Search Results (229)

The efficient recognition of unpaved roads from remote sensing (RS) images holds significant value for tasks such as emergency response and route planning in outdoor environments. However, unpaved roads often face challenges such as blurred boundaries, low contrast, complex shapes, and a lack of publicly available datasets. To address these issues, this paper proposes a novel architecture, Swin-FSNet, which combines frequency analysis and spatial enhancement techniques to optimize feature extraction. The architecture consists of two core modules: the Wavelet-Based Feature Decomposer (WBFD) module and the Hybrid Dynamic Snake Block (HyDS-B) module. The WBFD module enhances boundary detection by capturing directional gradient changes at the road edges and extracting high-frequency features, effectively addressing boundary blurring and low contrast. The HyDS-B module, by adaptively adjusting the receptive field, performs spatial modeling for complex-shaped roads, significantly improving adaptability to narrow road curvatures. In this study, the southern mountainous area of Shihezi, Xinjiang, was selected as the study area, and the unpaved road dataset was constructed using high-resolution UAV images. Experimental results on the SHZ unpaved road dataset and the widely used DeepGlobe dataset show that Swin-FSNet performs well in segmentation accuracy and road structure preservation, with an IoU_road of 81.76% and 71.97%, respectively. The experiments validate the excellent performance and robustness of Swin-FSNet in extracting unpaved roads from high-resolution RS images. Full article

Shape-sensing optical fibers have become increasingly important in applications requiring flexible navigation, spatial awareness, and deformation monitoring. Fiber Bragg Grating (FBG) sensors inscribed in multi-core optical fibers have been democratized over the years and nowadays offer a compact and robust platform for shape reconstruction. In this work, we propose a novel, computationally efficient method for determining the 3D tip position of a bent multi-core FBG-based optical fiber using a second-order polynomial approximation of the fiber’s shape. The method begins with a calibration procedure, where polynomial coefficients are fitted for known bend configurations and subsequently modeled as a function of curvature using exponential decay functions. This allows for real-time estimation of the fiber tip position from curvature measurements alone, with no need for iterative numerical solutions or high processing power. The method was validated using miniaturized test structures and achieved sub-millimeter accuracy (<0.1 mm) over a 4.5 mm displacement range. Its simplicity and accuracy make it suitable for embedded or edge-computing applications in confined navigation, structural inspection, and medical robotics. Full article

Object detection is a mature problem in autonomous driving, with pedestrian detection being one of the first commercially deployed algorithms. It has been extensively studied in the literature. However, object detection is relatively less explored for fisheye cameras used for surround-view near-field sensing. The standard bounding-box representation fails in fisheye cameras due to heavy radial distortion, particularly in the periphery. In this paper, a generic object detection framework is implemented using the base YOLO (You Only Look Once) detector to systematically explore various object representations using the public WoodScape dataset. First, we implement basic representations, namely the standard bounding box, the oriented bounding box, and the ellipse. Secondly, we implement a generic polygon and propose a novel curvature-adaptive polygon, which obtains an improvement of 3 mAP (mean average precision) points. A polygon is expensive to annotate and complex to use in downstream tasks; thus, it is not practical to use it in real-world applications. However, we utilize it to demonstrate that the accuracy gap between the polygon and the bounding box representation is very high due to strong distortion in fisheye cameras. This motivates the design of a distortion-aware optimal representation of the bounding box for fisheye images, which tend to be banana-shaped near the periphery. We derive a novel representation called a curved box and improve it further by leveraging vanishing-point constraints. The proposed curved box representations outperform the bounding box by 3 mAP points and the oriented bounding box by 1.6 mAP points. In addition, the camera geometry tensor is formulated to provide adaptation to non-linear fisheye camera distortion characteristics and improves the performance further by 1.4 mAP points. Full article

