、第三平面镜7、9和第二、第四平面镜8、10分别在两路光中起到改变光路方向的作用，最终将光反射进入内置CCD的条纹相机11中；内置CCD的条纹相机11接收来自第三、第四平面镜9、10反射的光，并在内置CCD相机成像；解码器12存储CCD相机图像，对时间偏移的二维图像进行三维解码重构，并且输出解码后的动态图像。The functions and requirements of each device of the present invention: the beam splitter 4 is connected with thefirst lens 1, and should strictly meet the energy of the reflected and transmitted two-path optical radiation, that is, 50% of the light is reflected and 50% of the light is transmitted; The first and secondspatial light modulators 5 and 6 (Spacial Light Modulator, SLM for short) respectively receive the two paths of light split by the beam splitter 4, perform spatial intensity modulation on them, and send the light into the second and second spatial light modulators, respectively. For thethird lenses 2 and 3, the positions of the first and secondspatial light modulators 5 and 6 require that the second spatial light modulator 6 should be placed directly on the image plane of thefirst lens 1. The first spatial light modulation Thedevice 5 should also be placed on the equivalent image plane according to the optical path of the two optical paths; In order to change the direction of the light path, the light is finally reflected into thestreak camera 11 with built-in CCD; thestreak camera 11 with built-in CCD receives the light reflected from the third andfourth plane mirrors 9 and 10, and images in the built-in CCD camera; thedecoder 12 stores the CCD camera image, performs three-dimensional decoding and reconstruction on the time-shifted two-dimensional image, and outputs the decoded dynamic image.

本发明中所有光学器件的位置都有要求，除了应满足同轴放置之外，还应该满足两路光线的光程应该相等。The positions of all optical devices in the present invention have requirements, in addition to coaxial placement, the optical paths of the two light rays should be equal.

本发明的工作过程分为正演过程和反演过程，正演过程为：高速动态场景先由第一透镜成像，然后到达分束器形成两路光路；两路光路一路经由第一空间光调制器进行编码后再进入第二透镜，另一路经由第二空间光调制器进行编码后进入第三透镜；第一平面镜和第三平面镜将第一条光路中从第一空间调制器出来的光信息通过反射使其进入条纹相机，第二平面镜和第四平面镜将另一条光路中从第二空间调制器出来的光信息进行方向改变也使其进入条纹相机；条纹相机对不同时刻进入的编码后的图景进行纵向位置移动，最后压缩成像在CCD相机的显示屏上。反演过程为：解码器接受来自CCD上的编码压缩后的二维数据图像并利用TwIST算法，反演重构出含有高速动态过程的三维数据(x-y-t)。The working process of the present invention is divided into a forward process and an inversion process. The forward process is as follows: the high-speed dynamic scene is first imaged by the first lens, and then reaches the beam splitter to form two optical paths; one of the two optical paths is modulated by the first spatial light. The second lens enters the second lens after being encoded by the second spatial light modulator, and the other path enters the third lens after being encoded by the second spatial light modulator; By reflection, it enters the streak camera, and the second plane mirror and the fourth plane mirror change the direction of the light information from the second spatial modulator in the other optical path to make it enter the streak camera; The picture is moved longitudinally, and finally compressed and imaged on the display screen of the CCD camera. The inversion process is as follows: the decoder accepts the coded and compressed two-dimensional data image from the CCD and uses the TwIST algorithm to invert and reconstruct the three-dimensional data (x-y-t) containing high-speed dynamic processes.

实施例Example

本实施例对单纳米颗粒发光过程进行模拟。This example simulates the luminescence process of a single nanoparticle.

