Visible Light Images Research Articles

UAV-Derived Spectral Indices for the Evaluation of the Condition of Rye in Long-Term Field Experiments

The aim of the study was to evaluate the effect of various fertilization treatments, including nitrogen, potassium and phosphorus fertilization, in long-term experiments for selected UAV (unmanned aerial vehicle)-derived spectral vegetation indices (NDVI—Normalized Difference Vegetation Index, NDRE—Normalized Difference Red Edge Index, VARI—Visible Atmospherically Resistant Index, TGI—Triangular Greenness Index, SIPI2—Structure Insensitive Pigment Index 2, LCI—Leaf Chlorophyll Index, BNDVI—Blue Normalized Difference Vegetation Index, GNDVI—Green Normalized Difference Vegetation Index, MCARI—Modified Chlorophyll Absorption in Reflective Index) based on multispectral (bands in the range of visible light and near infra-red) images of winter rye. The strongest effect on the studied vegetation indices was nitrogen fertilization, which discriminated values of most of the vegetation indices. The effect of phosphorus and potassium fertilization on the studied vegetation indices was much weaker. The treatments with nitrogen fertilization had significantly higher values of most vegetation indices in comparison to treatments without nitrogen. This was confirmed by principal component analysis (PCA), in which treatments without nitrogen fertilization were very different in comparison to all other treatments where nitrogen fertilization was applied. The effect of phosphorus and potassium fertilization on most of vegetation indices was relatively weak and not significant in most experiments. Only for rye cultivated in monoculture was the effect of phosphorus fertilization significant for most of vegetation indices in early growth stages. In later growth stages (heading and flowering) the effect of phosphorus fertilization was significant in rye monoculture for the SIPI2 vegetation index. Mean SIPI2 was higher for the fertilization treatment CaNPK in comparison to CaKN (without P fertilization). The effect of potassium fertilization on the studied vegetation indices was very weak, and in most cases not significant. The effect of nitrogen fertilization on vegetation indices was much stronger than effect of both potassium and phosphorus fertilization.

Performance evaluation of common-aperture visible and long-wave infrared imaging system based on a comprehensive resolution

Abstract In this article, a new method based on a comprehensive resolution is proposed to evaluate the performance of a common-aperture visible and long-wave infrared (LWIR) imaging system. The influence of the comprehensive resolution on target recognition range is analyzed. Under the condition of common-aperture and a combination of four comprehensive resolutions, the main factors affecting the ratio of target recognition range between the visible imaging system and the LWIR imaging system are discussed. Combined with the constraints of the optical system and the detector, the impact of the comprehensive resolution on the target recognition range is analyzed. The ratio of the target recognition range of the two imaging systems is modeled using the comprehensive resolution of the visible imaging system. The experimental system is built and the experimental results are acquired. The predictive results and experimental results show that when the comprehensive resolution of the visible light imaging system increases, the target recognition ranges of the visible light imaging system and the LWIR imaging system increase, as well as the ratio of the target recognition ranges of the two systems increases.

A Hierarchical Bayesian Fusion Method of Infrared and Visible Images for Temperature Monitoring of High-Speed Direct-Drive Blower

The high-speed direct-drive blower is a high-efficiency, energy-saving, and environmentally friendly blower, and it is widely applied in industrial fields. The temperature changes significantly affect the performance of the high-speed direct-drive blower, and it is crucial to monitor its temperature. Infrared and visible light cameras are simultaneously used to measure and characterize temperature information. The measurement method can have both the temperature information of the infrared image and the contrast information of the structure and contour of the visible light image. This article establishes an image fusion representation and probabilistic generation model of infrared and visible images. Then, the information of the infrared and visible images is fused under the Bayesian framework. A hierarchical prior model using the Haar wavelet transform is proposed. The joint maximum a posteriori criterion is chosen, and an appropriate alternate optimization algorithm is designed to achieve information fusion. The proposed method is validated in industrial scenarios of high-speed direct-drive blowers. The experimental results demonstrate the robustness and effectiveness of the proposed method.

