Light Environment Research Articles

Behavior recognition algorithm based on a dual-stream residual convolutional neural network

Abstract In the process of behavior recognition, the recognition operation may be carried out in various environments such as sunny, cloudy, and night. Since traditional recognition algorithms are judged by identifying the pixels of the image, the intensity of the light will affect the image. The brightness and contrast of the display thus interfere with the recognition results. Therefore, traditional algorithms are easily affected by the lighting environment around the recognition object. To improve the accuracy and recognition rate of the behavior recognition algorithm in different lighting environments, a convolutional neural network (CNN) algorithm using a dual-stream method of time flow and spatial flow is studied here. First, we collect behavioral action data sets and preprocess the data. The core of the behavior recognition algorithm of the dual-stream residual CNN is to use the time stream and the spatial stream to fuse behavioral features and eliminate meaningless data features. After processing Perform feature selection on the data, select the acoustic wave and light-sensing features of the data, and finally, use the extracted features to classify and identify using the two-stream residual CNN and the traditional behavior recognition method. The behavior recognition algorithm based on the dual-stream residual CNN was tested on the data of four groups of people. For the behavioral feature map with a data volume of 50, the behavior recognition algorithm of the dual-stream residual CNN was effective in various environments under different lighting conditions. The recognition accuracy can reach 83.5%, which is 12.3% higher than the traditional. The behavior recognition algorithm of the dual-stream residual CNN takes 17.25 s less than the conventional recognition algorithm. It is concluded that behavior recognition based on dual-stream residual CNNs can indeed improve the recognition accuracy and recognition speed in environments with different lighting conditions than traditional behavior recognition.

Qualitative Mechanisms of Perceived Indoor Environmental Quality on Anxiety Symptoms in University

The indoor environment is widely acknowledged as a non-pharmacological tool for regulating residents’ mental health. In dormitory environments with relatively high residential density, the mental health of university students requires particular attention. This study surveyed 445 students from a northern Chinese university and used structural equation modeling (SEM) to analyze the impact of perceived indoor environmental quality (IEQ)—including thermal, lighting, acoustics, indoor air quality, and overcrowding—on self-reported anxiety symptoms. The results indicated the following: (1) students’ perceptions of dormitory IEQ significantly affected anxiety symptoms, explaining 40% of the variance; (2) anxiety symptoms associated with the IEQ were mainly characterized by anxiety and panic (r = 0.91, p < 0.001); (3) subjective perceptions of the acoustic environment (r = −0.55, p < 0.001) and indoor air quality (r = −0.15, p < 0.05) were key predictors of anxiety, while thermal environment, lighting environment, and overcrowding were not significant. The findings enrich the IEQ system and provide directions for optimizing the dormitory indoor environment from the perspective of student mental health, with implications for other types of residential buildings.

White Light Generation and Stability Analysis of High-Power Blue LDs with Remote YAG Phosphors

This paper presents a comprehensive analysis of white light generation and the associated aging dynamics using high-power blue laser diodes (LDs) combined with transmissive single crystal remote YAG phosphors. By systematically varying input currents (ranging from 0.6 A to 3 A) and phosphor thicknesses (250 μm and 500 μm), this study elucidates the optical and electrical characteristics of LD-phosphor systems under diverse operating conditions. The results highlight the system’s potential for stable and efficient white light generation, making it suitable for high-power lighting applications. Experimental setups included both single LDs and a 4 × 2 LD array. For the single LD, a peak optical output of 4.16 W was achieved at 3 A, corresponding to an initial luminous flux of approximately 700 Lm and a correlated color temperature (CCT) of 4653 K, with minimal color temperature shift observed during a 60 min aging process. The 4 × 2 LD array demonstrated consistent white light output across varying phosphor thicknesses, with maximum luminous fluxes of 1857 Lm at 1.4 A and 2622 Lm at 1.6 A for phosphor thicknesses of 250 μm and 500 μm, respectively. Importantly, the phosphor exhibited excellent thermal stability throughout the aging process, with the CCT maintained within a range of 4600 K to 5500 K. These findings underscore the reliability and applicability of LD-based white light systems in demanding high-power lighting environments, offering a promising alternative to conventional light sources for automotive, industrial, and specialized lighting applications.

