Page 1 (data 1 to 30 of 339) | Displayed ini 30 data/page
1 2 3 4 5 6 7 8 9 10 11 12 NEXT >>

1	Topic A: General Remote Sensing	ABS-269
Object Based Classification Approach for impervious and non-impervious surfaces in satellite images Sujuma Basumatary (1), R.D. Garg (2), Pankaj Pratap Singh (3*) 1) Department of Computer Science & Engineering, Central Institute of Technology Kokrajhar, India 2) Geomatics Engineering, Department of Civil Engineering, Indian Institute of Technology Roorkee, India 3) Department of Computer Science & Engineering, Central Institute of Technology Kokrajhar, India *pankajp.singh[at]cit.ac.in Abstract In the last few years, satellite image classification is gaining more attention due to the availability of remotely sensed imagery^s in high spatial resolution. This paper approached a non-linear type object classification approach based on object basis which incorporates K-Nearest Neighbour (KNN) algorithm for segmentation and classification. This proposed approach is based on object based image analysis (OBIA) technique. Spatial information is playing an important role in this technique. In this work, various features are extracted and utilized for the classification of non-linear objects. Spectral features of the training image objects are extracted using region of image (ROI) based samples which are used in KNN algorithm for segmentation and classification with a good level of accuracy. Images are classified in five types of objects such as road, building, land, water body, and vegetation also. In addition, parking lots are also having sometimes similar types of spectral reflectance as road due to similar material in both. The primarily focus of this work is to extract the non-linear objects by avoiding misclassification in a compact manner and also to improve the visibility of object. Keywords: Object-based Classification- Multiresolution Segmentation- K-Nearest Neighbour- Impervious surfaces- Non-impervious surfaces Share Link | Plain Format | Corresponding Author (PANKAJ PRATAP SINGH)

2	Topic A: General Remote Sensing	ABS-23
Wood Extraction from Urban Street Tree Point Clouds using a Deep Learning Approach Andrew Egbert Wiryawan (a), Chi-Kuei Wang (a*), Mohamad Bagas Setiawan (a), Michael Vashni Immanuel Ryadi (a) (a) Department of Geomatics, National Cheng Kung University, Tainan City 70101, Taiwan R.O.C Abstract Accurate tree structure measurement is important for urban forest carbon assessment. For this purpose, Terrestrial Laser Scanner (TLS) enables these measurements through 3D point clouds that capture detailed tree structure. However, extracting wood components from point clouds has remained challenging due to incomplete data from occlusions and species diversity. Recent advances in deep learning have shown promising results for wood extraction. Most of the proposed deep learning techniques for extracting the wood point clouds use point-based methods, voxel-based methods, and projection-based methods. Among these three methods, the projection-based method has its potential for further exploration. Therefore, this study proposes a workflow that projects 3D voxelised point clouds into multiview 2D cross-sectional images. Our method employs Swin-UPerNet networks on the 2D cross-sectional images to perform wood extraction in urban tree point clouds. We apply average voting to determine the final predicted class of point clouds. This method examines three different urban scenario datasets, i.e. one virtual dataset and two real-world datasets. Our results show a promising accuracy of wood point clouds. Furthermore, the findings demonstrate the potential of projection-based approaches in tackling various species and incomplete data due to occlusion challenges for extracting the wood structure from urban forest scenes. Keywords: multiview projection-based, Swin-UPerNet, terrestrial laser scanning, urban forest, wood extraction Share Link | Plain Format | Corresponding Author (Andrew Egbert Wiryawan)