本实施例各部件参阅图1设置。光线经由第一透镜1到达分束器4，然后一分为二，变成等强度的两路光线，反射的一路经由第一空间光调制器5的调制后到达第二透镜2，同样地，透射的一路经由第二空间光调制器6的调制后到达第三透镜3，通过放置四个平面镜7、8、9、10使两路光线都进入内置CCD的条纹相机11内，最后解码器12接收来自条纹相机内置CCD所拍摄的像，运用TwIST算法进行图像解码重构。The components of this embodiment are set up with reference to FIG. 1 . The light reaches the beam splitter 4 through thefirst lens 1, and then is divided into two parts to become two rays of equal intensity. The reflected one reaches thesecond lens 2 after being modulated by the firstspatial light modulator 5. Similarly, The transmitted one is modulated by the second spatial light modulator 6 and then reaches thethird lens 3. By placing fourplane mirrors 7, 8, 9, and 10, both light rays enter thestreak camera 11 with built-in CCD, and finally thedecoder 12 The image captured by the built-in CCD of the streak camera is received, and the TwIST algorithm is used to decode and reconstruct the image.

本实施例的工作过程：The working process of this embodiment:

高速动态场景先由第一透镜1成像，然后到达分束器4形成两路光路；两路光路一路经由第一空间光调制器5进行编码后再进入第二透镜2，另一路经由第二空间光调制器6进行编码后进入第三透镜3；第一平面镜7和第三平面镜9将第一条光路中从第一空间调制器5出来的光信息通过反射使其进入条纹相机11，第二平面镜8和第四平面镜10将另一条光路中从第二空间调制器6出来的光信息进行方向改变也使其进入条纹相机11；条纹相机11对不同时刻进入的编码后的图景进行纵向位置移动，最后压缩成像在CCD相机的显示屏上。最后再由解码器12接受来自CCD上的编码压缩后的二维数据图像并利用TwIST算法，反演重构出含有高速动态过程的三维数据(x-y-t)。The high-speed dynamic scene is first imaged by thefirst lens 1, and then reaches the beam splitter 4 to form two optical paths; one of the two optical paths is encoded by the firstspatial light modulator 5 and then enters thesecond lens 2, and the other path passes through the second space. The light modulator 6 enters thethird lens 3 after encoding; the first plane mirror 7 and thethird plane mirror 9 reflect the light information from the firstspatial modulator 5 in the first optical path to make it enter thestreak camera 11, and the second Theplane mirror 8 and thefourth plane mirror 10 change the direction of the light information coming out of the second spatial modulator 6 in another optical path to make it enter thestreak camera 11; thestreak camera 11 performs longitudinal position movement on the encoded scene entered at different times , and finally compressed and imaged on the display screen of the CCD camera. Finally, thedecoder 12 receives the encoded and compressed two-dimensional data image from the CCD and uses the TwIST algorithm to invert and reconstruct three-dimensional data (x-y-t) containing high-speed dynamic processes.

本发明中的空间光调制器(SLM)对光辐射强度调控的多元化体现在其在各像素的位置处的编码不一定是0或者1，而是可以随机选取的一个数值。如矩阵(1)表示，所以光辐射的信息将更加多样化。The multiplicity of the spatial light modulator (SLM) in the present invention in regulating the intensity of light radiation is reflected in that the coding at the position of each pixel is not necessarily 0 or 1, but a value that can be randomly selected. As represented by matrix (1), the information of light radiation will be more diverse.

于此实施例中，参考图2和图3，图2是高斯光束先增强后减弱的过程图，图3是基于本发明模拟的高斯光束先增强后减弱的过程图。In this embodiment, referring to FIG. 2 and FIG. 3 , FIG. 2 is a process diagram of the Gaussian beam being enhanced first and then weakened, and FIG. 3 is a process diagram of the Gaussian beam being enhanced first and then weakened based on the simulation of the present invention.

