INFORMATION TECHNOLOGY

Good Mentors and Friends, Study and Swim – Interview with Professor Cao Liangcai of Tsinghua University


Editor’s note

Holographic optics is a technology based on the principle of light wave interference and diffraction to achieve image recording and reproduction, which can highly restore the three-dimensional characteristics of objects and provide an immersive visual experience. Since holographic optics was proposed in 1947 and won the Nobel Prize in Physics in 1971, it has gradually developed into two major research directions: digital holographic imaging and computational holographic display, deeply empowering the development of 6G communication, smart medical care, MR commercial headsets and other fields. The general solution ideas of optical inverse problems contained in it also provide theoretical support for the extensive combination of holographic optics with computational lithography, optical metamaterials, optical neural networks, OAM and other directions in recent years, showing great potential research and application value.

We are honored to invite Professor Cao Liangcai, a leading scientist in the field of holographic optics, to deeply interpret the opportunities and challenges faced by the development of holographic optical technology. During the interview, Professor Cao Liangcai showed us his latest scientific research achievements, took us back to the past and present of holographic optics, shared his wonderful stories during academic visits and exchanges, and introduced the “congyou culture” he advocates in talent training. Next, please follow the Light characters to appreciate the outstanding demeanor of Professor Cao Liangcai.

Professor Cao Liangcai

Cao Liangcai is a professor and doctoral supervisor of the Department of Precision Instruments of Tsinghua University. He is a Fellow of the International Society of Optical Engineering SPIE and the American Optical Society OPTICA, and a Distinguished Professor of Education Minister Jiang Scholar. He received his Ph.D. in optical engineering from Tsinghua University in 2005, and has been a visiting scholar at the University of California, Santa Cruz and the Massachusetts Institute of Technology, mainly teaching “Information Optics” and “Modern Optical Experiments” courses, mainly focusing on holographic optical imaging and display technology. He has presided over the key projects of the Natural Science Foundation of China and the key R&D projects of the Ministry of Science and Technology, published more than 100 academic journal papers, applied for more than 40 invention patents, won the first prize of Beijing Science and Technology Award, the first prize of technical invention of China Instrumentation Society, the “Academic Newcomer Award” of Tsinghua University and the “Advanced Worker” of Tsinghua University, and won the champion tutor of the National Academic League for Optical and Optical Engineering Doctoral Students hosted by Light Academic Publishing Center in 2021.

Interviewer: Cao Liangcai (Tsinghua University)

Interviewer: Sun Tingting

Sun, T. Light People: Professor Liangcai Cao. Light Sci Appl 12, 138 (2023).

https://doi.org/10.1038/s41377-023-01194-3

Q: Can you briefly introduce the current research focus? What important progress has been made?

A: My research group is the Holographic Optics Laboratory, also known as Hololab. I entered Tsinghua University in 2001 to pursue a Ph.D. in optical engineering, and under the guidance of Academician Jin Guofan and Professor He Qingsheng, I devoted myself to the research of volume holographic optical storage and optical computing. I have been doing research in this area since graduating in 2005. In 2013, I had the honor to participate in the projects led by Academician Dai Qionghai and Professor Wang Yongtian respectively, and began to turn to the research of holographic optical imaging and holographic optical display technology. Over the past decade, Hololab’s main research interests have focused on computational holographic displays and digital holographic imaging. Computational holography (CGH) display can convert the digital model of the object into a two-dimensional hologram through a computer, and directly project the original three-dimensional object using liquid crystal spatial light modulators or metasurface devices. Digital holographic (DH) imaging is the opposite, which uses a photoelectric image sensor to record the optical information of the object, collects the light field information to obtain a two-dimensional hologram, and then reconstructs the hologram through the computer to obtain a digital model of the object. These processes involve relatively basic mathematical theories and diffractive optical calculations, and some research results can be generalized to other optical imaging and display technologies, but they all belong to the category of information optics and holographic optics. We have successfully proposed computational holographic display algorithms and systems based on angle spectrum theory and deep learning, and also developed lensless imaging methods based on compressed sensing and deep learning, which have been recognized by peers and have been initially applied in the industry.