3	Topic A: General Remote Sensing	ABS-25
ASSESSING BEACH CARRYING CAPACITY: APPLYING OBJECT DETECTION TECHNOLOGY AT SAI KAEW BEACH, KOH SAMET THAILAND Chomchid Phromsin(a*), Kunlasatee pookpanich (a), Benjamapawn Kitjao (a), Nanticha Poonpanich (a), Punyanut Traiyatha (a) a) Kasetsart University, Department of Geography, Faculty of Social Science, Bangkok, Thailand *fsoccci[at]ku.ac.th Abstract Carrying capacity refers to the maximum number of people that can utilize a specific area. Assessing the carrying capacity of a beach involves determining the areas capacity or the maxi-mum number of visitors who can engage in activities within that beach area during a specific time period. Counting the crowd accurately poses a significant challenge in carrying capacity assessments. The aim of this paper is to apply Mask R-CNN (Region Based Convolutional Neural Networks) for detecting people and counting the number of visitors on the beach in order to evaluate the carrying capacity of Sai Kaew beach, located in Samet Island (Koh Samet), Rayong Province, Thailand between December 2019 and February 2020. This study used photo images taken by a digital camera to capture tourists on the beach by walking 4 rounds per day. The results of the research, which aimed to evaluate the accuracy of object detection using the Mask R-CNN model, indicated a precision value of 96.47% and a recall value of 92.15%. Through the application of Mask R-CNN, the study estimated the people at one time (PAOT) to be approximately 4,048 tourists per day within an 8,580 m2 area of the beach. These results contribute to the evaluation of the carrying capacity of Sai Kaew beach and provide valuable insights for managing visitor numbers and activities in the area. Keywords: carrying capacity- beach assessment- Mask R-CNN- tourist detection Share Link | Plain Format | Corresponding Author (Chomchid PHROMSIN)

4	Topic A: General Remote Sensing	ABS-35
Reflection Asymmetric Target Detection Based on Correlation Coefficients Ken Yoong LEE, Soo Chin LIEW, Santo V. SALINAS and Kim Hwa LIM Centre for Remote Imaging, Sensing and Processing, National University of Singapore Abstract Nghiem et al. (1992) presented an in-depth investigation on symmetric properties of geophysical media encountered in polarimetric remote sensing. Recently, Connetable et al. (2022) made use of statistical test for independence in evaluating the reflection symmetry in multi-look PolSAR data. They derived the so-called block-diagonality test statistic, where its asymptotic distribution is a chi-squared distribution. In this paper, the mathematical relationship between the block-diagonality test statistic and multiple correlation R was first investigated. The former can actually be expressed in form of the latter, which arises from multiple linear regression. Moreover, the exact distribution of the squared multiple correlation R^2 for a 3x3 polarimetric covariance matrix is a beta distribution, which depends only on the number of looks L. As the number of looks tends to infinity, the limiting distribution of LR^2 is a gamma distribution with unit scale parameter and a shape parameter of two. In this study, the reflection asymmetric target detection was carried out using multi-look ALOS-2 PALSAR-2 quad-polarisation data, which covered two separate test sites in Southeast Asia, namely, Singapore and Penang (Malaysia). Meanwhile, quantitative assessment was reported by using simulated PolSAR data through Monte-Carlo method following the procedure suggested by Lee et al. (1994). The experimental results confirmed the usefulness of reflection symmetry property of geophysical media, particularly in detecting man-made objects, such as bridge, slightly oriented buildings, vessels, aquaculture farms, etc. Keywords: ALOS-2. complex Wishart distribution, multiple correlation, polarimetric synthetic aperture radar, reflection symmetry Share Link | Plain Format | Corresponding Author (Ken Yoong Lee)

5	Topic A: General Remote Sensing	ABS-36
Identification of the Cold Pool Phenomenon at Soekarno-Hatta Airport Site Soni Soeharsono Master Program in Earth Sciences, Faculty of Earth Sciences and Technology, Bandung Institute of Technology, Bandung, Indonesia Abstract Characteristics of the cold pool over the Soekarno-Hatta airport site area have been documented based on the analysis of C-band Doppler radar images (6.16S, 106.64E). The cold pool phenomenon that was successfully observed originated from convective clouds. The existence of this phenomenon mostly disrupts aircraft landing activities above the Soekarno-Hatta airport area, resulting in aircraft failing to land and performing go-around maneuvers. One of the documented cold pool phenomena triggered the growth of convective clouds that resulted in flooding at the Soekarno-Hatta airport area. The results of research from lidar ceilometer observation data using cloud base height analysis showed that the cloud base height decreased after the cold pool propagation passed through the Soekarno-Hatta airport area. Analysis of cold pool energy and its propagation using wind profiler observation data calculations showed that the respective values were C^2 >= 100 m^2/s^2 and C >= 5 m/s. Analysis of the observed cold pool depth (H) shows that the cold pool phenomena that were successfully observed had an H value of >= 1300 m. Keywords: cold pool, gusty, convective clouds Share Link | Plain Format | Corresponding Author (Soni Soeharsono)

