Light Variations Research Articles

An Improved UNet-Based Path Recognition Method in Low-Light Environments

The fruit industry is a significant economic sector in China, with modern orchards gradually transitioning to trellis orchards. For mechanized orchard operations, automatic vehicle navigation is essential. However, in trellis orchards, the shading from trees results in low average light intensity and large variations in lighting, posing challenges for path navigation. To address this, a path navigation algorithm for trellis orchards is proposed based on the UNet-CBAM model. The network structures of UNet, FCN, and SegNet are compared to identify and select the optimal structure for further improvement. Among the three attention mechanisms of channel attention, spatial attention, and combined attention, the most effective mechanism is identified. The optimal attention mechanism is incorporated into the optimized network to enhance the model’s ability to detect path edges and improve detection performance. To validate the effectiveness and generalizability of the model, a total of 400 images were collected under varying lighting intensities. The experimental results show that this method achieves an accuracy of 97.63%, a recall of 93.94%, and an Intersection over Union (IoU) of 92.19%. These results significantly enhance path recognition accuracy in trellis orchards, particularly under low light under conditions. Compared to the FCN and SegNet algorithms, this method provides higher detection accuracy and offers a new theoretical foundation and research approach for path recognition in low-light environments.

Smart modeling of soil-foundation interaction using coupled mechanisms: a numerical framework for liquefaction risk mitigation

This study investigates the impact of nearby structures on the cyclic settlement mechanisms of shallow foundations in liquefiable soils using a numerical model based on Biot’s porous media theory. The model predicts excess pore water pressure and settlement by coupling equilibrium and continuity equations, solved using an implicit time integration scheme. Soil nonlinearity under cyclic loading is represented using generalized plasticity, boundary surfaces, and non-associated models. Three scenarios are simulated to study the effect of spacing between light and heavy foundations and variation in acceleration intensity. Results show that as spacing between foundations increases, lateral displacement and settlement decrease. Excess pore water pressure generation also decreases with increased foundation spacing. Soil just below the foundation exhibits maximum settlement, decreasing with depth. When input acceleration increases from 0.1 g to 0.15 g and 0.2 g, settlement increases by 40%–55% and 90%–110% respectively for both light and heavy foundations, regardless of spacing. Excess pore water pressure also increases sharply with higher acceleration intensity. The findings highlight the importance of considering foundation-soil-foundation interaction effects in liquefaction-prone urban settings and provide insights for designing resilient shallow foundations. The advanced numerical modeling approach offers engineers a more informed way to mitigate liquefaction risk and build safer, more durable structures in earthquake-prone areas.

Unique structural attributes of the PSI-NDH supercomplex in Physcomitrium patens.

Cyclic electron transport around photosystem I (PSI) is essential for the protection of the photosynthetic apparatus in plants under diverse light conditions. This process is primarily mediated by Proton Gradient Regulation 5 protein/Proton Gradient Regulation 5-like photosynthetic phenotype 1 protein (PGR5/PGRL1) and NADH dehydrogenase-like complex (NDH). In angiosperms, NDH interacts with two PSI complexes through distinct monomeric antennae, LHCA5 and LHCA6, which is crucial for its higher stability under variable light conditions. This interaction represents an advanced evolutionary stage and offers limited insight into the origin of the PSI-NDH supercomplex in evolutionarily older organisms. In contrast, the moss Physcomitrium patens (Pp), which retains the lhca5 gene but lacks the lhca6, offers a glimpse into an earlier evolutionary stage of the PSI-NDH supercomplex. Here we present structural evidence of the Pp PSI-NDH supercomplex formation by single particle electron microscopy, demonstrating the unique ability of Pp to bind a single PSI in two different configurations. One configuration closely resembles the angiosperm model, whereas the other exhibits a novel PSI orientation, rotated clockwise. This structural flexibility in Pp is presumably enabled by the variable incorporation of LHCA5 within PSI and is indicative of an early evolutionary adaptation that allowed for greater diversity at the PSI-NDH interface. Our findings suggest that this variability was reduced as the structural complexity of the NDH complex increased in vascular plants, primarily angiosperms. This study not only clarifies the evolutionary development of PSI-NDH supercomplexes but also highlights the dynamic nature of the adaptive mechanisms of plant photosynthesis.