对比例Comparative ratio

早期压缩超高速摄影装置(CUP)中只使用一个DMD对光路进行调控，因而只利用了分束器中的单一光路。数字显微器件(DMD)对光辐射强度的调控可以用矩阵(2)表示(以5*5像素来说明)，当从前一个器件输出的光通过DMD时，DMD相当于光开关，1通过，0不通过，所以携带图像信息的光辐射会加入一个伪随机的二元化编码。Early compression ultra-high-speed photographic units (CUPs) used only one DMD to steer the light path, thus utilizing only a single light path in the beam splitter. The regulation of the light radiation intensity by the digital microscope device (DMD) can be represented by matrix (2) (illustrated by 5*5 pixels). When the light output from the previous device passes through the DMD, the DMD is equivalent to an optical switch, 1 passes through, 0 does not pass, so a pseudo-random binary code is added to the optical radiation carrying the image information.

事实上，如果将光束分成若干束，并且每束都通过一个DMD，然后再同时叠加到CCD相机上也可以实现类似的效果，如式(3)所示，两个0-1随机的矩阵相加就出现了多于两种的情况，所以多个DMD作用效果的叠加也会出现这种光辐射信息多元化的效果。In fact, if the beam is divided into several beams, and each beam is passed through a DMD, and then superimposed on the CCD camera at the same time, a similar effect can be achieved, as shown in equation (3), two 0-1 random matrices phase In addition, there are more than two cases, so the superposition of multiple DMD effects will also have the effect of diversifying the light radiation information.

由于无法连续地对多个DMD进行调控，会导致实验有较大误差，并且实验规模也会过于冗余。The inability to continuously control multiple DMDs will lead to large errors in the experiment, and the scale of the experiment will be too redundant.

相比于现有的高速摄影技术，本发明具有非常明显的优势，它是单次相机拍照测量一个x-y-t事件，其中x和y是空间坐标，而t是时间坐标，因此可以观察瞬态发生事件，条纹相机的应用使得时间分辨率达到皮秒量级。此外，类似于传统成像，本发明仅仅是接收图像，因此并不像其它单次拍照成像那样需要特别的有源照明，因而可以对发光过程进行成像，例如物体荧光或者生物发光。Compared with the existing high-speed photography technology, the present invention has very obvious advantages. It measures an x-y-t event by taking a single camera shot, where x and y are the spatial coordinates, and t is the time coordinate, so the transient events can be observed. , the application of the streak camera enables the time resolution to reach the picosecond level. In addition, similar to traditional imaging, the present invention only receives images, so it does not require special active illumination like other single-shot imaging, and thus can image luminescence processes, such as object fluorescence or bioluminescence.

本发明对比目前最快速的CUP系统，既实现了双通道采样，又克服了CUP系统中使用数字显微器件(DMD)只能提供伪随机的0-1二元化编码的缺陷。在采样信息更全面的基础上还可以通过SLM来精密调控光强，使得光强调控分布可以由原来的二元化变为多元化，因而可以更加明显地区分出两幅图的差别，甚至是在很短的时间间隔内相邻两幅图之间的微小差别。具体的做法是采用空间光调制器SLM(Spacial Light Modulator))来取代原来的DMD，并且条纹相机同时接收两路经由SLM调控的光辐射。由于SLM能连续地调控光强度，所以通过调控空间光调制器，可以达到采集光信息多元化的目的，这就等同于多通道采样。Compared with the current fastest CUP system, the invention not only realizes dual-channel sampling, but also overcomes the defect that the digital microscope device (DMD) in the CUP system can only provide pseudo-random 0-1 binary coding. On the basis of more comprehensive sampling information, the light intensity can also be precisely controlled by SLM, so that the light intensity control distribution can be changed from the original binary to diversified, so the difference between the two pictures can be more clearly distinguished, even the Small differences between two adjacent images within a short time interval. The specific method is to replace the original DMD with a spatial light modulator (SLM (Spacial Light Modulator)), and the streak camera simultaneously receives two channels of light radiation regulated by the SLM. Since the SLM can continuously adjust the light intensity, by adjusting the spatial light modulator, the purpose of collecting optical information can be diversified, which is equivalent to multi-channel sampling.