6	Topic A: General Remote Sensing	ABS-40
Development of an alternative empirical relationship and parameters to model the absorption coefficient of phytoplankton in Singapore waters Ryan Tan1*, Amihan Yson Manuel1, Sandric C. Y. Leong2, and Soo Chin Liew1 1Centre for Remote Imaging, Sensing and Processing, National University of Singapore, Singapore 2Tropical Marine Science Institute, National University of Singapore, Singapore Abstract The presence of phytoplankton contributes significantly to ocean color, and their abundance makes them the largest and most important primary producers in the ocean. Through photosynthesis, phytoplankton are responsible for most of the transfer of carbon dioxide from the atmosphere to the ocean, serving as a natural carbon sink. The absorption coefficient of phytoplankton is also directly linked to the concentration of chlorophyll-a in the ocean, which is an important water quality parameter. For these reasons, it is important that we can accurately model the absorption coefficient of phytoplankton. Current methods to model the absorption coefficient of phytoplankton make use of empirical equations and parameters that were determined from the waters of their respective study areas and may not be universally applicable. Hence, in this study, we obtain lab measurements of water samples from field stations in Singapore to develop an alternate empirical relationship and parameters that are more representative of Singapore waters. Our results show that with this new empirical model, we are able to better estimate water quality parameters in Singapore using remote sensing. Keywords: phytoplankton, water quality, empirical, modelling Share Link | Plain Format | Corresponding Author (Ryan Tan)

7	Topic A: General Remote Sensing	ABS-44
Depth Refinement in 3D Mapping of Construction Sites Using a Stereo Camera Ryunosuke Ishiguro (a*), Junichi Susaki (a), Yoshie Ishii (a) Department of Civil and Earth Resources Engineering, Graduate School of Engineering, Kyoto University *ishiguro.ryunosuke.62w[at]st.kyoto-u.ac.jp Abstract Construction industry in Japan faces a critical challenge due to a severe labor shortage and a decline in skilled crane operators, creating a pressing need for automated 3D environmental mapping to support crane operations. Conventional stereo matching methods, such as Semi-Global Block Matching (SGBM), are prone to failure in texture-poor regions and occlusions, which are common at construction sites. Meanwhile, deep learning models, typified by PSMNet, require large-scale labeled training data that is difficult to acquire. To overcome these challenges, this research proposes a practical and robust method for generating high-fidelity 3D maps that circumvents the need for large-scale training data and compensates for the weaknesses of SGBM. The core of our proposed 3D reconstruction pipeline is a depth refinement process based on the PatchMatch MVS framework, which incorporates three key enhancements. First, it is initialized with a dense depth map and normal vectors derived directly from a high-quality stereo disparity map to establish a robust initial state. Second, instead of conventional alternating propagation, it employs a priority-based propagation scheme that expands outwards from pixels with the lowest initial depth error. This approach suppresses error propagation and enables a systematic refinement process. Third, during the random search, the search range is adaptively adjusted for each pixel according to its estimated depth error, efficiently focusing computational resources where refinement is most needed. To validate the effectiveness of our method, we utilized a synthetic dataset generated from a detailed 3D model that simulates a crane work site. A quantitative evaluation against the ground-truth data confirmed that the proposed method generates highly accurate depth maps, significantly reducing the Mean Absolute Error (MAE) from 0.41 to 0.17 compared to standalone SGBM. Our method produces a high-fidelity, dense point cloud while avoiding the heavy data requirements and potential generalization issues associated with machine learning models. This provides an essential foundation for the future development of sophisticated automated crane control systems. Keywords: Depth Map, PatchMatch MVS, Stereo Camera, Construction Stites, Stereo Matching Share Link | Plain Format | Corresponding Author (Ryunosuke Ishiguro)