Figure 1: Academician Jin Guofan instructs Professor Cao Liangcai to carry out holographic optical research (September 2008)

Q: As a lensless imaging technology, coding imaging has the advantages of light structure and easy construction, and how to establish the connection between the scene and the image, and invert the image by solving the inverse problem is a key problem in coding imaging. You and Professor George Barbastathis of MIT jointly proposed a coding imaging technique using a monolithic Fresnel ribbon to achieve lensless imaging under incoherent illumination (Light Sci Appl 9, 53 (2020).) and published in the top international journal Light: Science & Applications (Light). Can you explain the main research idea of this work? What are the advantages of this technology?

A: Coding imaging technology is to place a mask with a specific pattern into the imaging system, and the incident light is modulated by the mask pattern to form a seemingly chaotic encoded image on the image sensor, and finally the computer algorithm recovers the original image from the encoded image. How to establish the connection between the scene and the image, and invert the image by solving the inverse problem, is a difficult problem in coding imaging. While working with Professor David Brady on digital holography, our team published a paper in Physics Review Letters that used a compressed sensing algorithm to reveal the physical mechanism of twin removal (Phys. Rev. Lett. 121, 093902 (2018)). So we started thinking about whether we could use this method for lensless imaging. After trying, it is found that in the waveband sheet lensless imaging system, according to the sparseness of the twin image and the original image in the gradient domain, the TV regularization constraint can be added to the objective function, which finally effectively solves the discomfort of the problem.

In this Light work, we used a single Fresnel bandpiece to achieve lensless-free imaging under incoherent illumination. The imaging setup consists of only an image sensor and a Fresnel bandpiece placed a few millimeters in front of the sensor. We take advantage of the similarity between the point source hologram and the Fresnel band sheet structure, encode the incident light by using the Fresnel band sheet as a coding mask, and the encoded image has the same form as the coaxial hologram, and the inherent twin image noise of holographic reconstruction is eliminated by the compressed sensing algorithm, and high-quality image reconstruction without twin image is realized. Compared with other coding imaging technologies, this technology has the advantage of imaging without calibration and single exposure, and can be integrated with mobile terminals, security monitoring, automatic car driving and other equipment in the future.

Q: Digital holographic imaging uses computational methods to reconstruct the amplitude and phase information of the original object light wave, which can quantitatively analyze the interaction between light and matter, can you talk about the main application fields of digital holographic imaging technology? What are the important impacts of its development on our real lives?

A: The phase of the light wave mainly reflects the information of the optical path difference, that is, the change of the refractive index or thickness of the object. In biomedical imaging, there are certain differences between the refractive index of different biological tissues and the external environment, and digital holographic imaging technology can use this difference to achieve label-free observation of life phenomena, which has less impact on the normal physiological activities of cells than chemical staining, fluorescent staining and other observation methods. In optical detection and metrology, the thickness fluctuation of the sample can also be reflected in the phase change, and the digital holography can achieve nanometer-level precision surface shape measurement through the phase change. In addition, because the digital holography records all the information of the light field, the light field can be further inverted to achieve three-dimensional imaging, which is also one of the hot issues in current academic research.

Because digital holographic imaging has high imaging resolution and can provide phase information that cannot be obtained by traditional imaging technology, it has been widely used in interferometry, fine particle detection, biomedical imaging and other fields. The digital holographic microscope developed based on digital holographic technology has gradually matured from the research and development stage. The unique advantages of digital holography technology in data processing will combine it with a variety of technologies to provide strong support for precision metrology, microstructure imaging detection, medical diagnosis and other fields.

Q: As a leading scientist in the field of digital holographic imaging, can you talk about the research status of digital holographic imaging technology? In your opinion, what does the future hold?