Soil texture is one of the most important physical properties of soil and plays a crucial role in determining its suitability for crop cultivation. Currently, supervised classification machine learning methods are most commonly used in digital soil mapping. However, these methods may not yield optimal predictive performance due to the limited number of soil samples. Therefore, we propose using Constrained K-Means Clustering to combine a small number of labeled samples with a large amount of unlabeled data, thereby achieving improved prediction in soil texture mapping. In this study, we focused on a typical hilly region in northern Jurong City, Jiangsu Province, China, and used Constrained K-Means Clustering as our mapping model. GF-2 remote sensing imagery and the ALOS digital elevation model (DEM), along with their derived variables, were employed as environmental variables. In Constrained K-Means Clustering, the choice of distance method is a key parameter. Here, we used four different distance methods (euclidean, maximum, manhattan, and canberra) and compared the results with those of the random forest (RF) and multilayer perceptron (MLP) models. Notably, the euclidean distance method within Constrained K-Means Clustering achieved the highest overall accuracy ($OA$), $Kappa$ coefficient, and $MacroF1Score$, with values of 0.77, 0.68, and 0.75, respectively. These methods were higher than those obtained by the RF and MLP models by 0.12, 0.18, and 0.12, and 0.18, 0.26, and 0.18, respectively. This indicates that Constrained K-Means Clustering demonstrates strong predictive performance in soil texture mapping. Moreover, land use (LU), multi-resolution of ridge top flatness index (MRRTF), topographic position index (TPI), and plan curvature (PlC) emerged as the key environmental variables for predicting soil texture. Overall, Constrained K-Means Clustering proves to be an effective digital soil mapping approach, offering a novel perspective for soil texture mapping with limited samples. Full article

Multi-focal microlens arrays provide notable advantages over mono-focal counterparts, such as multi-scale imaging capabilities and optical aberration correction. However, existing multi-focal microlens arrays fabricated on continuous surfaces are incapable of achieving confocal imaging. As a result, multiple focus adjustments are required to acquire comprehensive image data, thereby complicating system design and increasing operational duration. To overcome this limitation, a stepped confocal surface microlens array is proposed, capable of simultaneously capturing images with multiple depths of field, various field-of-view scales, and different resolutions—without the need for additional focus adjustments. A combination of femtosecond laser processing and chemical etching was employed to fabricate microlenses with varying curvatures on a stepped fused silica substrate, which was subsequently used as a mold. The final stepped confocal microlens array was replicated via polydimethylsiloxane (PDMS) molding. Preliminary experimental analyses were carried out to determine the relationship between processing parameters and the resulting focal lengths. By precisely controlling these parameters, the fabricated stepped confocal microlens array successfully enabled confocal imaging, allowing for the simultaneous acquisition of diverse image data. This microlens array shows great potential in advancing lightweight, integrated, and highly stable optical systems for applications in optical sensing, spatial positioning, and machine vision. Full article

Microlens arrays (MLAs) are essential for light collection, extraction, and high-resolution imaging. However, most reported MLAs have a limited filling factor. Here, we demonstrate MLAs using three different UV-curing optical adhesives based on the electrohydrodynamic inkjet (E-jet) printing technique. The highest filling factor of 89.91% is achieved. By controlling the curvature of the microlens via the surface treatment of the substrate, a series of MLAs with different numerical apertures can be obtained. With the high-consistency printing technique, the demonstrated high filling factor MLAs could be potentially useful to improve the performance of optical imaging and sensing systems. Full article

An evaporation duct is a kind of atmospheric event with a refractive index exceeding the curvature of the Earth, which mostly exists on the ocean surface. Evaporation ducts have a great influence on radar, such as causing blind zones or achieving over-the-horizon detection. However, there is a lack of effective technology for evaporation duct detection, especially for passive methods. Global Navigation Satellite System Reflectometry (GNSS-R) has demonstrated potential in various remote sensing applications. However, its utilization for evaporation duct retrieval has not yet been successfully achieved. This study investigates the impact of evaporation ducts on GNSS-R delay maps (DMs), demonstrating that they elevate the non-specular point region, with the extent of this rising zone correlating with the evaporation duct height (EDH). Through semi-physical simulation, the rise signal is modeled. During a four-day experiment, GPS-R DMs with obvious features of evaporation ducts were repeatedly observed. Additionally, this study attempts to find the maximum code delay in the experimental data. The EDH is retrieved using the maximum code delay and GPS elevation angle, exhibiting a 4 m error relative to the reference model under the condition that all effective waveforms are successfully received. The results demonstrate that the GNSS-R offers a promising passive method for evaporation duct detection. Full article