LEAFusion: An Infrared and Visible Light Image Fusion Network Resilient to Harsh Light Environment Interference Based on Harsh Light Environment Aware

LEAFusion: An Infrared and Visible Light Image Fusion Network Resilient to Harsh Light Environment Interference Based on Harsh Light Environment Aware

Evolution of Large Eyes in Stromboidea (Gastropoda): Impact of Photic Environment and Life History Traits.

Eyes within the marine gastropod superfamily Stromboidea range widely in size, from 0.2 to 2.3mm - the largest eyes known in any gastropod. Despite this interesting variation, the underlying evolutionary pressures remain unknown. Here, we use the wealth of material available in museum collections to explore the evolution of stromboid eye size and structure. Our results suggest that depth is a key light-limiting factor in stromboid eye evolution; here, increasing water depth is correlated with increasing aperture width relative to lens diameter, and therefore an increasing investment in sensitivity in dim light environments. In the major clade containing all large-eyed stromboid families, species observed active during the day and the night had wider eye apertures relative to lens sizes than species observed active during the day only, thereby prioritising sensitivity over resolution. Species with no consistent diel activity pattern also had smaller body sizes than exclusively day-active species, which may suggest that smaller animals are more vulnerable to shell-crushing predators, and avoid the higher predation pressure experienced by animals active during the day. Within the same major clade, ancestral state reconstruction suggests that absolute eye size increased above 1mm twice. The unresolved position of Varicospira, however, weakens this hypothesis and further work with additional markers is needed to confirm this result.

Simulation of light environments inside lunar holes and their associated subsurface caverns

Vertical holes of several tens of meters in diameter and depth have been discovered on the Moon, potentially serving as skylights into subsurface volcanic caverns resembling lava tubes. Because of their scientific importance and potential for use as future lunar bases, these lunar holes and caverns are expected to be targets of intensive exploration in the near future. Using numerical simulations, this study investigates the light environment within a directly and/or indirectly sunlit subsurface cavern accessed through a skylight hole. We specifically analyze the Mare Tranquillitatis Hole (MTH), one of the largest lunar vertical holes, situated near the Moon's nearside meridian. The floor and walls of the hole, and the walls and ceiling of its associated subsurface cavern are lit by reflected sunlight from other parts of the hole and cavern, such as the floors, depending on the solar elevation angle. Furthermore, this paper presents estimation results for camera imaging for a case in which solar elevation angle is 45° in the morning, which is appropriate timing for exploration. Assuming reflectance of 0.1 for the lunar hole and cavern surfaces, we estimate the incident energy onto each pixel of a camera as it descends to the hole's floor. We specifically simulate conditions for a camera with a 12-bit dynamic range, similar to the wide-angle optical navigation camera (ONC-W) onboard the Hayabusa 2 spacecraft. The results suggest that a camera with a fixed gain would struggle to capture both directly and indirectly sunlit areas simultaneously, without saturating bright areas and ensuring minimum incident energy resolution (say, a 10 digital number) for darker areas. To overcome this challenge, adjusting the camera gain based on the hole's and cavern's illumination conditions is necessary.Graphical abstract

Optimizing light environments in aquatics center to enhance the performance of Olympic champions and other elite swimmers: an experimental study

Several major international sports events in the next five years require new sports venues, with lighting environments being a crucial factor. Typically, the lighting systems in sports venues are primarily for illumination. However, since light has been indicated to affect human psychology and physiology, we investigated whether the light environments in sports buildings impact athletic performance. Additionally, swimming is the indoor sport with the most gold medals in the Olympics. Therefore, we conducted a within-subject, randomized crossover study in an aquatics center with fourteen elite swimmers, including the Olympic champions and Asian champions (6 female and 8 male; 19.3 ± 3.4 years of age). During the experiment, the elite swimmers were exposed to either a common correlated color temperature (CCT) light environment (controlled condition: 5780 K) or an experimental light environment (8512 K) in the aquatics center, with both having the same illuminance (1020 lx). The neurobehavioral and swimming test results showed that the Olympic champions and other elite swimmers had significantly faster responses, fewer lapses, greater arousal, less visual fatigue, quicker reactions at start, and higher swimming speeds in the experimental condition compared to the controlled condition (p < 0.05). Our findings suggest that the light environment in the aquatics center affects swimmers' non-visual performance and enhancing the CCT of the light environment could improve swimmers' performance. We propose the concepts of Sports Lighting Beyond Illumination and Sports Human-Centric Lighting to enhance athletes’ health and performance, and to improve the sports environment for a better experience for all participants.