Probing ozone effects on European hornbeam (Carpinus betulus L. and Ostrya carpinifolia Scop.) leaf water content through THz imaging and dynamic stomatal response

We investigated the impact of ozone exposure on Hornbeam using a novel dual approach based on Terahertz (THz) imaging in a free-air ozone exposure experiment (three ozone levels: ambient; 1.5 times ambient; twice ambient). The research aims at unraveling the physiological responses induced by elevated ozone levels on water dynamics. THz imaging unveiled dynamic changes in leaf water content, providing a non-invasive approach to leaf water monitoring. Leaf gas exchange measurements assessed stomatal responses to light variation. Our findings showcase a compelling correlation between elevated ozone levels and reduction in photosynthetic rate and impairment of stomatal function, i.e. “stomatal sluggishness”, indicative of nuanced regulatory mechanism. Stomatal sluggishness was particularly evident in Carpinus betulus (CB) compared to Ostrya carpinifolia (OC) and was linked to reduction in photosynthetic capacity. THz-based imaging techniques confirmed this result indicating a negative effect of O3 on leaf-level total water content. In addition, spatial analysis of leaf water status using these techniques also highlighted that the negative effect of O3 on water status was progressing even in less sensitive OC plants though visible foliar injury was not detected. In fact, OC showed a relative dry area of 1.6 ± 1.6 % in the control group and 3.8 ± 1.3 % under high ozone levels. THz-based imaging techniques provided a deep understanding of O3 behavior in plants and may be recommended for precision biosensing in the early detection of O3-induced damage. The integration of THz imaging and physiological analysis resulted in comprehensive understanding of Hornbeam acclimation response to ozone exposure.

Tailhook Recognition for Carrier-Based Aircraft Based on YOLO with Bi-Level Routing Attention

To address the problems of missed and false detections caused by target occlusion and lighting variations, this paper proposes a recognition model based on YOLOv5 with bi-level routing attention to achieve precise real-time small object recognition, using the problem of tailhook recognition for carrier-based aircraft as a representative application. Firstly, a module called D_C3, which combines deformable convolution, was integrated into the backbone network to enhance the model’s learning ability and adaptability in specific scenes. Secondly, a bi-level routing attention mechanism was employed to dynamically focus on the regions of the feature map that are more likely to contain the target, leading to more accurate target localization and classification. Additionally, the loss function was optimized to accelerate the bounding box regression process. The experimental results on the self-constructed CATHR-DET and the public VOC dataset demonstrate that the proposed method outperforms the baselines in overall performance.

KIC 8840638: A Newly Discovered Eclipsing Binary with δ Scuti–Type Oscillations

In this paper, we analyze the light variation of KIC 8840638 using high-precision time-series data from the Kepler mission. The analysis reveals that this target is a new detached eclipsing binary system with a δ Scuti component, rather than a single δ Scuti star as previously known. The frequency analysis of short-cadence data reveals 95 significant frequencies, most of which lie in a frequency range of 23−32 day−1. Among them, seven independent frequencies are detected in the typical frequency range of δ Scuti stars, and they are identified as pressure modes. In addition, a possible large separation value of Δν = 36.5 ± 0.1 μHz is also detected with the Fourier transform (FT) and autocorrelation function (AC) analysis. The orbital frequency f orb (= 0.320008 day−1) and its harmonics are also detected directly in the frequency spectrum. The binary modelings derived from PHOEBE indicate that this binary system is in detached configurations with a mass ratio of q=0.33−0.04+0.06 , an inclination angle of 40.19−2.84+3.96 ​​​​​°. The derived parameters and binary evolutionary model suggest that the primary star is an object on the verge of leaving the main sequence with a temperature of ∼7600 K, while the secondary appears to be a cool component entering the giant branch with a temperature ∼3100 K lower than the primary. Moreover, this system may have undergone a mass ratio reversal, where the more massive star is the gainer component and the less massive one is the donor star.

Analysis the State-of-art Exoplanets Exploring Schemes: Radial Velocity, Transit and Gravitational Microlensing

Since the first planet beyond the solar system managed to be examined in 1995, the concept of “exoplanet” has been created. The exoplanet, officially defined as extrasolar planets circling other stars that are not part of the solar system, quickly arose an influential curiosity, as scientists began to consider how many stars and planets truly exist in the periphery of the solar system and what the properties of these objects have. In the following years, a series of detection schemes appeared and applied to this theme. This study has listed five of them that are utilized most in current decades, providing a specific view of the Radial velocity method, Transit method, and Gravitational microlensing method. Each of these three explores exoplanets depends on different theories, like Doppler shifts of the RV method, the light variations of the transit method caused by the shadow of planets, and the subtle peaks of luminance of the microlensing method. By outlining the information of several instruments, the goal is to obtain a comprehensive understanding and concise synopsis of the methods employed in the exoplanetary area, which may provide with fresh inspiration to create more sophisticated appliances by understanding the underlying principles of the ones that already exist.