This study proposes an advanced intelligent vehicle path-following control system using deep reinforcement learning, with a particular focus on the role of road geometry perception in motion planning and control. The system is structured around a three-degree-of-freedom (3-DOF) vehicle model, which facilitates the extraction of critical dynamic features necessary for robust control. The longitudinal control architecture integrates a Deep Deterministic Policy Gradient (DDPG) agent to optimise longitudinal velocity and acceleration, while lateral vehicle control is handled by a Deep Q-Network (DQN). To enhance situational awareness and adaptability, the system incorporates key input variables, including ego vehicle speed, speed error, lateral deviation, lateral error, and safety distance to the preceding vehicle, all in the context of road geometry and vehicle dynamics. In addition, the influence of road curvature is embedded into the control framework through perceived acceleration (sensed by vehicle occupants), allowing for more accurate and responsive adaptation to varying road conditions. The vehicle control system is tested in a simulated environment with a lead car in front with realistic speed profiles. The system outputs continuous values for acceleration and steering angle. The results of this study suggest that the proposed intelligent control system not only improves driver assistance but also has potential applications in autonomous driving. This framework contributes to the development of more autonomous, efficient, safety-aware, and comfortable vehicle control systems. Full article

Shape sensing with optical fiber sensors is an emerging technology with broad applications across various fields. This study evaluates the metrological performance of shape sensing cables in the presence of fiber core failures, a critical issue in scenarios where cable replacement is impractical due to technological and economic constraints. The impact of core failure is quantified by comparing the uncertainty in key parameters, such as curvature and bending angle, between pristine and damaged cables through Monte Carlo simulations. Results indicate that while core failure degrades performance, shape reconstruction remains achievable. However, the reconstruction becomes sensitive to bending direction due to the loss of core symmetry. Additionally, simulations of how measurement noise propagates into uncertainty in the 3D shape reconstruction are carried out. Analysis of specific shapes, including a circle and a right-handed helix, shows that increasing the number of sensing cores significantly mitigates the adverse effects of core failure. The most notable improvement occurs when the number of cores is increased from four to five. These findings show how shape reconstruction is still possible even in the presence of core damage, and how this changes the behavior of the sensing process. Full article

This study investigates the mechanisms driving current generation, power output, and charge storage in carbon nanotube springs under mechanical strain, addressing the gap between experimental observations and theoretical modeling, particularly in asymmetric electrical responses. Leveraging the Dirac equation in curved spacetime, we analyze how curvature-induced scalar and pseudo-gauge potentials shape two-dimensional electron gases confined to carbon nanotube springs. We incorporate applied mechanical strain by introducing time-dependent variations in the Lamé coefficient and curvature parameters, enabling the analysis of mechanical deformation’s influence on electrical properties. Our model clarifies asymmetric electrical responses during stretching and compression cycles and explains how strain-dependent power outputs arise from the interplay between mechanical deformation and curvature effects. Additionally, we demonstrate mechanisms by which strain influences charge redistribution within the helically coiled structure. We develop a new equivalent circuit model linking mechanical deformation directly to electronic behavior, bridging theoretical physics with practical electromechanical applications. The analysis reveals asymmetric time-dependent currents, enhanced power output during stretching, and strain-dependent charge redistribution. Fourier analysis uncovers dominant frequency components (primary at $\Omega$, harmonic at $2\Omega$) explaining these asymmetries. Theoretical investigations explain the mechanisms behind the curvature-driven time-dependent current source, the frequency-dependent peak power, the characteristics of open-circuit voltage with strain, and the asymmetric electrical property response under applied strain as the generated current and the charge distribution within the carbon nanotube springs. These findings highlight carbon nanotube springs applied to energy harvesting, wearable electronics, and sensing technologies. Full article

Plant colonization patterns on deglaciated terrain give insight into the factors influencing alpine ecosystem development. Our objectives were to use a chronosequence, extending from the Little Ice Age (~1850) terminal moraine to the present glacier terminus, and biophysical predictors to characterize vegetation across Sperry Glacier’s foreland—a mid-latitude cirque glacier in Glacier National Park, Montana, USA. We measured diversity metrics (i.e., richness, evenness, and Shannon’s diversity index), percent cover, and community composition in 61 plots. Field observations characterized drainage, concavity, landform features, rock fragments, and geomorphic process domains in each plot. GIS-derived variables contextualized the plots’ aspect, terrain roughness, topographic position, solar radiation, and curvature. Overall, vegetation cover and species richness increased with terrain age, but with colonization gaps compared to other forelands, likely due to extensive bedrock and slow soil development, potentially putting this community at risk of being outpaced by climate change. Generalized linear models revealed the importance of local site factors (e.g., drainage, concavity, and process domain) in explaining species richness and Shannon’s diversity patterns. The relevance of field-measured variables over GIS-derived variables demonstrated the importance of fieldwork in understanding alpine successional patterns and the need for higher-resolution remote sensing analyses to expand these landscape-scale studies. Full article

