Optical flow and stereo disparity, both are fundamental in the perception pipeline of robotic systems, enabling 3D understanding and motion estimation of the robot itself and the dynamic objects around. In this context, event cameras offer great potential to reduce latency and improve the efficiency of the perception pipeline. However, existing event-based approaches often deal with flow and disparity estimation tasks separately, potentially leading to redundant computations and reduced efficiency. In this work, we developed a joint flow and depth estimation network, featuring shared feature encoders for high computational efficiency. Moreover, we introduce a novel Bidirectional Mamba module to enhance feature expressiveness by increasing the spatial receptive field to capture global context with significantly lower computational overhead than vision transformers. We further improve efficiency by incorporating <italic xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">informed priors</i> to reduce the number of refinement iterations in the sequential estimation task. Additionally, we extend our framework to combine the estimated flow and disparity to predict 3D scene flow, an important task in many robotics applications.

Precise 3D crustal displacement retrieval in GCP-free environments: A geodetic and deep learning–assisted integration of InSAR and optical stereo data near the Denali Fault

Net-cage aquaculture faces challenges from biofouling, which reduces water exchange and threatens structural integrity. Automated cleaning robots provide an alternative to human divers but require effective, low-cost localization. Conventional acoustic–inertial systems are expensive and complex, while vision-only or IMU-based methods suffer from drift in turbid, low-texture waters. This paper presents a lightweight Visual–Inertial–Depth (VID) fusion framework for underwater net-cage cleaning robots. Built on the VINS-Fusion system the method estimates scene scale using optical flow and stereo matching, and incorporates IMU pre-integration for high-frequency motion prediction. A pressure-based depth factor constrains Z-axis drift, and reflective-anchor initialization ensures global alignment. The system runs in real time on a Jetson Orin NX with ROS. Experiments in air, tank, pool, and ocean settings demonstrate its robustness. In controlled environments, the mean anchor coordinate error (ACE) was 0.05–0.16 m, and loop-closure drift (LCD) was ≤0.5 m per 5 m. In ocean trials, turbulence and biofouling led to drift (LCD 1.32 m over 16.05 m, 8.3%), but IMU and depth cues helped maintain vertical stability. The system delivers real-time, cost-effective localization in structured underwater cages and offers insights for improvements in dynamic marine conditions.

Abstract The Ilyak fault in Tajikistan is an E‐W striking strike‐slip fault, which forms the northern boundary of the actively deforming Tajik Basin fold and thrust belt. Determining the activity of the Ilyak fault is important for understanding the hazard to the nearby capital city of Dushanbe, and more generally in assessing the role of lithology in controlling fault behavior. Using a bespoke Multi‐Temporal InSAR approach with 5 years of Sentinel‐1 satellite radar acquisitions, we produce regional surface velocity maps with a horizontal spatial resolution of 130 m and numerical uncertainties below 1 mm/yr. Our rate maps reveal previously unrecognized details of creep, fault structure, and slip partitioning. We document creep on the Ilyak fault with a right‐lateral rate of 8.1–8.7 mm/yr in the east reducing to 4.2–4.5 mm/yr in the west, as slip is progressively transferred to adjacent folds and thrusts in the Tajik Basin. Creep on the Ilyak fault and adjacent thrusts appears to be controlled by lithology, in particular the presence of evaporite‐rich layers. Nonetheless, moderate earthquakes within the basement beneath the Ilyak fault, including a destructive M5.5 event in 1989, suggest potential earthquake hazards from deeper crustal layers. Using high‐resolution Digital Elevation Models derived from optical stereo satellite data we document fault scarps north of the Ilyak fault, which may constitute an additional source of hazard. Filtered velocity maps also show evidence of soil creep and active, slow‐moving landslides across the basin and in the mountains, constituting another hazard to local communities and infrastructure.

Abstract. Currently, two types of satellite laser altimetry systems are in operation: the full waveform linear system and the single-photon-counting system, which exhibit significant differences in both data format and processing methodologies. It remains uncertain which satellite laser altimetry technology offers greater advantages in assisting stereo mapping of high-resolution optical imagery. To evaluate the effectiveness of those two satellite laser altimetry technologies in supporting optical imagery stereo mapping, this study extracted laser elevation control points from both GF-7 satellite and ICESat-2 laser altimetry data, and designed a specialized bundle adjustment workflow for GF-7 satellite stereo images incorporating laser elevation control points. Comparative experiments were conducted across both flat and mountainous regions. Results demonstrate that laser elevation control points derived from both GF-7 satellite and ICESat-2 effectively improved the elevation accuracy of stereo mapping of GF-7 satellite images. However, the ICESat- 2 laser altimetry data has advantages in quantity and distribution, and the combined bundle adjustment accuracy of the two survey areas is slightly better than that of the GF-7 satellite laser altimetry data.