Discussion on effect of observation circumstances for image in digital radiography

The discriminability of the IQI in the film radiography is affected by the observation circumstance and the luminance of film viewer. ISO17636-2 states that the images of digital radiograph shall be examined in a dimmed room. However, it does not describe requirements for the observation environment in digital radiography, despite the fact that the environment is similar to that of a normal professional PC and the display of the images can be easily adjusted. In this report, the influence of the observation environment in digital radiography was compared as the parameters the wire diameter and the WL value of a display adjustment value when the minimum wire diameter of the image quality indicator (IQI) was percepted, using images of welded steel at t25mm and t50mm. For welded steel of t25mm, differences due to SNRN values and differences due to the observation environment of light and dark room were compared. For t50mm, differences due to the presence or absence of the observation mask for the single wire of wire IQI which is places in front of the display, differences due to the resolution and brightness of the PC monitor, and differences when the WL was adjusted arbitrarily or fixed were compared. For the Digital Radiographic testing (RT-D), the basic spatial resolutions (SRb) were obtained by measuring the profile and applying of visual observation with the duplex wire type IQI. On the other hand, when observing the image of wire type IQI, the observation ways that one was obtain the perceptible image for all wires and another was to get the perceptible single wire image by using the observation mask for watching each wire were applied. Moreover, the influence of window width (WW) and window level (WL) adjustment on the discrimination of wire-type IQI was examined by quantifying it as WW/WL. As a result, it was confirmed that the discriminality in the dimmed room and the adjustment of WW and WL were effective. In addition, using the observation a mask showed that the perception of the adjacent wire image affected observer cognition in IQI discrimination during image observation. When using viewer software profiling, SRb measurements were not affected by the observation environment, but the visual observation was affected by the environment, and when observing images in the dimmed room with RT-D, it was found that adjusting WW and WL had a greater effect on discrimination for images with low SNRN and low contrast.

Photoinhibition Sensitivity of Abies sachalinensis Saplings Under Different Leaf Mass per Area Conditions: Insights into Acclimation to Current and Future Light Environments

Photoinhibition, a common physiological obstacle in forest regeneration, results from rapid light elevations after removing upper-layer trees, thereby impeding sapling growth and survival. The leaf mass per area (LMA), which reflects the current light environment, may predict saplings’ responses to future elevations in light intensity. However, the relationship between the LMA and photoinhibition sensitivity in boreal forest evergreen conifers remains unclear. In this study, we explored whether the LMA was related to photoinhibition sensitivity in A. sachalinensis saplings. Saplings with a high LMA exhibited higher excitation pressure (1−qP) under the current light environment, leading to increased stress even at low relative light intensities of 3%–17%. A model illustrating the dependence of photosynthetic physiological parameters on the photosynthetic photon flux density (PPFD) revealed that 1−qP significantly decreased with an increasing LMA, while the electron transfer rate (ETR) and non-photochemical quenching (NPQ) significantly increased. Additionally, regardless of the LMA, 1−qP reached near saturation (≈1.0) at a PPFD of 1000 μmol m−2 s−1, likely owing to the inefficient consumption of excess energy by electron transfer, as indicated by the low maximum ETR (approximately 25 μmol m−2 s−1). These findings suggest that although A. sachalinensis exhibits high sensitivity to photoinhibition, a high LMA may reflect physiological acclimation to forthcoming elevations in light exposure, thereby reducing the susceptibility to photoinhibition at present and in the future.

Crystallization Control to Prepare Uniform CsPbI2Br Thin Films for High-Efficiency Perovskite Solar Cells.