A: Holographic imaging technology has a history of more than 70 years since its birth, and the corresponding imaging technology has developed relatively maturely, but limited by complex imaging optical paths and heavy optical components, holographic imaging technology mostly remains in the laboratory. As a computational imaging technology, digital holographic imaging technology has received more and more attention in recent years. For example, the Digital Hoography and 3D Imaging Conference is one of the theme conferences of the Optical Society of America, which is held every year around the world, and researchers from academia and industry around the world come together to exchange and discuss.

With the maturity of digital holographic imaging technology, many companies have begun to industrialize beneficial attempts, aiming at biomedical imaging, optical detection and measurement and other application scenarios, Switzerland’s Nanolive, South Korea’s Tomocube and other companies, as well as domestic Beijierui, Nanjing University of Science and Technology Intelligent Computing Imaging Institute and many other enterprise units have successively launched a series of digital holographic microscopy imaging instruments. Regarding the future of digital holographic imaging technology, from the perspective of application, in order to truly achieve industrial landing, it is necessary to deeply integrate with downstream application needs. For example, how to specifically link optical phase information with information of interest to biologists or doctors is one of the key issues driving digital holography technology to the biomedical laboratory.

In the future, we also hope to introduce optical micro-processing technology into holographic imaging technology, integrate the holographic imaging optical path on a chip like semiconductor technology, make the imaging structure more compact, and give full play to the advantages of holographic imaging such as large field of view and high resolution.

Figure 2: Inviting domestic and foreign counterparts in the field of holographic optics to carry out academic exchanges at Tsinghua University (July 2018)

Q: In recent years, computational optical imaging technology has developed rapidly, can you take us to understand the past and present of computational optical imaging technology? Your team recently published snapshot optical imaging and inspection technology based on compressed digital holography at Light: Advanced Manufacturing 4, 6 (2023), which received a lot of attention from peers. You represent the introduction of compressed sensing and deep learning for the study of computational imaging techniquesNew ideas have been opened, what are the specific manifestations?

A: The development of computational imaging technology is inseparable from the development of semiconductor technology and Internet technology. On the one hand, with the continuous development of submicron, deep submicron and nanotechnology processes and the improvement of device structure, the image sensor technology for light intensity response is becoming more and more mature, and the resolution, signal-to-noise ratio, dynamic range and other performance of image sensors are constantly improving. In addition, with the help of other optical components or additional imaging devices and optical paths, other dimensional information such as phase, polarization, and spectrum can be converted into light intensity information, which can be recorded by image sensors to achieve all-round perception of light field information in the physical world. On the other hand, with the rise of the mobile Internet and the improvement of computer computing power, research fields such as image processing, computer vision, machine learning, and big data processing have developed by leaps and bounds. The collision and integration of these two research fields has promoted the birth and development of computational imaging.

Compressed sensing and deep learning, as typical representatives of model-driven and data-driven methods, are widely used in various computational imaging techniques. Compressed sensing and deep learning verify theoretically and practically that prior knowledge in signals can be summarized and utilized, so that complex signals can be expressed with less information. As early as 2009, Professor David Brady theoretically proved that the coding method of holographic diffraction meets the conditions of compressed sensing, and proposed the concept of “compressed holography”, and subsequently, the application of compressed sensing in the field of optical imaging has emerged everywhere, such as single-pixel imaging, spectral imaging, depth imaging and so on. However, compressed sensing also has shortcomings, and its dependence on prior knowledge and long computing time also makes its development bottleneck. Deep learning technology can learn the potential mapping relationships contained in the dataset through the training of a large number of datasets. As a data-driven computing method, deep learning technology gets rid of the dependence of analytical and numerical solving methods on forward models and prior information, and provides a new solution for the solution of optical inverse problems. In some imaging fields with complex models, such as scattered medium imaging and dark and low light imaging, superior results have been achieved. However, in recent years, the uninterpretability of deep learning has gradually attracted people’s attention, and how to organically combine imaging models and the superior performance of deep learning is an important direction for future thinking and research.