8	Topic A: General Remote Sensing	ABS-45
Potential Use of Sub-250 Gram Class UAVs for Rapid and Cost-Effective Aerial Mapping: A Case Study at Walimpong Dam, Indonesia Mohammad Fuady Rusnadi (a,b)- I Gde Budi Indrawan (a)- Ferian Anggara (a) a) Geological Engineering Department, Faculty of Engineering, Gadjah Mada University b) Ministry of Public Works Abstract Topographic mapping plays a crucial role in a wide range of applications, from regional development to disaster mitigation. Aerial mapping using unmanned aerial vehicles (UAVs) has become increasingly popular due to its efficiency in covering large areas within a short time frame. However, conventional UAV-based mapping often involves high operational costs and is subject to stringent regulatory requirements, necessitating certified personnel for its operation. Recent advancements have introduced affordable sub-250 gram UAVs, which fall into a weight class with fewer regulatory constraints, that are capable of performing automated aerial mapping. This situation presents an opportunity to enhance the accessibility of aerial mapping technologies, particularly for non-specialist users. This study investigates the potential of sub-250 gram UAVs for aerial mapping by conducting a case study at the proposed site of the Walimpong Dam in Soppeng Regency, South Sulawesi Province, Indonesia. The UAV was flown autonomously in a grid pattern at a predetermined altitude. The resulting data were then compared to mapping outputs from a UAV weighing over 250 grams. The comparative analysis demonstrates that the mapping accuracy achieved by the sub-250 gram UAV closely approximates that of its heavier counterpart. These findings suggest that lightweight UAVs can serve as a viable alternative for aerial mapping tasks, potentially broadening participation and fostering inclusivity in geospatial data acquisition and mapping sciences. Keywords: Remote Sensing- Aerial Mapping- Unmanned Aerial Vehicle- Walimpong Dam Share Link | Plain Format | Corresponding Author (Mohammad Fuady Rusnadi)

9	Topic A: General Remote Sensing	ABS-49
Analysis of Disparity Accuracy Factors in Disparity Image Generated Using Moving Images Taken From A Rotating Crane Kazama Kojiro (a*), Susaki Junichi (a), Ishii Yoshie (a), Nakaoka Shohei (b) a) Department of Civil and Earth Resources Engineering, Graduate School of Engineering, Kyoto University C1, Kyoto-Daigaku Katsura, Nishikyo-Ku, Kyoto, Japan *kazama.kojiro.34n[at]st.kyoto-u.ac.jp b) Advanced Technology Development Department of Tadano Ltd., Nittamachi-cho, Takamatsu-shi, Kagawa, Japan Abstract Rapid generation of three-dimensional (3D) maps is essential for automating construction sites. One promising technique involves generating disparity images from moving images captured by a monocular camera mounted vertically downward on the tip of a crane boom. Our proposed method calculates disparity by applying Semi-Global Matching (SGM) to image pairs rectified using the camera^s rotation angle, based on the assumption of a vertically oriented camera. However, the precise conditions under which the accuracy of these disparity images varies remain poorly understood. The factors influencing accuracy are significantly more complex for a rotating crane camera compared to standard translational stereo systems. For instance, while a larger baseline to height ratio typically improves accuracy in translational setups, with a rotating crane, this benefit can be counteracted by increased vertical disparity caused by the crane^s rotational movement, which in turn degrades SGM performance. Therefore, this study aims to establish the optimal conditions for acquiring high quality disparity images considering the practical application. To establish these optimal conditions, this study employs an approach combining experiments with actual images and comprehensive simulations. First, an experiment with actual images captured at a specific radius and height revealed a key trade-off, whichi is, as the baseline-height ratio increases, the height map^s Root Mean Squared Error (RMSE) initially decreases before increasing, indicating an optimal value exists. Building upon this finding, our simulation-based analysis systematically investigates a wider range of parameters including radius, rotation angle, and height. This analysis clarifies the general conditions for achieving high accuracy by assessing their impact on RMSE, disparity completeness, and vertical disparity. A thorough understanding of these influential factors is expected to facilitate the consistent and reliable generation of highly accurate 3D maps, advancing the automation of construction processes. Keywords: photogrammetry, disparity image, quality assessment, three dimentional mapping Share Link | Plain Format | Corresponding Author (Kojiro Kazama)