All-inorganic perovskite solar cells (PSCs) face challenges related to film inhomogeneity, which arises from nonideal crystallization. This issue significantly impedes the advancement of all-inorganic PSCs. In this study, we observed that during the crystallization process of CsPbI2Br, the Br-rich intermediate phase is often preferentially deposited, leading to an uneven distribution in the out-of-plane direction of the perovskite film. To address this issue, crown ether molecules (dibenzo-18-crown-6) were introduced into the perovskite precursor solution. The complexation of crown ether with Cs cations and Br anions optimizes the crystallization sequence of the perovskite, ensuring that the intermediate phase closely conforms to the standard stoichiometric ratio. This adjustment significantly mitigates the problem of uneven halogen ion distribution along the out-of-plane direction. Furthermore, the crown ether thermally decomposes during the high-temperature annealing process, thereby not affecting the composition of the final perovskite film. Following crown ether treatment, the efficiency of the PSCs reached 14.08%, and the unpackaged devices maintained 80% of their initial efficiency after 1000 h of exposure to light in an atmospheric environment.

The Impact of Varying Ratios of Light Quality Emitted by Light-Emitting Diodes (LEDs) on the Butter Lettuce

Light-emitting diodes (LEDs) are considered the optimal artificial light source for plant factories, yet further research is needed to understand the impact of LED light quality on plant growth. This experiment investigated the growth and development of butter lettuce under varying ratios of red and blue light provided by LEDs. The specific effects of different red and blue light quality ratios on butter lettuce growth were systematically tracked and recorded, with comprehensive measurements of growth parameters and photosynthetic characteristics. Empirical findings revealed that a higher red light component compared to blue light led to improved butter lettuce characteristics. These included increased leaf length and width, elevated plant height, higher relative chlorophyll content, and an increased net photosynthetic rate. The study also highlighted that the light quality requirements of lettuce fluctuate across different growth stages, allowing for adaptable modulation of lighting conditions to align with specific growth stage needs.Total fresh weight, total dry weight and dry matter content of lettuce reached the highest values in the R8B2 treatment with 80% red light and 20% blue light, and the percentage differences reached 29.2%, 23.1% and 13.9%, respectively. Additionally, the photosynthetic chlorophyll content and net photosynthetic rate exhibited maximum percentage differences of 22.22% and 14.6%, respectively. In conclusion, the experimental evidence supports that an R8B2 spectral composition (80% red and 20% blue light) in a controlled indoor setting represents the optimal light environment for the accelerated growth and superior quality of butter lettuce.

Visual fatigue prediction using classification model based on physiological responses of occupants under office lightings

Appropriate management of visual fatigue is crucial for eye health as it significantly impacts life quality throughout an individual's life cycle. The shift to digital tasks in modern office environments has increased occupants' exposure to visual fatigue, leading to various social problems. This study aimed to develop visual fatigue prediction models based on physiological responses and classification algorithms. Experiments were conducted to collect physiological responses and subjective visual fatigue under various lighting environments. Collected data was refined and reorganized as predictor variables by different time windows and target variables by scales. Then, visual fatigue prediction models were developed using supervised machine learning algorithms (i.e., artificial neural network, support vector machine, gradient boosting machine, and random forest). And improved by feature selection and adding subject-label variables. The resulting two-scale and three-scale visual fatigue prediction models demonstrated an average performance of 93.73 % and 94.64 % respectively. This research can contribute to reducing societal costs and enhancing productivity by proposing visual fatigue prediction models that help manage eye health in office environments.

The Adversarial Robust and Generalizable Stereo Matching for Infrared Binocular Based on Deep Learning

Despite the considerable success of deep learning methods in stereo matching for binocular images, the generalizability and robustness of these algorithms, particularly under challenging conditions such as occlusions or degraded infrared textures, remain uncertain. This paper presents a novel deep-learning-based depth optimization method that obviates the need for large infrared image datasets and adapts seamlessly to any specific infrared camera. Moreover, this adaptability extends to standard binocular images, allowing the method to work effectively on both infrared and visible light stereo images. We further investigate the role of infrared textures in a deep learning framework, demonstrating their continued utility for stereo matching even in complex lighting environments. To compute the matching cost volume, we apply the multi-scale census transform to the input stereo images. A stacked sand leak subnetwork is subsequently employed to address the matching task. Our approach substantially improves adversarial robustness while maintaining accuracy on comparison with state-of-the-art methods which decrease nearly a half in EPE for quantitative results on widely used autonomous driving datasets. Furthermore, the proposed method exhibits superior generalization capabilities, transitioning from simulated datasets to real-world datasets without the need for fine-tuning.