Recognizing objects in complex scenes: A Recent Systematic Review

This systematic review provides a comprehensive examination of recent advancements in object recognition within complex scenes, focusing on addressing challenges such as clutter, occlusion, and diverse environmental conditions. Leveraging the PRISMA framework, the study meticulously analysed a diverse range of literature from esteemed sources, employing advanced search methods on Scopus and WoS databases to identify and analyse primary research studies (n = 25). The review encompasses three key themes: Theme 1 concentrates on Image Noise Identification and Removal Techniques, while Theme 2 delves into Image Classification and Recognition under Noise. Additionally, Theme 3 explores Innovative Models and Approaches for Noise-Robust Image Analysis. Despite the progress achieved, contemporary recognition systems struggle with real-world complexities such as varied scales, lighting variations, and different viewpoints. The synthesis of findings emphasizes the necessity for innovative strategies that capitalize on contextual cues and harness the potential of deep learning to enhance precision in object recognition within intricate visual environments. The insights gleaned from this synthesis are poised to guide future research directions, informing the development of more resilient algorithms capable of navigating challenges and catalysing advancements in the field of object recognition.

Morphophysiological variation, vigor test, and influence of light and temperature on the germination of four cultivars of Glycine max (L.) Merrill

Soybean (Glycine max. (L.) Merrill) is a crop whose productivity depends on the quality of the seeds used. In this context, the objective of this study was to characterize the seeds physically, evaluate viability using tetrazolium, identify physiological vigor through an electrical conductivity test, and determine the influence of temperature and photoperiod on the germination of four cultivars (Ultra, Olimpo, Extrema and Foco). 200 seeds were used, measuring size, weight of a thousand seeds, moisture content, and fresh mass. In tetrazolium (1%), 100 seeds were soaked for 4 h at 25°C. For electrical conductivity, 50 seeds were submerged on 75 mL of deionized water at 25°C for periods of 4, 8, 12, 16, 20 and 24 h. In the germination process, the treatments (four replications of 50 seeds) were incubated (B.O.D.) at temperatures of 25°C, 27°C, 30°C and 35°C, under photoperiods of 12 and 14 h. The experimental design adopted was completely randomized for all experiments, and the means were subjected to analysis of variance and Tukey's test at 5% probability. The Ultra cultivar presented perimeter (38.21±3.72 mm), length of 9.98±0.24 mm, and width of 7.94±0.27 mm. The cultivars presented viability above 90%. For electrical conductivity, the 12 h period was effective for identifying vigor. The Ultra cultivar (25°C and 12 h of photoperiod) achieved 77% germination, 10 germinated seeds per day, an average germination time of 4 days and 43 germinated seeds by the third day.

Mass-transfer Outbursts Reborn: Modeling the Light Curve of the Dwarf Nova EX Draconis

EX Draconis is an eclipsing dwarf nova that shows outbursts with moderate amplitude (≃2 mag) and a recurrence timescale of ≃20–30 days. Dwarf novae outbursts are explained in terms of either a thermal-viscous instability in the disc or an instability in the mass-transfer rate of the donor star (MTIM). We developed simulations of the response of accretion discs to events of enhanced mass transfer, in the context of the MTIM, and applied them to model the light curve and variations in the radius of the EX Dra disc throughout the outburst. We obtain the first modeling of a dwarf nova outburst by using χ 2 to select, from a grid of simulations, the best-fit parameters to the observed EX Dra outbursts. The observed time evolution of the system brightness and the changes in the radius of the outer disc along the outburst cycle are satisfactorily reproduced by a model of the response of an accretion disc with a viscosity parameter α = 4.0 and a quiescent mass-transfer rate Ṁ2(quiescence)=4.0×1016 g s−1 to an event of width Δt = 6.0 × 105 s (∼7 days) where the mass-transfer rate increases to Ṁ2(outburst)=1.5×1018 g s−1.