ASTER GDEM provides the fundamental data for remote sensing identification of snow cover in mountainous areas. Due to its elevation accuracy being easily affected by optical stereo images and local terrain, many studies have utilized machine learning (ML) models for correction. However, most correction methods rely on a single ML model, which limits the improvement of DEM accuracy. Stacked ensemble learning (SEL) is a newly developed method of improving model performance by combining multiple ML models. This study proposes a DEM correction method based on SEL and ICESatand affiliations. -2/ATL08 products. Taking the Babao River Basin in Qilian Mountains as the study area, five ML models with good DEM correction effects (XGBoost, AdaBoost, LightGBM, BPNN, and CatBoost) were selected and trained using land cover and various terrain factors to obtain DEM errors, respectively. Then, the SEL algorithm was used to integrate the DEM errors of the five ML models and correct GDEM. Using 740 CORS measurements and 48,000 ATL08 points for accuracy validation, the results showed that the SEL achieved higher DEM accuracy than any single ML model. The root mean square error (RMSE) of the corrected GDEM decreased from 7.15 m to 4.13 m, while the mean absolute error (MAE) and mean bias error (MBE) values both decreased about by 38%. Furthermore, unmanned aerial vehicle (UAV) DEM data from five sample areas were selected for profile analysis, and it was found that the corrected GDEM was closer to the real surface. Further analysis revealed that the influence of slope, aspect, and land cover types on corrected DEM was weakened, with the most significant improvement in DEM accuracy observed in areas with slope ≥5°, north orientation, and bare land. This study can provide high-precision DEM scientific data for quantitative remote sensing, flood prediction, and other research.

Enhancing Large-Area DEM modeling of GF-7 stereo imagery: Integrating ICESat-2 data with Multi-characteristic constraint filtering and terrain matching correction

ABSTRACTTerrestrial ecosystem carbon monitoring satellite (CM‐1) equipped with multi‐beam LiDAR, multi‐angle multi‐spectral camera and other payloads, which can carry out high accuracy stereo mapping applications, has become a new member of China's civilian stereo mapping satellite series following GF‐7 and ZY‐3. Aiming at the composite mapping capability for the satellite, this paper studies the extraction of laser elevation control points and the composite surveying method combined with stereo images, and compares and analyzes the accuracy of composite surveying from three perspectives: different terrains (include flatland, hilly terrain, mountainous region, and alpine land), different angles of optical stereo images (38° and 82°), and different satellites (CM‐1, ZY3‐03, and GF‐7). The experimental results show that the elevation accuracy of the laser points after screening and extraction is decreased from 2.41 and 4.13 m to 0.44 and 0.607 m, respectively, which can be used as elevation control points for stereo mapping. The accuracy of composite surveying can be improved from 11.45 to 2.27 m, and different terrain conditions are significantly improved. Stereo images of different angles can be combined with laser for mapping, but the accuracy of the image of 38° convergence angle is better than 82°, and spatial resolution may be the main influencing factor. Among different Chinese satellites, the relative elevation accuracy of DSM generated by GF‐7 satellite is better than that of the other two satellites, and the elevation accuracy of DSM generated by CM‐1 satellite is equivalent to that of ZY3‐03 satellite. The relevant conclusions have reference value for giving full play to the application advantages of laser elevation control points and serving the production of real‐scene three‐dimensional terrain‐level products in China.

Abstract Rapid, transient, landscape‐scale changes associated with deglaciation can condition slopes for failure and trigger bedrock landslides. However, the mechanisms leading to paleo rock slope failures following the last glacial period are challenging to infer because observations of how both landsliding and potential driving factors were distributed in space and time are limited. Here, we map and analyze the spatiotemporal pattern of 676 post‐glacial bedrock landslides around Eyjafjörður in north‐central Iceland using 2‐m resolution digital elevation data generated from optical stereo satellite imagery. Frequency‐ratio analysis demonstrates that after controlling for slope, landslides are most overrepresented within 2.6 km horizontal distances from surface projections of major Tertiary bedrock structures and at land surface elevations within 300 m of a modeled lower limit to permafrost. Surface roughness analysis of landslide deposits indicates that peak landslide frequency of at least 0.2 landslides yr−1 in the 5,579 km2 study area lagged deglaciation by several thousand years. This timing aligns well with that of rapid permafrost degradation from the Younger Dryas (12.9–11.7 cal ky BP) through the Holocene Thermal Maximum (∼10–7 cal ky BP). Landslide frequency has averaged about 0.014 landslides yr−1 since the Holocene Thermal Maximum when the climate has generally been cooler and permafrost has been more extensive. However, present day warming is likely to reduce permafrost extent and increase the potential for bedrock landslides in north‐central Iceland, as has already been observed for several recent shallower landslides in regolith.