Unsupervised clustering and analysis of WISE spiral galaxies

ABSTRACT The Wide-Field Infrared Explorer (WISE) survey captured massive amounts of infrared astronomical object data, and different astronomical objects will show different characteristics. Especially spiral galaxies that are richer in colour than other astronomical objects. In addition, the Sloan Digital Sky Survey (SDSS) has obtained a large number of visible light images and their parameter. So this paper mainly explores the colour characteristics of WISE spiral galaxies, and uses SDSS, 2MASS, WISE, and MANGA parameters to analyse the differences between different classes and the commonality of the same class, so that provides help for astronomers to study infrared spiral galaxies. First, the RA and Dec. of the spiral galaxy samples were crossed in GalaxyZoo2. The flux data of w1, w2, and w3 bands in WISE were crossed for image synthesis. The Bootstrap Your Own Latent contrastive learning framework and K-means clustering were used to unsupervised classify the infrared spiral galaxy images. After multiple experiments, five classes of images were selected by referring to the best clustering results. Finally, the parameter of galaxies in SDSS, 2MASS, WISE, and MANGA catalogues were crossed, including redshift, 12 magnitudes that from visible light band to mid-infrared band, stellar formation rate, stellar metallicity, stellar velocity dispersion, etc. These parameters were qualitatively and quantitatively analyzed. The effectiveness of unsupervised clustering algorithms for handling unlabeled data is demonstrated and two special classes of galaxies are found. The analysis result shows that the distribution characteristics of different parameters of different classes of infrared spiral galaxies are different.

Research on 24-Hour Dense Crowd Counting and Object Detection System Based on Multimodal Image Optimization Feature Fusion

Motivation. In the environment of day and night video surveillance, in order to improve the accuracy of machine vision dense crowd counting and target detection, this paper designs a day and night dual-purpose crowd counting and crowd detection network based on multimode image fusion. Methods. Two sub-models, RGBD-Net and RGBT-Net, are designed in this paper. The depth image features and thermal imaging features are effectively fused with the features of visible light images, so that the model has stronger anti-interference characteristics and robustness to the light noise interference caused by the sudden fall of light at night. The above models use density map regression-guided detection method to complete population counting and detection. Results. The model completed daytime training and testing on MICC dataset. Through verification, the average absolute error of the model was 1.025, the mean square error was 1.521, and the recall rate of target detection was 97.11%. Night vision training and testing were completed on the RGBT-CC dataset. After verification, the average absolute error of the network was 18.16, the mean square error was 32.14, and the recall rate of target detection was 97.65%. By verifying the effectiveness of the multimode medium-term fusion network, it is found to exceed the current most advanced bimodal fusion method. Conclusion. The experimental results show that the proposed multimodal fusion network can solve the counting and detection problem in the video surveillance environment during day and night. The ablation experiment further proves the effectiveness of the parameters of the two models.

Low-voltage-modulated perovskite/organic dual-band photodetectors for visible and near-infrared imaging

Visible and near-infrared (NIR) light dual-band photodetectors (PDs) have potential applications in signal detection, bioimaging, optical communications and safety monitoring. Herein, we report an ultrafast perovskite/organic heterojunction dual-mode PD with a voltage-modulated photoresponse range in visible and NIR spectra. The PD, comprising a perovskite layer to absorb visible light (500–810 nm) and an organic bulk heterojunction layer for NIR light absorption (810–950 nm), exhibited a switchable spectral response in the visible or NIR bands. The voltage-modulated visible and NIR photoresponses of the PD were attributable to controlled charge photogeneration in perovskite and organic blend thin films under different bias polarities. The device exhibited peak responsivities of 93.5 and 102.2 mA/W in the visible and NIR bands, respectively; a high detectivity of 4.3 × 109 Jones (at forward bias of 0.7 V and incident 625 nm light) and 1.6 × 1012 Jones (at reverse bias of –1.5 V and incident 900 nm light); a fast microsecond response time; and a wide dynamic range (>120 dB) both in the visible mode and NIR mode. Also, this voltage-modulated dual-band PD shows promising applications in visible light and NIR imaging, which is proven by demonstrating a PD array with 25 pixels (5 × 5).

Single-kernel classification of deoxynivalenol and zearalenone contaminated maize based on visible light imaging under ultraviolet light excitation combined with polarized light imaging

Deoxynivalenol and zearalenone contamination in maize poses a threat to food safety. Thus, the development of a cost-effective and non-destructive method to classify deoxynivalenol and zearalenone contaminated maize is an important research issue. Natural deoxynivalenol and zearalenone contaminated maize were randomly selected for this experiment. Visible light imaging under ultraviolet light excitation combined with polarized light imaging was used to extract texture and color features from the endosperm and germ sides. And six classification models were established for the single-kernel classification of 1035 contaminated maize. The results indicated that the optimal detection surface was the endosperm side, and the optimal model was SVM with a classification accuracy of 88.1%. Then, local feature analysis of images and mycotoxin concentrations verification experiment were used to reduce secondary contamination and further evaluate the performance of the model, respectively. This method can be verified to be effective in identifying three deoxynivalenol and zearalenone contamination levels of maize, and can also reduce the costs of test stands.