Efficient Environment Map Rendering Based on Decomposition

AbstractThis paper presents an efficient environment map sampling algorithm designed to render high‐quality, low‐noise images with only a few light samples, making it ideal for real‐time applications. We observe that bright pixels in the environment map produce high‐frequency shading effects, such as sharp shadows and shading, while the rest influence the overall tone of the scene. Building on this insight, our approach differs from existing techniques by categorizing the pixels in an environment map into emissive and non‐emissive regions and developing specialized algorithms tailored to the distinct properties of each region. By decomposing the environment lighting, we ensure that light sources are deposited on bright pixels, leading to more accurate shadows and specular highlights. Additionally, this strategy allows us to exploit the smoothness in the low‐frequency component by rendering a smaller image with more lights, thereby enhancing shading accuracy. Extensive experiments demonstrate that our method significantly reduces shadow artefacts and image noise compared to previous techniques, while also achieving lower numerical errors across a range of illumination types, particularly under limited sample conditions.

Multi-Camera Lighting Estimation for Mobile Augmented Reality

Lighting estimation is an important and long-standing task in computer vision and graphics communities[1]. Over the past few years, we have observed a rising demand for creating immersive user experiences in mobile augmented reality (AR) applications. In particular, environment lighting estimation is essential to render visually coherent virtual objects that need to be overlaid on physical scenes in mobile AR. In this work, we explore the key research questions in lighting understanding for an emerging application domain, mobile AR on multi-camera handheld devices, particularly try-on applications where end users leverage handheld mobile devices, such as smartphones, to overlay products of interest on their faces. Mobile AR try-on apps promote online shopping and often require photorealism to provide user experiences on par with physical try-on.

Research on Target Image Classification in Low-Light Night Vision.

In extremely dark conditions, low-light imaging may offer spectators a rich visual experience, which is important for both military and civic applications. However, the images taken in ultra-micro light environments usually have inherent defects such as extremely low brightness and contrast, a high noise level, and serious loss of scene details and colors, which leads to great challenges in the research of low-light image and object detection and classification. The low-light night vision image used as the study object in this work has an excessively dim overall picture and very little information about the screen's features. Three algorithms, HE, AHE, and CLAHE, were used to enhance and highlight the image. The effectiveness of these image enhancement methods is evaluated using metrics such as the peak signal-to-noise ratio and mean square error, and CLAHE was selected after comparison. The target image includes vehicles, people, license plates, and objects. The gray-level co-occurrence matrix (GLCM) was used to extract the texture features of the enhanced images, and the extracted image texture features were used as input to construct a backpropagation (BP) neural network classification model. Then, low-light image classification models were developed based on VGG16 and ResNet50 convolutional neural networks combined with low-light image enhancement algorithms. The experimental results show that the overall classification accuracy of the VGG16 convolutional neural network model is 92.1%. Compared with the BP and ResNet50 neural network models, the classification accuracy was increased by 4.5% and 2.3%, respectively, demonstrating its effectiveness in classifying low-light night vision targets.

Effects of light quality on agronomic traits, antioxidant capacity and nutritional composition of Sarcomyxa Edulis

Sarcomyxa edulis, a notable edible and medicinal mushroom indigenous to northeast China, is celebrated for its high nutritional value and delightful taste. In this study, white light as a control and examined the effects of red, green, and blue light on the agronomic traits, antioxidant capabilities, and nutritional composition of S. edulis. The results showed that different monochromatic light qualities affected the color of S. edulis pileus, with blue light demonstrating particular efficacy. Furthermore, blue light also regulated pileus length, whereas red light was instrumental in significantly increasing stalk length. Regarding antioxidant capacity, compared with red and green light, the activities of POD, SOD, and CAT were significantly improved by blue light irradiation, decreased MDA levels, and improved free radical scavenging potential. Additionally, blue light exposure led to an increase in the contents of 15 amino acids. Green light treatment reduced the crude fat content. For the first time, we found that light quality is a key factor in controlling the color of S. edulis. Blue light is an effective way to regulate the color and pileus size of S. edulis, and improve the antioxidant properties. The photobiological characteristics and the response of nutritional quality to light environment of S. edulis were clarified.