Abstract. The Gaofen-7 (GF-7) satellite was launched on November 3, 2019. It can achieve 1:10000 scale stereo mapping and serve the application needs of basic surveying and mapping, natural resources monitoring and geographic information resource construction. These application rely on high-precision digital surface model (DSM) products as data support. Therefore, it is of great significance to analyze the quality of DSM products produced by Gaofen-7. This paper uses GF-7 optical stereo images from the regions of Zhaodong and Tianjin in China to extract DSMs. And the difference DSM is produced by subtracting the 0.5m resolution aerial reference DEM and the 30m resolution COP30 product in the same area. The quality of DSM products is evaluated by comparing and analyzing the difference DSM. The results show that when the DSM produced by the GF-7 satellite uses the aerial DEM as a reference, the elevation accuracy can reach 0.97 m in the Zhaodong area and 1.54 m in the Tianjin area, and the details of the objects are well depicted; when the COP30 product is used as a reference, the elevation positioning accuracy can reach 1.29 m in the Zhaodong area and 2.42 m in the Tianjin area, which can correctly show the overall situation of the surface objects. Among them, the elevation accuracy in the Tianjin area is relatively poor, mainly because there are many water areas and buildings in this area, and the COP30 product has data missing in this area, which affects the accuracy. In general, the quality of the DSM products generated by the GF-7 satellite is good. Through effective ground control and further processing, the DSM products can fully meet the needs of practical applications such as 1:10000 scale mapping.

Urban Canopy Parameters (UCPs) are crucial for urban microclimate modeling; however, the scarce availability of precise UCP data in developing regions limits their application for urban climates. This study investigated the use of multi-platform remote sensing data viz. very high-resolution satellite (VHRS) optical stereo and Unmanned Aerial Vehicle (UAV) datasets for the computation of UCPs in high-density urban scenarios in India, with varied development characteristics. The results demonstrated high accuracy in terms of building height and footprint extraction from both datasets, key inputs for UCP computation. However, UCPs from UAV data have displayed relatively high accuracy for building footprints (86%), building height (RMSE ~ 0.05 m), and land use/land cover classification (90%). Performance evaluation of computed UCPs against a 3D reference geodatabase showed high prediction accuracy for most UCPs, with overall biases, mean absolute error, and root-mean-square error values significantly better than 1 m, with strong correlation (0.8–0.9). It was concluded that VHRS optical stereo and UAV datasets offer a secure, reliable, and accurate solution for UCP computation in urban areas, particularly in developing regions. These findings have significant implications for urban climate research and the sustainable development of rapidly urbanizing areas facing resource and policy constraints.

Surge-type glaciers in the Himalayas have been extensively documented, however, understanding of the mechanisms behind glacier surges remain limited. Herein, based on multi-source remote sensing data, we systematically investigated the characteristics and subglacial processes of the unnamed glacier at Abi Gamin Peak in central Himalayas. Our approach integrated optical stereo photogrammetry, feature tracking, visual interpretation, and glacier velocity decomposition. Our study reveals that the glacier surge was initiated in August 2019 and lasted for less than six months, with rapid acceleration and deceleration phases. During the surge, the glacier transported approximately 0.233 km3 of mass from higher to lower elevations, leading to a thickness increase of >70 m at the glacier terminus. The glacier terminus advanced by >800 m, accompanied by significant expansion of the crevasses. Furthermore, we quantitatively revealed the evolution of basal stresses, strain rates, and sliding velocities during the surge, indirectly characterising the influence of water on it. We contend that the surge in this glacier was primarily driven by subglacial sliding induced by glacier surface meltwater. Significant changes in regional climate serve as external factors that perturb the glacial dynamic equilibrium. The observed characteristics of the unnamed glacier at Abi Gamin Peak indicate that its surge was controlled by a hydrological switch. Notably, our work contributes to an enhanced understanding of glacier surge mechanisms in High Mountain Asia.