10	Topic A: General Remote Sensing	ABS-58
A Comparison of Kernel-Driven BRDF Parameters Between AHI and VIIRS for Surface Reflectance Characterization Masayuki Matsuoka (a*), Hiroki Yoshioka (b), Kazuhito Ichii (c) a) Department of Information Engineering, Mie University 1577 Kurima-machiya, Tsu, 514-8507 Japan * matsuoka[at]info.mie-u.ac.jp b) Department of Information Science & Technology, Aichi Prefectural University 1522-3 Ibaragabasama, Nagakute, Aichi, 480-1198 Japan c) Center for Environmental Remote Sensing, Chiba University 1-33 Yayoi, Inage, Chiba, 263-8522 Japan Abstract Geostationary Earth orbit (GEO) satellites provide Earth observation data with a temporal resolution of several minutes throughout the day and night from a fixed position. In contrast, low Earth orbit (LEO) satellites with wide-swath sensors observe nearly the same time of day, but from different viewing angles. A bidirectional reflectance distribution function (BRDF) is a mathematical model that represents the relationship between spectral reflectance and the geometry of the sun, target, and sensor. Comparing BRDF model parameters helps integrate GEO and LEO data to characterize land surfaces. Of particular interest is whether the same parameters can represent the reflectance of both types of sensors. This study compared the kernel-driven BRDF model parameters of the Advanced Himawari Imager (AHI) sensor on a GEO satellite and the Visible Infrared Imaging Radiometer Suite (VIIRS) sensor on a LEO satellite. The target area was Japan. The AHI was analyzed using six hours of diurnal time series data (145 scenes). These data were corrected for the atmospheric effects and reprojected to a latitude-longitude projection. For VIIRS, the ^BRDF/Albedo Model Parameters^ products were used for both the National Polar-orbiting Partnership (Suomi-NPP) and the Joint Polar Satellite System 1 (JPSS-1). Comparing three BRDF parameters (fiso, fvol, and fgeo) revealed a remarkable terrain-dependent pattern in the AHI data. This pattern was caused by shadows resulting from the sun^s movement throughout the day. VIIRS showed a clear dependence on land cover. The different features of the GEO-LEO BRDF model parameters provide useful information for characterizing land surfaces, such as albedo and FPAR. This study also helps to develop an accurate BRDF model by integrating reflectance observed under different observation geometries. Keywords: bidirectional reflectance distribution function, geostationary earth orbit, low earth orbit, kernel-driven BRDF model, observation geometry Share Link | Plain Format | Corresponding Author (Masayuki Matsuoka)

11	Topic A: General Remote Sensing	ABS-66
Detecting Land Cover Changes Using Multi-Temporal Radar Imagery: A Case Study of Taiwan^s Western Coastal Region Ting-Yu Lai 1*, Kuo-Hsin Tseng 1,2 1 Department of Civil Engineering, National Central University 2 Center for Space and Remote Sensing Research, National Central University Abstract Satellite remote sensing data have been widely used to identify surface coverage- however, traditional optical images are limited in monitoring long-term surface changes due to cloud cover and atmospheric moisture interference. In contrast, Synthetic Aperture Radar (SAR) is less affected by atmospheric conditions and sunshine, and is more sensitive to surface texture changes, which can complement optical satellite imagery by providing long-term and stable observation of surface changes. Therefore, in this study, the C-band Sentinel-1 GRD images from 2016 to 2024 and the X-band TerraSAR-X data from 2021 to 2025 were collected in the western coastal area of Taiwan to observe the temporal changes of features and water bodies. The study is divided into two parts: land and coastal areas. Firstly, all the images are using SNAP software for a standardized preprocessing workflow, and then the land section involves stacking multi-temporal data and composing false-color images using selected polarization bands from specific periods to observe changes in urban infrastructure and agricultural activities. For the coastal area, three to five low-tide images are selected every year based on the tide level data, and the average backscatter intensity values are calculated. These annual averages are assigned to different channels of pseudo-color images. The results reveal erosion and sedimentation hotspots along the west coastline of Taiwan. By integrating multi-temporal radar data, this study provides a foundational basis for future applications such as land classification, change detection, and spatial planning. Keywords: Synthetic Aperture Radar (SAR), Sentinel-1, TerraSAR-X, Change Detection Share Link | Plain Format | Corresponding Author (TING YU LAI)