Semantic Segmentation Model-Based Boundary Line Recognition Method for Wheat Harvesting

The wheat harvesting boundary line is vital reference information for the path tracking of an autonomously driving combine harvester. However, unfavorable factors, such as a complex light environment, tree shade, weeds, and wheat stubble color interference in the field, make it challenging to identify the wheat harvest boundary line accurately and quickly. Therefore, this paper proposes a harvest boundary line recognition model for wheat harvesting based on the MV3_DeepLabV3+ network framework, which can quickly and accurately complete the identification in complex environments. The model uses the lightweight MobileNetV3_Large as the backbone network and the LeakyReLU activation function to avoid the neural death problem. Depth-separable convolution is introduced into Atrous Spatial Pyramid Pooling (ASPP) to reduce the complexity of network parameters. The cubic B-spline curve-fitting method extracts the wheat harvesting boundary line. A prototype harvester for wheat harvesting boundary recognition was built, and field tests were conducted. The test results show that the wheat harvest boundary line recognition model proposed in this paper achieves a segmentation accuracy of 98.04% for unharvested wheat regions in complex environments, with an IoU of 95.02%. When the combine harvester travels at 0~1.5 m/s, the normal speed for operation, the average processing time and pixel error for a single image are 0.15 s and 7.3 pixels, respectively. This method could achieve high recognition accuracy and fast recognition speed. This paper provides a practical reference for the autonomous harvesting operation of a combine harvester.

Generation and Clearance of Superoxide Radicals at Buried Interfaces of Perovskites with Different Crystal Structures.

One of the most notorious issues with classic perovskite (MAPbI3) is its rapid degradation caused by generating superoxide radicals (O2 •-) on its surface under light and oxygen environments (light/O2). The differences in O2 •- generation rate and tolerance to O2 •- among perovskite with different structures are pending. For the first time it is validated through solid-electron paramagnetic resonance (EPR) that MAPbI3 and Cs0.175FA0.75MA0.075PbI3 (PVSK) crystals can generate O2 •- in an air atmosphere. The rapid degradation of perovskite buried interfaces caused by O2 •- dominates the nonexposed air aging process of SnO2-based perovskite film, and the degradation rate of MAPbI3 film is faster than that of PVSK film. The fullerene pyridine derivatives (C60OPD), which function as a buffer layer between SnO2 and PVSK to scavenge O2 •- and prevent degradation at the buried interface of the PVSK film, reduce the density of defect states, and accelerate the transmission of photogenerated electrons. The photoelectric conversion efficiency (PCE) of perovskite solar cells (PSCs) optimizes with C60OPD increased from 21.15% to 23.11% while significantly improving the stability in light/O2. This work reveals the hidden degradation of perovskite-buried interfaces caused by O2 •- and explores efficient ways for perovskite to resist O2 •-.

Railway foreign object tracking and detection with spatial positioning and feature generalization enhancement

The existing deep learning foreign object tracking and detection algorithm is easily affected by complex environments and target occlusion, resulting in problems such as missed detection and low detection accuracy. A railway foreign object tracking and detection algorithm with spatial positioning and feature generalization enhancement is proposed. First, a multi-scale cascade improved GhostNet feature network is proposed to improve the feature extraction capability of infrared targets. Secondly, the spatial positioning and feature generalization enhancement module is designed by using the spatial position positioning and generalized morphological information of foreign objects to enhance the detection accuracy of targets with position movement and tracking trajectory changesin complex scenes. Then, a pyramid prediction network is constructed to obtain the detection anchor frame, type and confidence information of infrared railway foreign objects. Finally, by improving the DeepSORT tracking algorithm with improved category and confidence display, combined with Kalman filtering and Hungarian algorithm, railway foreign object tracking and detection in infrared weak light environment is realized. Experimental results show that the proposed method has a tracking and detection accuracy of railway foreign objects 83.3%, with an average detection rate of 11.3 frames per second. Compared with the comparison method, the proposed method has higher detection accuracy and has better performance in railway foreign object tracking and detection in infrared weak light scenes.