Three-dimensional (3-D) stereo images can be generated via computer-based image processing of CORONA stereo pairs. To a certain extent, important terrain and surface feature data extracted from these stereo images can improve the survey of archaeological sites and the identification and mapping of major landscapes. In this study, we focused on the identification of the archaeological ruins of Liangzhu City. An optical stereo model (red/blue stereo image) of the Liangzhu site was created through computer-based mosaicking and processing of CORONA remote-sensing stereo pairs taken in the 1960s and 1970s. By importing the optical stereo model into mobile phones, tablet computers, and other mobile devices, the research team undertook real-time locating of ruins via human observation, on-site investigation, and image overlay during a field survey and identified several Liangzhu-period dams, some of which have been confirmed via archaeological field investigations. The research team later applied the same method to the identification of tombs at the site of the mausoleums of the six emperors of the Southern Song dynasty. The results further prove that this method is feasible and reliable and can be widely promoted and used for the identification of archaeological ruins.

Abstract. Observations of glacier mass changes are key to understanding the response of glaciers to climate change and related impacts, such as regional runoff, ecosystem changes, and global sea level rise. Spaceborne optical and radar sensors make it possible to quantify glacier elevation changes, and thus multi-annual mass changes, on a regional and global scale. However, estimates from a growing number of studies show a wide range of results with differences often beyond uncertainty bounds. Here, we present the outcome of a community-based inter-comparison experiment using spaceborne optical stereo (ASTER) and synthetic aperture radar interferometry (TanDEM-X) data to estimate elevation changes for defined glaciers and target periods that pose different assessment challenges. Using provided or self-processed digital elevation models (DEMs) for five test sites, 12 research groups provided a total of 97 spaceborne elevation-change datasets using various processing approaches. Validation with airborne data showed that using an ensemble estimate is promising to reduce random errors from different instruments and processing methods but still requires a more comprehensive investigation and correction of systematic errors. We found that scene selection, DEM processing, and co-registration have the biggest impact on the results. Other processing steps, such as treating spatial data voids, differences in survey periods, or radar penetration, can still be important for individual cases. Future research should focus on testing different implementations of individual processing steps (e.g. co-registration) and addressing issues related to temporal corrections, radar penetration, glacier area changes, and density conversion. Finally, there is a clear need for our community to develop best practices, use open, reproducible software, and assess overall uncertainty to enhance inter-comparison and empower physical process insights across glacier elevation-change studies.

Generally, the classic three-dimensional (3D) geometric positioning of optical satellite imagery uses the least squares principle to calculate the coordinates of a ground point by minimizing the sum of the squares of the distances between two imaging rays, which requires the standard stereo data with good imaging geometric conditions. As for unconventional stereo images, the undesirable and ubiquitous weak intersection phenomena exist in data will lead to bad results or even calculation failures for the conventional method. By selecting the highest precision intersection point in block adjustment, a new method that can solve the 3D coordinates with higher accuracy and stability was proposed. Tests of two data sets covering different landscapes validated the effectiveness of the method. The results showed that the geo-positioning performance and robustness of the proposed method was better than that of the conventional method, and this advantage is even greater in areas with more undulating terrain and more images with weak convergence.

本研究では，ステレオカメラの設置深度や撮影時刻により，計測個体数や計測平均尾叉長が変化する可能性の検証を目的に，養殖ブリ生簀内の異なる深度に同時にステレオカメラを設置して12時間の撮影を行い，計測個体数や計測平均尾叉長のパネルデータ分析を行った。その結果，ステレオカメラの設置深度や撮影時刻により計測個体数や計測平均尾叉長が変化することが示唆された。この事実は，画角の限られるカメラ計測の特徴を示しており，正確なモニタリング実施には，ステレオカメラの設置深度の慎重な検討が必要であることが示唆された。

Acquiring disparity maps by dense stereo matching is one of the most important methods for producing digital surface models. However, the characteristics of optical satellite imagery, including significant occlusions and long baselines, increase the challenges of dense matching. In this study, we propose an end-to-end edge-guided multi-scale matching network (EGMS-Net) tailored for optical satellite stereo image pairs. Using small convolutional filters and residual blocks, the EGMS-Net captures rich high-frequency signals during the initial feature extraction phase. Subsequently, pyramid features are derived through efficient down-sampling and consolidated into cost volumes. To regularize these cost volumes, we design a top–down multi-scale fusion network that integrates an attention mechanism. Finally, we innovate the use of trainable guided filter layers in disparity refinement to improve edge detail recovery. The network is trained and evaluated using the Urban Semantic 3D and WHU-Stereo datasets, with subsequent analysis of the disparity maps. The results show that the EGMS-Net provides superior results, achieving endpoint errors of 1.515 and 2.459 pixels, respectively. In challenging scenarios, particularly in regions with textureless surfaces and dense buildings, our network consistently delivers satisfactory matching performance. In addition, EGMS-Net reduces training time and increases network efficiency, improving overall results.