Figure 3: Professor Cao Liangcai was awarded the Fellow of the International Society for Optical Engineering (August 2019)

Figure 4: Professor Cao Liangcai was awarded the Fellowship of the Optical Society of America (October 2020)

Q: Many of your team’s work is based on deep learning technology to achieve fast, high-quality 3D imaging and display. What role can deep learning technology play? What are the opportunities and challenges facing holographic dynamic 3D display technology?

A: The ideal human-computer interaction device should be able to respond to user instructions in a timely manner and switch three-dimensional display screens in a high-quality and fast manner. This process not only involves imaging and display, but also places high demands on the fineness and timeliness of technologies such as modeling, machine vision, motion capture, and semantic understanding. Deep learning techniques provide innovative solutions to these problems. For example, 3D imaging is often limited by the device and environment, resulting in color distortion, uneven brightness, and depth distortion. Fixing these problems requires a lot of computing power, and the speed of repair is difficult to meet the needs of immersive interaction. Imaging enhancement technology based on deep learning can automatically screen and repair abnormal images, improve the speed and quality of 3D imaging, and is an important support for immersive human-computer interaction.

Compared to two-dimensional displays, three-dimensional displays can provide image content that is closer to the real world. Computational holographic 3D display can provide all kinds of depth clues, the visual experience is more friendly, the 3D effect is more realistic, is one of the ideal solutions for 3D display. Compared with the current commercial headset scheme based on the principle of binocular vision or light field display, computational holography technology avoids the occurrence of radial adjustment conflicts in principle, so that viewers can obtain an immersive interactive experience without vertigo and other visual fatigue. Current problems include limited hologram reconstruction quality, limited modulation performance of wavefront modulation devices, limited spatial bandwidth of holographic display systems, and limited holographic 3D content sources. With the breakthroughs of the above challenging problems one by one, holographic three-dimensional display technology will have broad application prospects in the fields of intelligent manufacturing, distance education, remote office and entertainment and social networking.

Q: Your team has proposed a new lensless fiber optic microendoscopic imaging technique (Light Sci Appl 11, 204 (2022).), which is only the size of an embroidery needle but can achieve an ultra-high magnification of 1000 times, allowing doctors to see tissue surface cells while greatly reducing patient pain. Can you tell us about the innovation of this work? What accuracy can be achieved today? What are the key questions that remain to be addressed for follow-up research?

A: This work combines coaxial holographic technology, phase recovery algorithm and fiber beam imaging to achieve quantitative phase imaging based on fiber endoscope. Quantitative phase imaging can provide rich image contrast information, such as cell dynamic contrast, reflectivity contrast, refractive index contrast, phase information, etc., and can simulate the effect of fluorescence staining, so that cancer cells can be clearly distinguished from complex backgrounds. In addition, important physical parameters such as cell volume, refractive index, and mass can also be calculated from quantitative phase reconstruction maps, providing more valuable auxiliary information for clinical diagnosis and research. It is worth mentioning that the working distance of this lensless fiber microendoscope has been significantly increased from 10 microns to 10 millimeters, and the smallest object can be “seen” by 1 micron, and the nanoscale three-dimensional reconstruction is realized. In follow-up studies, we will further study the transmission characteristics of optical fibers and improve the lighting method to make this technology better suitable for in situ observation.

Q: We know that you have been a visiting scholar at UCSC and MIT, can you share this story with us? What was the biggest takeaway from these experiences? Compared with China, what are the characteristics of foreign scientific research atmosphere?

A: I had the privilege of visiting Professor Claire Gu’s research group at UC Santa Cruz in 2009 and Professor George Barbastathis at MIT in 2014. This has been very helpful for my international academic exchanges, and I have had the opportunity to meet many international scientists and very active young scholars. Both overseas experiences also opened up new research directions for me and had a profound impact on my research after returning to China. After returning home, I have been in touch with tutors abroad, asking them to continue to give guidance and cooperate effectively.

