The digital economy cannot thrive without technological support. The development of modern agriculture calls for technological innovation, and people’s yearning for a better life also requires ever-advancing science and technology to safeguard it. Hyperspectral remote sensing technology, an emerging science integrating spectroscopy and remote sensing, is demonstrating boundless possibilities and vitality in the development of the digital economy. At the recently held Forum on Innovation and Development of Aerospace Information and Digital Economy, Zhang Liren, academician of the Australian Academy of Engineering, introduced hyperspectral remote sensing technology and its applications. He noted that against the backdrop of interdisciplinary integration, research on hyperspectral remote sensing technology will integrate 5G, the Internet of Things (IoT), and cloud computing platforms to continuously serve industrial development and market demands. Spectral analysis is an important means of analysis in natural sciences. A spectrum is a pattern formed by monochromatic light, dispersed by a dispersion system from composite light, arranged in order of wavelength (or frequency). It includes not only visible light perceivable by the human eye but also light invisible to the naked eye. Through spectral analysis, human perception beyond visual range can be greatly expanded. Traditional spectral analysis is generally conducted on a single point, analyzing the interaction between the object to be measured and its self-emitted light or an external light source. Hyperspectral imaging technology, however, combines spectroscopy and imaging technology, integrating spectral resolution and graphic resolution to achieve surface spectral analysis in the spatial dimension. Its hyperspectral nature, with more spectral bands, makes it easier to distinguish subtle features of materials, resulting in more complex and accurate images of objects than ever before. Remote sensing technology refers to non-contact, long-distance detection technology. It generally involves using sensors/remote sensors to detect the electromagnetic radiation and reflection characteristics of objects. Remote sensing uses instruments sensitive to electromagnetic waves (remote sensors) to detect target objects from a distance without contact, acquiring information on their reflected, radiated, or scattered electromagnetic waves (such as electric fields, magnetic fields, electromagnetic waves, seismic waves, etc.), and then extracting, determining, processing, analyzing, and applying such information—a discipline and technology in its own right. By combining remote sensing technology with hyperspectral imaging analysis of objects, each object can be given a unique "fingerprint," through which indicators such as the object’s physical structure and chemical composition can be uniquely presented. Based on this technology, Zhang Liren’s team developed the Deep Blue Intelligent Spectrum Hyperspectral Remote Sensing Platform. "This technology features multi-machine collaboration, flexibility, and low cost. It uses a 5G IoT platform + cloud computing + database to break geographical restrictions and significantly shorten data processing time. Additionally, it can flexibly measure within a range of 30 to 500 meters according to different needs," Zhang Liren explained. Moreover, the team has achieved GPS grid synchronization, expanding sub-meter satellite mapping to centimeter-level precision. In Zhang Liren’s view, technology must be applied in practice and continuously improved in the market. Currently, his team has applied hyperspectral remote sensing technology in wheat fields to expose wheat stripe rust at its early stages. Wheat stripe rust is a typical airborne fungal disease with strict temperature requirements: it thrives at 5–12°C and cannot survive when field temperatures drop below -7°C or exceed 23°C. It overwinters, oversummers, and spreads over long distances via air currents. Temperature, humidity, and wind are key factors contributing to its spread. The infection process of wheat stripe rust fungus has four stages: contact, invasion, incubation, and onset. In the early stages, the pathogen absorbs nutrients, spreads, and reproduces inside the host, with no visible symptoms to the naked eye—making early identification critical. Missing this stage allows the disease to damage leaves, leaf sheaths, stems, and even ears, glumes, and awns. It destroys chlorophyll, causing bright yellow spots on leaves, reducing photosynthetic efficiency, robbing the plant of nutrients and water, increasing transpiration, and hindering grain filling. Yield reductions typically range from 20% to 30%, and can exceed 50% in severe cases. In mature plants, early infection leads to numerous bright yellow uredinia on the leaf surface; in later stages, the leaf epidermis ruptures, releasing rust-colored powder, and the leaves eventually wither and die. Wheat stripe rust is one of the most important wheat diseases worldwide and a major biological hazard affecting China’s safe wheat production, primarily occurring in Hebei, Henan, Shandong, Shanxi, Shaanxi, and