Understory topography serves as a crucial data source, playing an instrumental role in numerous forest ecosystem applications. However, the use of synthetic aperture radar interferometry and optical stereo for the acquisition of ASTER (Advanced Spaceborne Thermal Emission and Reflection Radiometer), SRTM (Shuttle Radar Topography Mission), and GLO-30 (Copernicus Digital Elevation Model) DEM presents unique challenges, particularly in forested environments. These challenges are primarily due to limitations in penetration capability and the effects of foreshortening. ICESat-2/ATLAS, with its higher spatial sampling rate and strong penetrability, presents a new opportunity for estimating forest height parameters and understory terrain. We assessed the vertical accuracy of ASTER, SRTM, GLO-30, and ATLAS in the forest study areas of the United States compared to the reference dataset DTM provided by G-LiHT and we will further discuss the influence of different ground altitudes, forest types, slopes, and aspects on vertical accuracy. The study reveals that in a forested environment, ICESat-2 ATL03 exhibits the highest accuracy at the footprint scale, with a correlation coefficient (R2) close to 1 and Root Mean Square Error (RMSE) = 1.96 m. SRTM exhibits the highest accuracy at the regional scale, with an R2 close to 0.99, RMSE = 11.09 m. A significant decrease in accuracy was observed with increasing slope, especially for slopes above 15°. With a sudden increase in altitude, such as in mountainous situations, the accuracy of vertical estimation will significantly decrease. Aspect and forest cover indeed influence the accuracy of the four DEM products, but this influence lacks a clear pattern. Our results show that ICESat-2 and SRTM data might show sufficient and stable vertical accuracy in a forested environment.

This study aims to explore factors related to optical navigation that interfere with the accuracy of robot-assisted surgery, specifically focusing on the TIANJI Robot system. A measurement model was created to assess the accuracy of the TIANJI Robot system in simulated screw placement. Deviation between actual and planned positions was measured using a three-coordinate machine. Various experiments were conducted to investigate the impact of different optical navigation factors on screw placement accuracy. Deviations were measured at different distances (ranging from 1.2 to 2.2m) between the optical navigation stereo camera and the tracker, with each distance being tested 50 times. The distance between the optical camera and patient tracker was set at 1.4m. Deviations were also measured at different angles between the camera and robot tracker, repeated over 25 times for each angle. Data were analyzed using mean and standard deviation, with line charts illustrating deviation changes based on distance and angle details. Within the range of the TIANJI Robot system's optical navigation (1.2-2.2m), deviation increased as distance increased (χ2 = 479.107, P < 0.001). The robotic system demonstrated high and consistent accuracy (mean deviation: 0.332mm ± 0.067mm) when the relative angle between the optical camera and tracker was below 40°. The accuracy of the TIANJI Robot system was found to be influenced by relative distance and angle between the optical camera and tracker during screw placement procedures. Surgeons are recommended to set a relative distance of 1.4-1.5m between the optical camera and patient tracker, with a relative angle below 40° when placing and adjusting optical tracking devices.

High-resolution optical satellite stereo image pairs have been a challenge for dense matching due to their characteristics, such as long baseline and significant occlusion. This letter proposes a semiglobal matching with optimized penalty and interpolation (OPI-SGM), which includes the segmental adaptive adjustment (SAA) for the penalty term and the dual-path interpolation (DPI) for invalid regions. We introduce <inline-formula xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink"> <tex-math notation="LaTeX">$\gamma $ </tex-math></inline-formula> as the maximum absolute gray value difference between pixels in the same plane and combine it with the logarithmic function to make SAA less sensitive to penalty adjustment in the aggregation function. DPI uses two paths to improve interpolation reliability and uses <inline-formula xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink"> <tex-math notation="LaTeX">$\gamma $ </tex-math></inline-formula> as a threshold to filter weighted valid disparities, reducing the number of erroneous disparities near dense building areas. Experimental results show that OPI-SGM outperforms other SGM variants. When processing new data, it produces more accurate disparity maps than two end-to-end matching networks, demonstrating the effectiveness of the proposed algorithm.