Sinkholes are significant geohazards in karst regions that pose risks to landscapes and infrastructure by disrupting geological stability. Usually, sinkholes are mapped by field surveys, which is very cost-intensive with regard to vast coverages. One possible solution to derive sinkholes without entering the area is the use of high-resolution digital terrain models, which are also expensive with respect to remote areas. Therefore, this study focusses on the mapping of sinkholes in arid regions from open-access remote sensing data. The case study involves data from the Sentinel missions over the Mangystau region in Kazakhstan provided by the European Space Agency free of cost. The core of the technique is a multi-scale curvature filter bank that highlights sinkholes (and takyrs) by their very special illumination pattern in Sentinel-2 images. Marginal confusions with vegetation shadows are excluded by consulting the newly developed Combined Vegetation Doline Index based on Sentinel-1 and Sentinel-2. The geospatial analysis reveals distinct spatial correlations among sinkholes, takyrs, vegetation, and possible surface discharge. The generic and, therefore, transferable approach reached an accuracy of 92%. However, extensive reference data or comparable methods are not currently available. Full article

In this paper, a high-linear-sensitivity fiber curvature sensor based on the sphere-shaped misaligned structure (SSMS) with few-mode fiber (FMF) and single-mode fiber (SMF) was proposed and demonstrated. A spherical structure was prepared at one end of a few-mode fiber, which could effectively excite higher-order modes and generate interference in the misaligned cascade. When external environmental parameters changed, the resonance peaks formed by intermodal interference were displaced, and the shifts generated by different resonant peaks were also different. The experimental results show that the maximum curvature sensitivity was −2.220 nm/m⁻¹, and the linear fitting coefficient reached up to 0.991, which is an extremely high sensitivity among wavelength-modulated curvature sensors. Meanwhile, the strain sensitivity of the sensor was as low as 7.99 pm/μ$\overline{\mathsf{\epsilon }}$, and the temperature sensitivity was 3.958 pm/°C, which is a low temperature sensitivity and low strain sensitivity, and solves the cross-sensitivity problem. With advantages of simple manufacture, low cost, and favorable stability, the sensor is expected to be one of the best candidate instruments for measuring curvature and inclination. Full article

Electronic skin is widely employed in multiple applications such as health monitoring, robot tactile perception, and bionic prosthetics. In this study, we fabricated millimeter-scale electronic skin featuring compact sensing units using the Boston Micro Fabrication S130 (a high-precision additive manufacturing device) and the template removal method. We used a gallium-based liquid metal and achieved an inner channel diameter of 0.1 mm. The size of the sensing unit was 3 × 3 mm². This unit exhibited a wide linear sensing range (10–22,000 Pa) and high-pressure resolution (10 Pa) even on an ultra-curved surface (radius of curvature was 6 mm). Sliding was successfully detected at speeds of 8–54 mm/s. An artificial nose with nine sensing units was fabricated, and it exhibited excellent multitouch and sliding trajectory recognition capabilities. This confirmed that the electronic skin functioned normally, even on an ultra-curved surface. Full article

Flexible sensing materials have become a hot topic due to their sensitive electrical response to external force or temperature and their promising applications in flexible wear and human–machine interaction. In this study, a PDMS/capillary GaInSn flexible sensing material with high force and thermal sensitivity was prepared utilizing liquid metal (LM, GaInSn), flexible silicone capillary, and polydimethylsiloxane (PDMS). The resistance (R) of the flexible sensing materials under the action of different forces and temperatures was recorded in real-time. The electrical performance results confirmed that the R of the sensing material was responsive to temperature changes and increased with the increasing temperature, indicating its ability to transmit temperature signals into electrical signals. The R was also sensitive to the external force, such as cyclic stretching, cyclic compression, cyclic bending, impact and rolling. The ΔR/R₀ changed periodically and stably with the cyclic stretching, cyclic compression and cyclic bending when the conductive pathway diameter was 0.5–1.0 mm, the cyclic tensile strain ≤ 20%, the cyclic tensile rate ≤ 2.0 mm/min, the compression ratio ≤ 0.5, and the relative bending curvature ≤ 0.16. Moreover, the material exhibited sensitivity in detecting biological signals, such as the joint movements of the finger, wrist, elbow and the stand up-crouch motion. In conclusion, this work provides a method for preparing a sensing material with the capillary structure, which was confirmed to be sensitive to force and heat, and it produced different types of R signals under different deformations and different temperatures. Full article