other regions. Reports indicate that it can have a devastating impact on wheat production, causing over 40% yield loss or even total crop failure in epidemic years. Losses caused by wheat stripe rust have long been a focus of attention in wheat production. Given its latent infection characteristics, rapid identification and detection during the incubation period are of great significance for disease control, safe pesticide use, and environmental protection. Past measures to control stripe rust included large-scale field spraying and pesticide reconnaissance, but existing reconnaissance technologies suffer from lag in disease diagnosis and identification, and need improvement in neural network convergence speed and accuracy. Additionally, mathematical statistical models and prediction effects are uncertain, and wide-area forecasting based on the spatiotemporal dynamics of pathogen spores is still underdeveloped. In other words, past stripe rust control was lagging and imprecise. "So we thought of using hyperspectral remote sensing technology to identify early wheat stripe rust," Zhang Liren further explained. "What is invisible to the naked eye on early wheat leaves becomes exposed with the help of spectroscopy." First, remote sensing imaging is conducted on the target wheat field. Then, spectral features are preprocessed to eliminate interference from factors such as light angle, shadows, background, water mist, shaking, and wind, resulting in a hyperspectral image of the target area. Superpixel clustering is used to segment hyperspectral data similar to stripe rust. The energy probability distribution of the spectral features of the segmented targets is normalized, and the hyperspectral features of stripe rust are fused. Using this as a standard, early stripe rust-infected wheat plants can be quickly identified and matched. "During this process, the team created and used three core technologies," Zhang Liren noted. Window-rolling multimodal data sampling enhances effectiveness and real-time performance; a multivariate Bayesian learning model dynamically estimates the weights of interfering factors such as temperature, humidity, and wind direction, improving the accuracy of distributed sparse estimation; and this is the first time a time-frequency domain jump algorithm has been applied to small UAV-borne hyperspectral remote sensing imaging. Since spectroscopy is in the frequency domain and images are in spatiotemporal domains, their organic integration better utilizes hyperspectral remote sensing data to accurately capture wheat stripe rust pathogens. Color regions indistinguishable to the naked eye are clearly visualized using sensitive bands of 550–700 nm and 750–940 nm. The research team found that since chlorophyll content is highly sensitive to reflected spectra, early identification of stripe rust infection using the correlation between infection time and leaf chlorophyll content can accurately pinpoint affected areas. Identifying the source of infection early enables effective control. Beyond vast wheat fields, hyperspectral remote sensing technology also plays a role in vineyards in the Southern Hemisphere. In Australia’s Barossa Valley, Zhang Liren’s team has applied hyperspectral remote sensing technology. The tannin content of grapes is a key factor affecting wine quality. By analyzing nitrogen content in grapevines using hyperspectral remote sensing and inferring tannin distribution, the team has transitioned from manual tasting to hyperspectral-based deduction, making grape tannin content more visible and wine quality more controllable. "Additionally, hyperspectral remote sensing technology combined with UAVs can be used for water quality remote sensing detection. From land to rivers and oceans, from agriculture to industry, the 'fingerprints' of all things have extremely wide applications," Zhang Liren summarized. Next, the team plans to develop miniature hyperspectral spectrometers, building on the Deep Blue Intelligent Spectrum Miniature Hyperspectral Detection Technology Development Project to bring hyperspectral technology into homes. Leveraging hyperspectral remote sensing’s ability to distinguish the spectral characteristics of molecules or elements in observed substances, Zhang Liren’s team is committed to researching and applying miniature hyperspectral remote sensing combined with 5G IoT and cloud computing platforms, making it more flexible, convenient, safe, reliable, and low-cost. Further reducing the size and weight of spectrometers will integrate them into people’s daily lives, allowing more families to benefit from the technology. Miniature hyperspectral remote sensing will have extensive applications. By embedding miniaturized devices into smartphone systems, spectral data can be collected, preprocessed for storage and transmission, analyzed with cloud processing, and finally implemented on the terminal via mobile apps. "In the future, miniature hyperspectral remote sensing devices will play a huge role in scenarios such as detecting food additive content, identifying genuine and counterfeit alcohol, measuring pesticide