12	Topic A: General Remote Sensing	ABS-68
Spatio-Temporal Analysis of Savanna Changes and its Correlation with Deer and Bull Populations in Baluran National Park Himayah, S., Somantri, L., Fadhilah, A., and Sulastri, A. Geography Information Science, Faculty of Social Science Education, Universitas Pendidikan Indonesia Abstract Baluran National Park, known as ^Little Africa in Java,^ is a rich conservation area featuring diverse ecosystems, including savanna. The savanna ecosystem within this national park faces a serious threat of land cover change from savanna to non-savanna. This change is primarily driven by the invasion of species like acacia, which was initially planted in the 1960s for forest and land fire mitigation. This invasion not only disrupts the original ecosystem but also threatens the survival of key herbivores such as deer and bull, which heavily rely on savanna vegetation as their primary food source. To address this issue, this research aims to: (1) analyze the changes of savanna vegetation using remote sensing technology, and (2) analyze the correlation between these changes and the population dynamics of deer and bull in Baluran National Park. The methods employed include the Random Forest algorithm for land cover change identification, the Modified Soil-Adjusted Vegetation Index (MSAVI) for savanna vegetation density identification, and linear regression for analyzing the correlation. Preliminary results indicate that the savanna land cover area has continuously narrowed over the years, correlating with changes in deer and bull populations. This research is expected to enrich the application of remote sensing in savanna studies and support Sustainable Development Goal (SDG) number 15 regarding the restoration and sustainable management of terrestrial ecosystems. Keywords: remote sensing, Baluran National Park, land cover change, savanna, deer and bull populations. Share Link | Plain Format | Corresponding Author (Shafira Himayah)

13	Topic A: General Remote Sensing	ABS-69
Mapping Benthic Habitat Composition and Estimating Seagrass Percent Cover Using Stepwise and Machine Learning Regression Methods: A Case Study from Kwandang Bay, North Gorontalo, Indonesia Setiawan Djody Harahap (a*), Huwaida Nur Salsabila (b), Abhista Fawwaz Sahitya (c), Jennifer Wijaya (c), Safina Rajwaa Ananda (d), Pramaditya Wicaksono (e), Nurul Khakhim (e), Muhammad Kamal (e), Prima Widayani (e), Muhammad Banda Selamat (f) a) Master in Remote Sensing, Faculty of Geography, Universitas Gadjah Mada, Sekip Utara, Sleman, Daerah Istimewa Yogyakarta 55281, Indonesia *setiawandjody99[at]mail.ugm.ac.id b) Department of Geography, Faculty of Mathematics and Natural Sciences, Universitas Negeri Makassar, Makassar, South Sulawesi 90224, Indonesia c) Blue Carbon Research Group, Faculty of Geography, Universitas Gadjah Mada, Sekip Utara, Sleman, Daerah Istimewa Yogyakarta 55281, Indonesia d) Cartography and Remote Sensing, Faculty of Geography, Universitas Gadjah Mada, Sekip Utara, Sleman, Daerah Istimewa Yogyakarta 55281, Indonesia e) Faculty of Geography, Universitas Gadjah Mada, Sekip Utara, Sleman, Daerah Istimewa Yogyakarta 55281, Indonesia f) Faculty of Marine Science and Fisheries, Hasanuddin University, Perintis Kemerdekaan km.10 Tamalanrea, Makassar, South Sulawesi 90245, Indonesia Abstract Understanding the spatial distribution of benthic habitats and seagrass ecosystems is essential for effective coastal ecosystem management. Furthermore, spatial information on seagrass percent cover (PCv) is an important parameter for monitoring the condition and health of seagrass meadows, as it reflects indicators of seagrass abundance that serve as measurable proxies for evaluating ecosystem resilience. This study presents a comprehensive approach to mapping benthic habitat composition and estimating seagrass PCv in Kwandang Bay, North Gorontalo, Indonesia. Using Sentinel-2 imagery, this study employed Random Forest classification to generate the benthic habitat composition map, while the seagrass PCv map was estimated using three regression techniques: Stepwise Regression (SWR), Random Forest Regression (RFR), and Support Vector Machine Regression (SVR). The field data used in this study were collected using a georeferenced photo-transect method. The performance of the benthic composition map was assessed using a confusion matrix, while the performance of each seagrass PCv model was evaluated using standard accuracy metrics, including the coefficient of determination (\(R^{2}\)), root mean squared error (RMSE), and 1:1 graphical plot to assess prediction reliability. This study also aims to compare the performance of classical and machine learning models in mapping seagrass PCv. This study underscores the potential of both traditional and modern analytical techniques for mapping seagrass PCv, where the resulting benthic composition map and seagrass PCv estimations provide a valuable spatial baseline for monitoring coastal ecosystems and seagrass dynamics, thereby supporting conservation strategies in Kwandang Bay Keywords: Benthic habitat mapping, Seagrass percent cover mapping, Random Forest, Stepwise regression, Support Vector Machine regression, Sentinel-2 Share Link | Plain Format | Corresponding Author (Setiawan Djody Harahap)