Does the distribution of light intensity within the barn impact broiler production and welfare?

ABSTRACT 1. The objective of this study was to determine if rearing broilers under variable light intensity (VLI) impacted their welfare or productivity. 2. Ross 308 broilers (n = 7,256) were reared until 35 d of age and exposed to a uniform intensity of 10 lux (CON) or VLI with low intensity areas of 2–5 lux proximal to the walls and high intensity areas of 84–133 lux proximal to feeders. 3. The data were analysed as a complete randomised design using an analysis of variance. Significance was declared when p ≤ 0.05. 4. Applying VLI resulted in increased feed intake early in life but had no impact on body weight. Overall efficiency was improved in the CON treatment. Mortality diagnoses of skeletal problems were reduced under VLI. Treatment had no impact on footpad, hock or gait score, heterophil to lymphocyte ratio or melatonin concentration. Birds performed certain behaviours in specific locations within the room, independent of light intensity treatment. 5. In conclusion, raising broilers under VLI had little impact on production or most welfare parameters assessed in this study. However, satisfying the bird’s preference for different light intensities may improve welfare.

D3L-SLAM: A Comprehensive Hybrid Simultaneous Location and Mapping System with Deep Keypoint, Deep Depth, Deep Pose, and Line Detection

Robust localization and mapping are crucial for autonomous systems, but traditional handcrafted feature-based visual SLAM often struggles in challenging, textureless environments. Additionally, monocular SLAM lacks scale-aware depth perception, making accurate scene scale estimation difficult. To address these issues, we propose D3L-SLAM, a novel monocular SLAM system that integrates deep keypoints, deep depth estimates, deep pose priors, and a line detector. By leveraging deep keypoints, which are more resilient to lighting variations, our system improves the robustness of visual SLAM. We further enhance perception in low-texture areas by incorporating line features in the front-end and mitigate scale degradation with learned depth estimates. Additionally, point-line feature constraints optimize pose estimation and mapping through a tightly coupled point-line bundle adjustment (BA). The learned pose estimates refine the feature matching process during tracking, leading to more accurate localization and mapping. Experimental results on public and self-collected datasets show that D3L-SLAM significantly outperforms both traditional and learning-based visual SLAM methods in localization accuracy.

CPH-Fmnet: An Optimized Deep Learning Model for Multi-View Stereo and Parameter Extraction in Complex Forest Scenes

The three-dimensional reconstruction of forests is crucial in remote sensing technology, ecological monitoring, and forestry management, as it yields precise forest structure and tree parameters, providing essential data support for forest resource management, evaluation, and sustainable development. Nevertheless, forest 3D reconstruction now encounters obstacles including higher equipment costs, reduced data collection efficiency, and complex data processing. This work introduces a unique deep learning model, CPH-Fmnet, designed to enhance the accuracy and efficiency of 3D reconstruction in intricate forest environments. CPH-Fmnet enhances the FPN Encoder-Decoder Architecture by meticulously incorporating the Channel Attention Mechanism (CA), Path Aggregation Module (PA), and High-Level Feature Selection Module (HFS), alongside the integration of the pre-trained Vision Transformer (ViT), thereby significantly improving the model’s global feature extraction and local detail reconstruction abilities. We selected three representative sample plots in Haidian District, Beijing, China, as the study area and took forest stand sequence photos with an iPhone for the research. Comparative experiments with the conventional SfM + MVS and MVSFormer models, along with comprehensive parameter extraction and ablation studies, substantiated the enhanced efficacy of the proposed CPH-Fmnet model in addressing difficult circumstances such as intricate occlusions, poorly textured areas, and variations in lighting. The test results show that the model does better on a number of evaluation criteria. It has an RMSE of 1.353, an MAE of only 5.1%, an r value of 1.190, and a forest reconstruction rate of 100%, all of which are better than current methods. Furthermore, the model produced a more compact and precise 3D point cloud while accurately determining the properties of the forest trees. The findings indicate that CPH-Fmnet offers an innovative approach for forest resource management and ecological monitoring, characterized by cheap cost, high accuracy, and high efficiency.