residues, and authenticating cultural relics," Zhang Liren concluded. In fruit markets, using hyperspectral remote sensing to determine fruit sweetness not only facilitates consumers’ shopping but also helps fruit merchants plan transportation schedules in advance to prevent natural decay due to excessive sweetness during transit. Hyperspectral technology in beauty is another promising direction: hyperspectral imaging can detect facial blemishes, evaluate the suitability of cosmetics, and provide personalized skincare recommendations. Imagine a future where a small device can determine food freshness, authenticity of items, and water quality—undoubtedly improving people’s quality of life. Zhang Liren emphasized: "Commercial success is the only criterion for testing technological success. We must always value the market and apply technology in it." This is his belief and a requirement for the research and development of hyperspectral remote sensing technology. "I once climbed Mount Everest, but stopped at 6,400 meters, very close to the summit," said Tong Qingxi, academician of the Chinese Academy of Sciences (CAS) and researcher at the CAS Aerospace Information Research Institute, using his personal experience to describe China’s position in hyperspectral technology: advanced, either keeping pace or leading; in a critical climbing phase, but close to the summit. Academician Tong Qingxi made these remarks at the unveiling ceremony of the Joint Laboratory for Innovative Applications of Hyperspectral Technology. On that day, the "Joint Laboratory for Innovative Applications of Hyperspectral Technology," co-founded by Inspur Cloud Information Technology Co., Ltd. (hereinafter referred to as "Inspur Cloud") and Zhongke Puguang Technology (Tianjin) Co., Ltd. (hereinafter referred to as "Zhongke Puguang"), was officially established. Positioned as a hyperspectral technology R&D center, innovative application center, and high-end think tank, the laboratory will promote the implementation of hyperspectral technology in industrial internet, digital agriculture, and "dual carbon" businesses. Experts believe this will "accelerate" the R&D and application of China’s hyperspectral technology. Different elements and their compounds on Earth have unique spectral characteristics, with spectra regarded as the "fingerprints" for identifying substances. Hyperspectral technology acts as "tinted glasses" helping humans see these "fingerprints" clearly. In the industry, CAS is regarded as the "Huangpu Military Academy" of hyperspectral research, with Academician Tong Qingxi as one of its "leaders." In an interview with *Science and Technology Daily*, Tong Qingxi listed typical application scenarios of hyperspectral technology: In cultural relic protection, for example, experts use hyperspectral scanning imaging and color fusion technologies to extract relic seals, identify authenticity, extract ink marks, and recognize pigments, providing effective technical support for recording and preserving history. In digital agriculture, experts perform band calculations, analysis, processing, and storage on collected hyperspectral data (such as key physiological parameters like biomass, leaf area index, and chlorophyll) based on a large database of surface object spectra, enabling real-time, efficient, and precise monitoring and management of crops, including fine classification, yield estimation, growth monitoring, pest detection, and activity monitoring. Additionally, hyperspectral technology can be used in material identification, target detection and recognition, mineral identification, and environmental monitoring of water and soil pollution. Zhang Lifu revealed to *Science and Technology Daily* that the laboratory focuses on three directions: "industrial internet, digital agriculture, and dual carbon businesses." For example, in the industrial internet, it provides coal calorific value detection and industrial equipment lubricating oil testing services; in smart agriculture, it offers detection of crop nitrogen, phosphorus, and potassium content, pest identification, and crop growth monitoring; and in "dual carbon" businesses, it provides carbon metering equipment and dual carbon control services. As a researcher at the CAS Aerospace Information Research Institute, Zhang Lifu has promoted the transformation of scientific and technological achievements through entrepreneurship. Now, the establishment of the new laboratory has given him new expectations: the laboratory will take on the responsibility of exploring efficient mechanisms for transferring and transforming scientific and technological achievements, building a model of industry-academia-research integration, and seizing the high ground in technological competition. He believes the laboratory should play three roles: a "highland" for talent gathering, a "position" for technological innovation, and a "base" for achievement transformation. Xiao Xue, on the other hand, hopes that the strong alliance between the two parties will further accelerate application innovation and achievement transformation based on cloud as the foundation, industrial