参考图4，同时参阅图2和图3。图4为对比例用CUP系统重构出单纳米颗粒发光过程的九幅图。显然，不论是从图像之间的变化趋势还是相邻两幅图之间的细微差别，本发明的模拟结果都要更加契合原始过程。这就说明这种双通道、多元化的信息采集可以使信号重构的精确度提升，同时也可以进一步解决更加精密微观瞬时过程的观测问题。Referring to Figure 4, see Figures 2 and 3 concurrently. FIG. 4 is nine pictures of the luminescence process of single nanoparticle reconstructed by the CUP system in the comparative example. Obviously, the simulation result of the present invention should be more in line with the original process, no matter from the change trend between the images or the subtle difference between two adjacent images. This shows that this dual-channel and diversified information acquisition can improve the accuracy of signal reconstruction, and can also further solve the observation of more precise microscopic instantaneous processes.

Claims (5)

1. The ultra-high-speed compression photographing device based on the compression sensing and stripe camera principle is characterized by comprising a first lens (1), a second lens (2), a third lens (3), a beam splitter (4), a first spatial light modulator (5), a second spatial light modulator (6), a first plane mirror (7), a second plane mirror (8), a third plane mirror (9), a fourth plane mirror (10), a stripe camera (11) with a built-in CCD and a decoder (12), wherein the first lens (1) is connected with the beam splitter (4) through an optical path; one path of the beam splitter (4) is connected with the first spatial light modulator (5), the second lens (2) and the first plane mirror (7) in sequence through light paths; the other path of the beam splitter (4) is connected with a second spatial light modulator (6), a third lens (3) and a second plane mirror (8) in sequence through light paths; the first plane mirror (7) is connected with the third plane mirror (9) through an optical path; the second plane mirror (8) is connected with the fourth plane mirror (10) through the light path; the third plane mirror (9) and the fourth plane mirror (10) are respectively connected with a stripe camera (11) with a built-in CCD; a stripe camera (11) with a built-in CCD is connected with a decoder (12); wherein:

the first plane mirror (7), the third plane mirror (9), the second plane mirror (8) and the fourth plane mirror (10) respectively change the direction of a light path, and light is reflected into a stripe camera (11) with a built-in CCD;

the high-speed dynamic scene is imaged by the first lens (1) and then reaches the beam splitter (4) to form two paths of light paths; one path of the two paths of light paths enters the second lens (2) after being coded by the first spatial light modulator (5), and the other path of light paths enters the third lens (3) after being coded by the second spatial light modulator (6); the first plane mirror (7) and the third plane mirror (9) reflect the light information coming out of the first spatial light modulator (5) in the first light path to enable the light information to enter the stripe camera (11), and the second plane mirror (8) and the fourth plane mirror (10) change the direction of the light information coming out of the second spatial light modulator (6) in the other light path to enable the light information to enter the stripe camera (11); the stripe camera (11) moves the longitudinal position of the coded scenes entering at different moments, and finally, the coded scenes are compressed and imaged on a display screen of the CCD camera; and then a decoder (12) receives the two-dimensional data image after the coding compression from the CCD, and three-dimensional data (x-y-t) containing a high-speed dynamic process is reconstructed by inversion.

2. The imaging apparatus according to claim 1, wherein the total optical length of the two optical paths split by the beam splitter (4) is equal.

3. The camera device according to claim 1, characterized in that the beam splitter (4) is such that the energy of the reflected and transmitted two light radiations is the same, i.e. 50% of the light is reflected and 50% of the light is transmitted.

4. The photographing apparatus according to claim 1, wherein the second spatial light modulator (6) is placed on an image plane of the first lens (1), and the first spatial light modulator (5) is placed on an equivalent image plane according to an optical path length of a reflected optical path; the first spatial light modulator (5) and the second spatial light modulator (6) encode image information at different times.

5. The camera device according to claim 1, wherein the built-in CCD stripe camera (11) receives the light reflected from the third plane mirror (9) and the fourth plane mirror (10), and obtains a time-shifted two-dimensional image, which is a superimposed image of several images in a single exposure time, in the built-in CCD camera.

Images