Three-Dimensional Force Sensor Based on Fiber Bragg Grating for Medical Puncture Robot

In medical puncture robots, visible light, infrared and ultrasound images are currently used to guide punctures. The lack of information about the interaction forces between the puncture needle and soft tissue in different directions during the puncture process can easily lead to soft tissue being damaged. The current three-dimensional force sensors are large and can only be mounted on the base of the puncture needle, which does not allow for easy integration. Moreover, the force transfer to the base introduces various disturbing forces and the measurement accuracy is low. To reduce the risk of soft tissue being damaged and to enhance the intelligent control strategy of the puncture robot, this paper designs a three-dimensional force sensor based on fiber Bragg gratings. The sensor is very small and can be integrated into the back end of the puncture needle to accurately measure the interaction forces between the puncture needle and the soft tissue in different directions. The puncture needle wall is designed with notched bending of a multilayer continuous beam, which can increase the sensitivity of axial stiffness, while maintaining the sensitivity of the sensor to lateral bending and torsion, and also reduce the crosstalk between the axial and lateral forces. The finite element method is used to optimize its structural parameters, and a BP neural network based on the global optimal fitness function is proposed to solve the decoupling problem between the three-dimensional forces, which effectively improves the detection accuracy of the force sensor. The experimental results show that the measurement error of the sensor is less than 1.5%, which can accurately measure the interaction force between the puncture needle and the soft tissue and improve the safety of the puncture process.

Tomography of the Solar Corona with the Metis Coronagraph I: Predictive Simulations with Visible-Light Images

The Solar Orbiter/Metis coronagraph records full-Sun visible-light polarized brightness (\(pB\)-) images of the solar corona. This work investigates the utility of a synoptic observational program of Metis for tomographic reconstruction of the three-dimensional (3D) distribution of the electron density of the global solar corona. During its lifetime, the mission’s distance to the Sun will range over \(\approx 0.3-1.0\,\mathrm{AU}\), while its solar latitude will span \(\approx \pm 33^{\circ}\). The limitations that this orbital complexity poses on tomographic reconstructions are explored in this work. Using the predicted orbital information of Solar Orbiter and 3D-MHD simulations of the solar corona using the Alfvén Wave Solar atmosphere Model (AWSoM), time series of synthetic Metis \(pB\)-images were computed and used as data to attempt tomographic reconstruction of the model. These numerical experiments were implemented for solar-minimum and solar-maximum conditions. In both cases, images were synthesized from three orbital segments, corresponding to extreme geometrical conditions of observation by Metis: aphelion, perihelion, and maximum solar latitude. The range of heights that can be reconstructed, the required data-gathering period, and the accuracy of the reconstruction, are discussed in detail for each case. As a general conclusion, a Metis synoptic observational program with a cadence of at least four images day−1 provides enough data to attempt tomographic reconstructions during the whole lifetime of the mission, a requirement well within the two- to three-hour cadence of the current synoptic program. This program will allow implementation of tomography experimenting with different values for the cadence of the time series of images used to feed reconstructions. Its cadence will also provide continuous opportunities to select images avoiding highly dynamic events, which compromise the accuracy of tomographic reconstructions.

Fifty years of nightly global low-light imaging satellite observations

The history of satellite remote sensing of nighttime lights traces it is roots back the U.S. Air Force Defense Meteorological Satellite Program (DMSP) which began flying polar orbiting satellites with cloud imaging sensors in the 1970’s. The meteorologists found that they could improve weather predictions with visible and thermal band imagery. The visible band channels on early DMSP satellites worked well in the daytime—but were unable to see clouds at nights. Visible band low light imaging was added in the mid-1970’s using photomultiplier tubes to amplify the signal, enabling the detection of moonlit clouds. In 1992, NOAA established a digital archive for DMSP data at the National Geophysical Data Center in Boulder, Colorado. The Earth Observation Group (EOG) was formed in 1994 and dedicated it is efforts to making global nighttime light data products for uses by the science and policy communities. Many of the shortcomings of DMSP data were addressed with the NASA/NOAA Visible Infrared Imaging Radiometer Suite (VIIRS) day/night band (DNB), which opened a new era in nighttime light studies. In total, EOG has produced 65 annual global nighttime light products and over 650 monthly products.