internet platform as support, and hyperspectral technology as the key technology. Hyperspectral remote sensing is basically non-destructive and non-contact detection. Most cultural relics and historic sites are ancient and irreplaceable, making them vulnerable to damage from contact measurement. In this regard, hyperspectral remote sensing aligns perfectly with the protection of cultural relics and historic sites. Integrating the two is a trend in modern scientific and technological development, as well as a demand and call from the archaeological community. Many cultural relics and historic sites have suffered some degree of damage, weathering, or corrosion on their surfaces or interiors after thousands of years. These relics often have high historical value and cannot be arbitrarily subjected to contact detection and research, posing great challenges to evaluating their damage. The Leshan Giant Buddha has a history of over a thousand years, with severe surface wear. According to experts’ estimates, the Buddha we see today is much "thinner" than the original, indicating severe weathering and corrosion. Using contact methods to assess surface damage is not only extremely difficult and ineffective but also risks further harming the surface, exacerbating weathering. Hyperspectral remote sensing solves this problem: as a non-contact detection method, it reduces detection costs while protecting the relics, making it highly reliable. We know that different substances respond differently to different bands of hyperspectral remote sensing images, a property of hyperspectral remote sensing. This allows us to obtain internal information about cultural relics that is difficult to detect from the surface. In practice, by identifying bands sensitive to the hidden information of the Buddha and conducting in-depth research using these bands, we can capture features inaccessible through ordinary methods, restoring lost information. This technology has been applied in practice: for example, Britain’s *Sunday Times* first revealed on May 28, 2006, that Seracini used multispectral imaging technology to discover the bloody scene behind Leonardo da Vinci’s *The Adoration of the Magi*. Applying hyperspectral remote sensing to protect the Leshan Giant Buddha will therefore be highly effective. On one hand, hyperspectral remote sensing data can help us understand the Buddha’s damage and develop better protection measures to prevent further harm. On the other hand, it can reveal more information to predict potential issues, such as identifying cracks or depressions, enabling prevention before irreversible damage occurs. Hyperspectral remote sensing is also effective in dating cultural relics. According to remote sensing knowledge, even cultural relics of the same type differ significantly in raw materials and processing techniques due to different ages, leading to distinct spectral characteristics in their compositions. Using hyperspectral remote sensing to analyze relics makes it easy to identify their compositional features and infer their approximate age. However, dating relics with hyperspectral remote sensing requires testing a large number of samples of the same material, accumulating extensive data, and establishing a comprehensive spectral fingerprint database. This database provides a reference standard for accurately determining the age of relics. Thus, building a sufficiently large database, ensuring it covers diverse aspects, and maintaining it are practical challenges in using hyperspectral remote sensing for dating. Leshan, a world-renowned region rich in cultural heritage, has a large number of relics, but experts remain divided on the age of many of them. Hyperspectral remote sensing can be considered for dating these undated relics. Compared with traditional methods, it better protects relics from secondary damage during dating and ensures measurement accuracy. Promoting its use in dating and accelerating the development of a more comprehensive spectral fingerprint database are of great practical significance in the long run. Digitalization has become a general trend in modern life, including in the field of cultural relics and historic sites. Many digital museums already exist, such as the Digital Forbidden City of the Palace Museum in Beijing. However, common digital museums today typically involve only visible light band images. While effective and practical, they leave room for further research compared to hyperspectral remote sensing. Hyperspectral remote sensing provides richer, more three-dimensional spectral information than visible light, making it an excellent choice for digital museums. Ordinary visible light-based digital stereoscopic images only record and reproduce the spatial information of relics, lacking comprehensive, three-dimensional preservation and reproduction of further details, limiting in-depth research. Hyperspectral remote sensing captures more information than visible light, including not only 3D and color information but also spectral information. Spectral information reveals many hidden details, such as the relic’s material composition, historical