An effective unsupervised domain adaptation for in-field potato disease recognition

Accurate disease recognition through computer vision is crucial for the intelligent management of potato production. Popular data-driven classification methods face challenges including limited labelled data and poor model portability. Unsupervised Domain Adaptation (UDA) addresses these challenges with a novel learning strategy. However, the complex field environment introduces a significant domain shift problem due to varying conditions. Existing UDA methods usually concentrate on aligning global data distribution and employ a single structure for disease feature extraction, thereby limiting their efficacy in true field environment. To tackle this challenge of potato disease recognition, the Multi-Representation Adaptive Network (MRSAN) based on subdomain alignment is presented. MRSAN effectively aligns feature distributions across diverse data by minimising distribution differences among relevant subdomains. Simultaneously, the multi-representation extraction method captures finer details from various perspectives in the disease images. The combination of these two approaches efficiently mitigates the adverse effects caused by various interference factors in field environment. Based on the acquisition conditions of light variation and disease progression, two field potato disease image datasets are created, containing five and six kinds of potato leaf disease, respectively. Extensive transfer experiments are conducted on the two datasets. MRSAN achieves average classification accuracies of 87.03% and 80.06% on the datasets for the corresponding transfer tasks, outperforming the other compared methods. This not only validates the effectiveness of MRSAN but also demonstrates its robust ability to generalise across changes in regard to light variation and disease progression.

High-Precision Disparity Estimation for Lunar Scene Using Optimized Census Transform and Superpixel Refinement

High-precision lunar scene 3D data are essential for lunar exploration and the construction of scientific research stations. Currently, most existing data from orbital imagery offers resolutions up to 0.5–2 m, which is inadequate for tasks requiring centimeter-level precision. To overcome this, our research focuses on using in situ stereo vision systems for finer 3D reconstructions directly from the lunar surface. However, the scarcity and homogeneity of available lunar surface stereo datasets, combined with the Moon’s unique conditions—such as variable lighting from low albedo, sparse surface textures, and extensive shadow occlusions—pose significant challenges to the effectiveness of traditional stereo matching techniques. To address the dataset gap, we propose a method using Unreal Engine 4 (UE4) for high-fidelity physical simulation of lunar surface scenes, generating high-resolution images under realistic and challenging conditions. Additionally, we propose an optimized cost calculation method based on Census transform and color intensity fusion, along with a multi-level super-pixel disparity optimization, to improve matching accuracy under harsh lunar conditions. Experimental results demonstrate that the proposed method exhibits exceptional robustness and accuracy in our soon-to-be-released multi-scene lunar dataset, effectively addressing issues related to special lighting conditions, weak textures, and shadow occlusion, ultimately enhancing disparity estimation accuracy.

Poliface: A Multi-pose Synchronous Imaging System

To enhance the accuracy of face recognition technology in real-world scenarios, it is necessary to train deep learning models on datasets that contain a large number of labeled human face images under multiple poses, lighting, and accessory variations. In this paper, we introduce a novel acquisition system named the Poliface. This system can capture multiple high-resolution images simultaneously around the human head. We designed this system with a well-built aluminum structure, control electronic circuits, and high-performing in-house software. The results demonstrate the precise operation and exceptional stability of this system. Using this Poliface system, we have collected over 6 million photos, which can be used to train and evaluate facial recognition models, and exploited for three-dimensional (3D) virtual face reconstruction.

MDAR: A Multiscale Features-Based Network for Remotely Measuring Human Heart Rate Utilizing Dual-Branch Architecture and Alternating Frame Shifts in Facial Videos

Remote photoplethysmography (rPPG) refers to a non-contact technique that measures heart rate through analyzing the subtle signal changes of facial blood flow captured by video sensors. It is widely used in contactless medical monitoring, remote health management, and activity monitoring, providing a more convenient and non-invasive way to monitor heart health. However, factors such as ambient light variations, facial movements, and differences in light absorption and reflection pose challenges to deep learning-based methods. To solve these difficulties, we put forward a measurement network of heart rate based on multiscale features. In this study, we designed and implemented a dual-branch signal processing framework that combines static and dynamic features, proposing a novel and efficient method for feature fusion, enhancing the robustness and reliability of the signal. Furthermore, we proposed an alternate time-shift module to enhance the model’s temporal depth. To integrate the features extracted at different scales, we utilized a multiscale feature fusion method, enabling the model to accurately capture subtle changes in blood flow. We conducted cross-validation on three public datasets: UBFC-rPPG, PURE, and MMPD. The results demonstrate that MDAR not only ensures fast inference speed but also significantly improves performance. The two main indicators, MAE and MAPE, achieved improvements of at least 30.6% and 30.2%, respectively, surpassing state-of-the-art methods. These conclusions highlight the potential advantages of MDAR for practical applications.