An Image Fusion Algorithm Based on Improved RGF and Visual Saliency Map.

To solve the artifact problem in fused images and the lack of enough generalization under different scenarios of existing fusion algorithms, the paper proposes an image fusion algorithm based on improved RGF and visual saliency map to realize fusion for infrared and visible light images and a multimode medical image. Firstly, the paper uses RGF (rolling guidance filter) and Gaussian filter to decompose the image into the base layer, interlayer, and detail layer by a different scale. Secondly, the paper obtains a visual weight map by the calculation of the source image and uses the guided filter to better guide the base layer fusion. Then, it realizes the interlayer fusion through maximum local variance and realizes the detail layer fusion through the maximum absolute value of the pixel. Finally, it obtains the fused image through weight fusion. The experiment demonstrates that the proposed method shows better comprehensive performance and obtains better results in fusion for infrared and visible light images and medical images compared to the contrast method.

Radiation Sensitivity Analysis of Ocean Wake Information Detection System Based on Visible Light Remote Sensing

Various ships and submerged moving objects in the ocean are key targets of numerous remote sensors. Wake has developed into one of the key detection targets of ocean visible light remote sensing as the visible trail information left by moving objects on the ocean surface. In the situation of slow ship speed, deep draft, and the existence of air clouds and fog, the wake target signal is weak, and the signal-to-noise ratio is low due to the low reflectivity of the sea surface and the interference of the background waves on the sea surface. This paper analyzes the radiative sensitivity of visible light imaging systems for the most crucial wake detection indicator in order to address the aforementioned issues. The noise equivalent reflectance difference, which is widely used to describe radiative sensitivity in engineering, is derived and numerically simulated by establishing the imaging link model based on TDICCD. We calculated the noise equivalent reflectivity difference for eight bands commonly used in ocean remote sensing; results show that the index is generally on the order of 10−4, and with the increase in the central wavelength, the value of noise equivalent reflectance difference also shows a downward trend and is stable within a certain value range. This research provides theoretical guidance for the engineering design of a visible spectrum imaging system for wake detection, aids in improving the imaging system’s capacity to detect weak wake signals, and provides a basis for subsequent wake detection and enhancement processing, removal of false wakes, and retrieval of ship information.

Extraction of agricultural plastic film mulching in karst fragmented arable lands based on unmanned aerial vehicle visible light remote sensing

Considering the fragmentation of karst mountainous areas, a Dajiang Innovation Mavic 2 Pro quadcopter drone was used to collect visible light images of a modern intensive agricultural planting area in response to the high labor intensity and low efficiency of manual assessment of agricultural plastic film mulching (PFM). First, the characteristics of RGB values of 10 main types of features in the sample area, such as PFM, crops, buildings, and roads, were analyzed based on the color index, geometric size, and texture characteristics. Second, based on the construction principle and form of the excess green index, we constructed the excess blue index (ExB), which comprehensively utilized the three visible light bands of red, green, and blue. By calculating the ExB of the image, the characteristics of the PFM were enhanced. Third, Gaussian High-Pass Filter (GHPF) was used for ExB images to retain high-frequency information of PFM targets and reduce the influence of noisy ground types such as plant ashes, blue-roofed buildings, asphalt roads, and water on identifying PFM. Then, combined with field measurements of PFM data, the GHPF images were segmented using Otsu and color slices. The polygon area method was used to eliminate polygons that were too large or too small. Combined with random samples from field surveys, the threshold segmented PFM polygons were verified. The accuracies of the PFM area by methods of Otsu and human–computer interaction threshold segmentation (HC-ITS) are 85.44% and 97.36%, respectively, and the average accuracies of the verification areas B-I and B-II are 90.03% and 82.32%, respectively. It is proved that the PFM ExB proposed in this study has good applicability. We find that the HC-ITS method is suitable for scenes with complex ground object types, and the Otsu method is suitable for scenes with fewer ground object types.

Visible and near-infrared dual band switchable metasurface edge imaging

Optical edge detection at the visible and near infrared (VNIR) wavelengths is deployed widely in many areas. Here we demonstrate numerically transmissive VNIR dual band edge imaging with a switchable metasurface. Tunability is enabled by using a low-loss and reversible phase-change material Sb2S3. The metasurface acts simultaneously as a high-pass spatial filter and a tunable spectral filter, giving the system the freedom to switch between two functions. In Function 1 with amorphous Sb2S3, this metasurface operates in the edge detection mode near 575 nm and blocks near infrared (NIR) transmission. In Function 2 with crystalline Sb2S3, the device images edges near 825 nm and blocks visible light images. The switchable Sb2S3 metasurfaces allow low cross talk edge imaging of a target without complicated optomechanics.