During the visit to the United States, research is relatively simple, simple and happy work, and the teachers and students of the research group are very concerned about the scientific problem itself, and carry out more important and pure research, which is very helpful for improving their academic level. The group I interviewed was not very large, but they were very interested in academics, and they were able to carry out long-term and in-depth research on a small problem, which produced higher level results. In addition, teachers and students abroad are more easy-going, everyone calls each other by their first names, teachers and students are more active and engaged in their work, and the degree of professionalism and professionalism of research work is relatively higher. Therefore, I strongly encourage the students of the lab to go abroad for exchange, which can cultivate their global perspective and improve their competitiveness in future career development.

Figure 5: Professor Cao Liangcai with Professor Claire Gu during his visit at UC Santa Cruz (August 2009)

Figure 6: Professor Cao Liangcai at MIT with Professor George Barbastathis (September 2014)

Q: In the 2021 National Optics and Optical Engineering Doctoral Academic League held by Light, you led the Beijing Division to achieve excellent results, and you yourself won the honorary title of champion supervisor. What do you think are the positive implications of holding such an academic event?

A:The Light Academic League has built a platform for outstanding doctoral students engaged in optics and optical engineering to display academic ability and exchange academic ideas. Students took advantage of this excellent platform to fully express their research motivation, research ideas and research results for their doctoral dissertation work, which is a very rare opportunity for cultivating an excellent doctoral student. The academic league provides an opportunity for academic reports, which are watched online by tens of thousands of listeners, with questions, suggestions and evaluations from judges and teachers, and outstanding students competing on the same stage. Therefore, to some extent, it surpasses the thesis proposal and defense report in the school, and also surpasses the usual academic conference report. Therefore, it will certainly be of great help to improve the level of doctoral dissertations and academic reports of students. The country is now cultivating a new generation of innovative optical talents, in addition to diligent thinking and daring to practice, they must also be good at expression. Giving back to your research by disseminating your ideas through expression. In this sense, each contestant should actively think about how to articulate an achievement in order to have an impact on academic peers and even society as a whole. In addition, the innovative competition format also promotes communication and understanding between participants, forming a good competition and cooperative relationship. This allows doctoral students to earn honors while also making a group of like-minded peers.

Figure 7: Professor Cao Liangcai won the 2021 Academic League Champion Tutor (May 2021)

Q: There are two important characteristics of innovative talents, one is to create, that is, to surpass themselves, and the other is to be new, that is, to be different. In terms of talent training, you advocate “teachers and students travel, innovation and progress together”, can you explain this unique training concept for us?

A: I strongly encourage undergraduate students to participate in surgical creation activities and research activities. Through communication and interaction in scientific research activities, undergraduates can continue to shape the correct value orientation and personality qualities such as courage, perseverance, self-confidence and team spirit, and firmly establish the concept of combining learning basic knowledge with the determination to devote themselves to scientific research in their professional field, so as to have the ability of lifelong learning and the potential for continuous development of innovation. Students come to Tsinghua after choosing their major after the college entrance examination, and their understanding of the major may not be deep enough, so I will focus on cultivating students’ professional interests and improving their scientific research quality and research ability. If there is no interest, no matter how hard and smart it is, it is futile. The impulse for scientific research should come from you discovering the imperfections in the world and then can’t wait to change the world to make it more perfect. The scientific research impulse of a large number of top scholars in Tsinghua is precisely the glory of Tsinghua. In addition, some students’ career development direction is a little vague, I always encourage students to clearly set big aspirations, go to the big stage, become a big career, according to the academic masters and typical cases in this field, with the eight words of “thinking of the world, illuminating life” to inspire students in the direction of optics to work hard, use “light” to illuminate life, and guide students to take the optical career as their lifelong development career. When choosing an industry, we should not only consider the interests of individuals and families, but “think of the world” and strive to make a difference.

Figure 8: Professor Cao Liangcai with graduate and undergraduate students attending CIOP2022 conference in Xi’an (August 2022)

Q: In July 2022, Light Science Popularization Square was officially established, providing clear explanations, credible answers and reliable comments for common problems in our lives. We know that there is still a long way to go to bring science to the public, and as a member of the Advisory Board of Scientists, what advice do you have for this?