LUOJIA Explorer: An Auto-UAV for Unexposed Space Exploration

Abstract. With the advancement of technology, unexposed spaces have emerged as a new type of strategic area, attracting significant attention from researchers. These spaces often present complex environments, such as extreme lighting variations, irregular geometries, and the lack of external positioning signals, making human entry hazardous. Fortunately, advancements in robotics and artificial intelligence have enabled unmanned systems to mitigate these challenges, allowing for safer exploration of unexposed spaces. UAVs, as vital unmanned system platforms, are extensively used across various industrial scenarios and are particularly effective for detecting unexposed spaces. To facilitate automated exploration of these areas, we propose the autonomous UAV system, LUOJIA Explorer. Equipped with a LiDAR sensor, data transmission module, and external anti-collision devices to protect the aircraft, LUOJIA Explorer is designed for effective unexposed space detection. We have integrated a laser point cloud positioning and mapping subsystem, along with a planning and control subsystem, into the LUOJIA Explorer. Through both simulation and actual experiments, the Luojia Explorer has demonstrated the capability to achieve stable flight based on self-positioning in unexposed spaces, with an exploration efficiency of 88.83 m3/s.

Nanostructural Influence on Optical and Thermal Properties of Butterfly Wing Scales Across Forest Vertical Strata.

Butterfly wing scales feature complex nanostructures that influence wing coloration and various mechanical and optical properties. This configuration plays a key role in ecological interactions, flight conditions, and thermoregulation, facilitated by interactions with environmental electromagnetic energy. In tropical forests, butterflies occupy distinct vertical habitats, experiencing significant light and temperature variations. While wing nanostructures have been widely studied, their variation across different vertical flight preferences remains underexplored. This study investigates the wing nanostructures of 12 tropical butterfly species from the Nymphalidae family, focusing on their optical, morphological, and thermal properties across different forest strata. We analyzed the optical response through diffuse reflectance in the UV, Vis, and NIR ranges, correlating these findings with nanostructural configuration and thermal stability using thermogravimetric analysis (TGA). Our results reveal a significant correlation between flight stratification and wing optical responses, alongside distinct nanostructural features within each stratum. This study demonstrates the variability in butterfly wing nanostructures along the vertical stratification of the forest to cope with environmental conditions, raising new questions for future research on eco-evolutionary flight and thermal adaptations. Additionally, this underscores the importance of understanding how these structural adaptations influence butterfly interactions with their environment and their evolutionary success across different forest strata.

Research on Cattle Behavior Recognition and Multi-Object Tracking Algorithm Based on YOLO-BoT.

In smart ranch management, cattle behavior recognition and tracking play a crucial role in evaluating animal welfare. To address the issues of missed and false detections caused by inter-cow occlusions and infrastructure obstructions in the barn environment, this paper proposes a multi-object tracking method called YOLO-BoT. Built upon YOLOv8, the method first integrates dynamic convolution (DyConv) to enable adaptive weight adjustments, enhancing detection accuracy in complex environments. The C2f-iRMB structure is then employed to improve feature extraction efficiency, ensuring the capture of essential features even under occlusions or lighting variations. Additionally, the Adown downsampling module is incorporated to strengthen multi-scale information fusion, and a dynamic head (DyHead) is used to improve the robustness of detection boxes, ensuring precise identification of rapidly changing target positions. To further enhance tracking performance, DIoU distance calculation, confidence-based bounding box reclassification, and a virtual trajectory update mechanism are introduced, ensuring accurate matching under occlusion and minimizing identity switches. Experimental results demonstrate that YOLO-BoT achieves a mean average precision (mAP) of 91.7% in cattle detection, with precision and recall increased by 4.4% and 1%, respectively. Moreover, the proposed method improves higher order tracking accuracy (HOTA), multi-object tracking accuracy (MOTA), multi-object tracking precision (MOTP), and IDF1 by 4.4%, 7%, 1.7%, and 4.3%, respectively, while reducing the identity switch rate (IDS) by 30.9%. The tracker operates in real-time at an average speed of 31.2 fps, significantly enhancing multi-object tracking performance in complex scenarios and providing strong support for long-term behavior analysis and contactless automated monitoring.