Efficient image-based soiling quantification using histograms of a checkered pattern

In this study, a model correction and a new brightness ratio calculation method were proposed to improve a previously reported checkered pattern image-based soiling quantification technique. The image’s black-white brightness ratio (BWR) was obtained for various levels of soiling and camera altitude angles, using only the histogram of the visible-light image of a checkered pattern, consequently denoted hBWR. Normal-incidence soiling loss of a PV reference cell subjected to the same soiling levels as the checkered pattern was determined using the short-circuit current method and a solar simulator. The hBWR quantity was found to be effectively identical to the BWR quantity determined from a more complex image processing method previously used. Also, a unified relationship was found between hBWR and soiling loss corrected by the sine of the camera altitude angle, for all soiling levels and all camera altitude angles. These findings show the feasibility of implementing the image-based soiling quantification technique using readily available photography equipment and generic image processing tools.

MilliNet: Applied Deep Learning Technique for Millimeter-Wave Based Object Detection and Classification

ABSTRACT Navigation and surveillance are crucial aspects of various fields such as the transportation, defence, and aeronautical industries. Imaging the terrain and the surrounding environment is a key part of a good navigation system. Some of the most widely used imaging techniques are visible light and Light Imaging Detection and Ranging (LiDAR) imaging, which covers two- and three-dimensional imaging. These imaging techniques face difficulties in adverse weather and atmospheric conditions, so there is a need for stronger radiation formation. Moreover, the data collected by radars, which emit these radiations cannot be used directly due to their disorganized characteristic. This article introduces the use of millimeter-wave imaging for the purpose of navigation and surveillance. Millimeter-wave images are captured by radar and fed to a novel neural network architecture, known as MilliNet, which is trained using LiDAR data (in point clouds) and can be used to classify and segment millimeter-wave imagery input involving features such as permutation and symmetric invariance. This provides a three-dimensional image in adverse weather conditions and conveniently uses that data for object detection and classification. The results obtained after testing the model on a dataset have been validated for different conditions and objects.

Synchrotron X-ray based particle image velocimetry to measure multiphase streamflow and densitometry

Synchrotron-based radiation techniques are increasingly used in the field of flow condition measurements. As a complicated multiphase flow, cavitation has been investigated in a wide variety of industrial fields, to prevent damages on spillways at large dams or intensify the bacteria eradication rate during water treatments. To control the detrimental effects of cavitating flows, the dynamics of and mechanisms affecting cavitation development shall be visualized and identified in real-time. Due to the large reflection and scattering effects of multiphase flow, standard visible-light imaging technologies cannot penetrate vapor bubbles and provide extremely limited information. In this study, the synchrotron source emitted high-flux X-ray pulses are used to capture flow motion and visualize internal structures with sufficiently high spatial and temporal resolutions. The image processing procedures on raw x-ray images are developed using 2D Fourier transform (FT) and wavelet transform (WT) to implement the contrast enhancement and de-noising. Through the edge detection algorithm, tracking either seeded particles or phase interfaces inside the opaque multiphase flow can be achieved to perform the particle image velocimetry. With decomposing particle-only and particle-eliminated subplots from the X-ray phase-contrast image, the time-resolved velocity and void fraction fields are obtained simultaneously, paving a way for further flow condition analysis such as densitometry.

Water photovoltaic plant contaminant identification using visible light images

The rapid advancement of photovoltaic (PV) technology offers a significant way to advance renewable energy. As a form of PV power generation that does not take up land resources to enhance the total amount of PV power generation, water PV shows excellent development prospects. However, water PV plants lack efficient monitoring solutions for PV panels. This paper proposes a method of using visible light images (RGB colour space) captured by UAVs to automatically segment PV panels and automatically classify contaminated PV panels. A dynamic threshold segmentation method combining grey space and the S component (in HSV colour space) is proposed. Then, the texture features and pixel features of the PV panels are considered for automatic detection. In the automatic classification phase, VGG16, VGG19, ResNet50, MobileNet V2, Xception and DenseNet121 networks are evaluated with transfer learning as a prerequisite. Data augmentation techniques considering texture features are used. In these networks, VGG16 obtains the best values for accuracy, recall, F1 score and balance accuracy. In particular, the model in this paper provides visual localization information and visual classification information. In the actual test phase, the recognition accuracy of the developed method for PV panels in visible light images is 96.5%. The accuracy of the classification is 86.8%, and the recall is 76.2%. These results prove the feasibility of this method and can provide guidance for the daily maintenance of water PV power plants.