A: Science popularization work is the job of every researcher, only by explaining their work clearly to the general population, this work has high social value. I also often encourage PhD students to simplify their work and introduce it to family and friends. Our research can only be widely disseminated when everyone around us can understand it. Like classroom teaching, teaching and learning are mutual, and the amount of knowledge taught by teachers is not the only criterion for teaching quality, what students really master is the most important. The same is true for scientific research, the number of scientific research results is not equivalent to the application of scientific research results, we are often asked to use one or two sentences to explain the complex work clearly in academic reports and project defense, which is the so-called road to simplicity. On the other hand, popular science work can not only stimulate the public’s interest in science and technology, but also promote more people to participate in scientific and technological innovation work. Government capital and taxpayers are more willing to invest funds in scientific research, and in the future, China will have more social capital and individuals or groups to support the country’s scientific and technological research through donations, which is also a long-term return on science popularization. I once visited the Optical Museum in Changchun, and it was very rewarding and insightful. It is hoped that Light Science Popularization Square can collect and provide more updated popular science knowledge, and at the same time, with the help of the International Day of Light to build a bridge with the world’s optical science popularization, so that more people can understand light, understand light, and chase light.

Q: Thank you for your continued support and contribution to Light, Light journals have gone through eleven years from ignorance to international renown, can you publish your message and expectations for Light?

A: The establishment of Light journal is a milestone in the history of Chinese optical development, and it is expected that Light will further go out of China and move towards the world, provide and display Chinese wisdom, Chinese solutions and Chinese strength in the process of world optical development, study light and use light, and make new and greater contributions to improving human life.

Figure 9: Hololab researchers (October 2022)

Thanks

Thank you to Sheng Yunlong (University Laval), Zhang Qihang (MIT), Liu Kexuan, Wu Jiachen, Sui Xiaomeng, Yang Yiqian, Gao Yunhui and others for their hard work and contributions to this article!

References

[1] Wu, J., Zhang, H., Zhang, W. et al. Single-shot lensless imaging with fresnel zone aperture and incoherent illumination. Light Sci Appl 9, 53 (2020). https://doi.org/10.1038/s41377-020-0289-9

[2] Zhang, W., Cao, L., Brady, D. et al. Twin-Image-Free Holography: A Compressive Sensing Approach. Phys. Rev. Lett. 121, 093902 (2018).

https://journals.aps.org/prl/abstract/10.1103/PhysRevLett.121.093902

[3] Gao, Y., Cao, L., Iterative projection meets sparsity regularization: towards practical single-shot quantitative phase imaging with in-line holography. Light: Advanced Manufacturing 4, 6 (2023).

https://doi.org/10.37188/lam.2023.006

[4] Sun, J., Wu, J., Wu, S. et al. Quantitative phase imaging through an

ultra-thin lensless fiber endoscope. Light Sci Appl 11, 204 (2022). https://doi.org/10.1038/s41377-022-00898-2

Light editing

Tingting Sun, Ph.D. in Engineering, associate researcher, currently working in the Light Academic Publishing Center of Changchun Institute of Optics, Fine Mechanics and Physics, Chinese Academy of Sciences (Changchun Institute of Optics and Mechanics), scientific editor of the leading journal of the Excellence Program “Light: Science & Applications”, assistant to the editor-in-chief of the new journal “eLight”, and jointly cultivate senior talents from the International Cooperation Bureau of the Chinese Academy of Sciences. As the project leader, he has presided over two scientific research projects, participated in a number of important scientific research projects, published a number of SCI and EI academic papers, and authorized two national invention patents.

(Source: China Optics WeChat public account)
 
Special statement: This article is reproduced only for the need to disseminate information, and does not mean to represent the views of this website or confirm the authenticity of its content; If other media, websites or individuals reprint and use from this website, they must retain the “source” indicated on this website and bear their own legal responsibilities such as copyright; If the author does not wish to be reprinted or contact the reprint fee, please contact us.



Source link

Related Articles

Leave a Reply

Your email address will not be published. Required fields are marked *

Back to top button