changes, 3D structure, and appearance. In short, hyperspectral remote sensing can use this information to assess damage, infer age, evaluate integrity, and reconstruct the original appearance of damaged relics. For example, Egyptian archaeologists used hyperspectral remote sensing to digitally reconstruct the sunken ancient city of Alexandria, achieving excellent results. Since the city is submerged underwater, contact measurement and evaluation are impossible, but hyperspectral remote sensing provided scientists with an effective means to showcase the thousand-year-old underwater city to the world. For Leshan, establishing a digital museum using hyperspectral remote sensing is also a necessary choice. A hyperspectral remote sensing digital museum of the Leshan Giant Buddha or Mount Emei would, on one hand, enable analysis of relics using their spectral information to better protect Leshan’s cultural heritage. On the other hand, displaying spectral information alongside ordinary relic data in a digital museum would provide more information, scientific analysis, and reasonable protection recommendations, resonating more strongly with audiences. For example, making the hyperspectral remote sensing digital museum of the Leshan Giant Buddha available online would allow more people to learn about Leshan’s precious historical resources. This is an excellent choice both for relic protection and promotional effects. The applications of hyperspectral technology are truly extensive; the above only briefly covers a few. Additionally, in the documentary *Masters in Forbidden City*, researchers used hyperspectral remote sensing technology to scan ancient calligraphy and paintings, extract ink marks, and identify pigments. Substances have unique spectral characteristics, like fingerprints. Hyperspectral remote sensing helps people see them more clearly, so it can be used to monitor natural disasters, classify land use, and conduct fine crop classification. Overall, China’s remote sensing technology started relatively late, but hyperspectral remote sensing is one of the few remote sensing technologies that keeps pace with international frontiers. In the early days of reform and opening up, CAS organized experts to develop remote sensing mineral exploration equipment using hyperspectral remote sensing. I worked with a group of outstanding experts from institutions like the Shanghai Institute of Technical Physics on this project. Based on the actual technological level at the time, we independently developed the first generation of infrared sub-spectral scanners. Though simple by today’s standards, it was a crucial step in China’s hyperspectral remote sensing development. Subsequently, CAS imported two "Citation" remote sensing aircraft. We fully refitted them for remote sensing and had them fly over the vast land of Xinjiang. Over two years, we identified multiple gold alteration zones and mineralized zones, delineating several prospecting targets. These efforts pushed China’s hyperspectral remote sensing into a period of rapid development. During their over 30 years of service, these aircraft also played important roles in monitoring soil erosion on the Loess Plateau, surveying the Three-North Shelterbelt, and national land resource surveys. Entering the 21st century, China’s hyperspectral remote sensing developed rapidly. Environmental satellites, Shenzhou spacecraft, and satellites under the national major project "High-Resolution Earth Observation" all carry hyperspectral remote sensing payloads. In recent years, the development of hyperspectral equipment such as "visible-shortwave infrared hyperspectral cameras" has advanced China’s hyperspectral remote sensing technology to a new stage. From its start to vigorous development, from exploratory research to in-depth application, China has always kept pace with international frontiers in hyperspectral remote sensing, with some technologies leading the world. Looking back on the development of hyperspectral remote sensing technology, it is clear that national support is the greatest driving force, and the nationwide system has provided strong impetus for many technological innovations. I remember that to track cutting-edge developments in Earth observation technology, CAS selected personnel from its institutes to establish an Earth resource satellite research team, joining forces with national research institutions to develop related technologies. When the country formulated science and technology research plans, there were many important research areas that needed urgent development and could yield quick results. Despite limited national financial resources at the time, hyperspectral remote sensing, as a frontier technology, still received significant support. Looking to the future, hyperspectral remote sensing still has obvious shortcomings in technological maturity and the breadth and depth of applications. However, with the efforts of the new generation of researchers, its prospects will undoubtedly grow brighter. Sources: *High Technology & Industrialization*, *Science and Technology Daily*, Laissen Optics, People’s Network