Analysis of hand and face images for the purpose of engineering support for Parkinson's disease diagnosis

Engineering support in the field of recognizing Parkinson's disease against the background of other diseases, its progression and monitoring the effectiveness of drugs is currently widely implemented as part of work devoted to the use of re-cording and analysis devices equipped with sensors of movement parameters attached to the patient's body, e.g. accelerometers and gyroscopes. This material touches on an alternative approach, in which the concept of using techniques for processing selected image data obtained during a clinical examination evaluating a patient using the unified UPDRS number scale is proposed. The research was conducted on a material that corresponded to selected components of the scale and included images of faces recorded in the visible light range and images of the outer surfaces of the hand recorded with a thermal imaging camera. This was aimed at assessing the possibility of differentiating persons in terms of detecting Parkinson's disease on the basis of registered modalities. Thus, tasks aimed at developing characteristics important in the binary classification process were carried out. The assessment of features was made in a modality-dependent manner based on available tools in the field of statistics and machine learning.

<italic>Operando</italic> monitoring microstructures and defects in laser fusion additive manufacturing of metals

<p indent="0mm">Laser fusion additive manufacturing of metals has been growing tremendously over the past decade. Despite its unrivaled capability to directly fabricate complex geometries, the accuracy, repeatability, and reliability of the build remain a major concern. From the micro perspective, it is the unfavorable microstructures and defects/anomalies that deteriorate the service performances such as fatigue life and corrosion resistance. The microstructures here involve not only melt pool and keyhole but also grain morphology and size, crystal orientation, interface structure, composition homogeneity, phase stability, etc. The defects/anomalies mainly include porosity, balling, lack of fusion, spattering, element loss, oxidation, micro residual stress and deformation, microcracking, and rough surface. Accurate detection and characterization of these features are critical for their understanding and control. Traditionally, we use optical microscope, electron microscope, X-ray tomography and diffraction, and so forth, to perform postmortem analysis and track the thermal traces the local material has experienced. However, such information and evidence are typically static and limited. This poses difficulties in recreating and interpreting the microstructural formation. In this review, we summarize recent progress on<italic> in situ</italic>/<italic>operando</italic> monitoring the laser fusion additive manufacturing of metals. It mainly covers high-speed visible light and thermal imaging, laboratory X-ray imaging, ultrasonic measurement, and high-speed synchrotron X-ray imaging and diffraction. Beyond traditional postmortem characterizations, the <italic>in situ</italic>/<italic>operando</italic> techniques have demonstrated the capability to capture the transient and dynamic formation and evolution processes of various microstructures and defects. They are revolutionizing the research mode in the field of metal additive manufacturing, from trial and error to mechanism oriented. In the former, we continue to optimize the objective function (e.g., grain morphology and size) until it arrives at a satisfactory solution, largely based on limited experience and knowledge and costly iterations. In the latter, we quest for the ultimate solution to manufacturing defect-free and perfect structures through direct exploration of the physical origins of the aforementioned features. Among the various techniques, we address high-speed synchrotron X-ray imaging and diffraction. It is an emerging approach in this field and is more suitable to monitor the internal microstructural evolution of metals, e.g., melt pool and keyhole dynamics, pore formation, cracking, rapid solidification, and phase transformation, because of the micrometer spatial resolution, sub-nanosecond temporal resolution, megahertz frame rate, and millimeter penetration capacity. Nevertheless, other techniques like high-speed visible light and thermal imaging have irreplaceable advantages and application scenarios. They can provide complementary information (e.g., thermal distribution on the melt pool surface, vapor plume dynamics, and three-dimensional spatter morphologies) for comprehensive understanding of the fundamental mechanisms in laser melting. Looking ahead to the next decade, the integration of multiple <italic>in situ</italic>/<italic>operando</italic> techniques and the translation among the resultant multi-dimensional signals, as well as big data computing and high precision modeling, will bridge understanding of mechanisms on the microscale and tailoring of properties on the macroscale. This will promote sustainable growth of the additive manufacturing industry.