14	Topic A: General Remote Sensing	ABS-326
Tropical SMAP Soil Moisture Validation using Triple Collocation: Analysing the error Chan, F.C. (a) , Mohd Reba, M.N. (a*) a) Faculty of Built Environment and Surveying, Universiti Teknologi Malaysia (UTM), 81310 UTM Johor Bahru, Johor, Malaysia *nadzri[at]utm.my Abstract Accurate soil moisture (SM) estimation is critical for understanding the land-atmosphere interactions, hydrological processes, and climate dynamics. Yet, the accurate SM retrieval in tropical areas remains challenging due to the heterogeneity of land cover and precipitation dynamics. Conventionally, satellite-retrieved SM products are validated against the collocated in-situ SM measurements, which are often assumed as the ground truth. However, the spatial representation error owing to the point-scale in situ measurements and areal-scale satellite footprint can degrade the reliability of in-situ measurements as a benchmark in the validation. Triple Collocation-based validation is found to minimize the scale issue between grid-scale satellite SM and point-scale in-situ SM. This study aims to expand the validation of SMAP SM products beyond conventional in-situ comparisons by employing TCA in a tropical environment. The objectives are twofold: (1) to evaluate the error characteristics of SMAP- (2) to investigate the potential errors introduced by precipitation and vegetation density in SMAP products. The study attempts to minimize the confounding factors related to surface heterogeneity by focusing on a homogeneous land cover. This study includes the hourly in-situ measurements, SM products from Soil Moisture Active Passive (SMAP) at 9 km spatial resolution, and modelled SM products from ERA5-Land reanalysis from March 2024 to March 2025 in a homogeneous tropical agricultural site. Results show SMAP demonstrates moderate correlation (R = 0.406) but a negative bias (-0.022m3/m3), and higher ubRMSE (0.0337m3/m3), attributed to signal attenuation under dense vegetation and rapid moisture fluctuations. Precipitation intensifies SMAP errors (ubRMSE peaks at 0.065m3/m3) during heavy rainfall. This study affirms the utility of TCA for SM validation in tropical areas and provides insights into the performance of SMAP in understudied tropical regions. The findings aim to contribute to the development of SM retrieval algorithms. Keywords: SMAP, soil moisture, passive microwave remote sensing, triple collocation analysis, tropics Share Link | Plain Format | Corresponding Author (Fong Chi Chan)

15	Topic A: General Remote Sensing	ABS-327
VALIDATION STRATEGY FOR SATELLITE-DERIVED SOIL MOISTURE IN TROPICAL REGION Chan Fong Chi (a) and Mohd Nadzri Md Reba (a*) a) Faculty of Built Environment and Surveying, Universiti Teknologi Malaysia (UTM), 81310 UTM Johor Bahru, Johor, MALAYSIA- Abstract Continuous and distributed in-situ soil moisture (SM) measurements are crucial in the calibration and validation of Soil Moisture Active Passive (SMAP) data. The existing public database of the International Soil Moisture Network (ISMN) in tropical regions compared to other climate zone, thus impeding the validation of the SMAP products in tropical regions. Several data records in tropical regions includes MySMnet in Malaysia and VDS network in Vietnam. Additionally, the tropical climate experiences frequent rainfall and higher humidity in the dense vegetation which poses distinct challenges to the calibration of satellite SM products. Such dynamics contribute to inconsistency in field data and are increase the diurnal and nocturnal variation of surface-to-atmosphere interactions. Therefore, the objectives of this research are twofold: first, to compare the correlation of the SMAP products with- (a) in-situ measurements different upscaling techniques on the, (b) in-situ measurement at different temporal integrations- and second, to validate the collocated SMAP SM with the in-situ SM. To address the scarcity of in-situ SM data in tropical regions, the research has deployed a MIE soil moisture network six stations and a weather station to collect soil moisture, temperature and humidity data hourly from May to November 2024. The in-situ measurement from MySMnet, VDS network and MIE network has been upscaled to the SMAP-scale using Voronoi Diagram, arithmetic average- while analysis for temporal variability by using median temporal integration was performed to compare with SMAP within a 2-hour, 3-hour, 4-hour proximity in both AM and PM variants. Statistical metrics, like unbiased Root Mean Square Error (ubRMSE), are utilised to evaluate the temporal variability and consistency between both data. The results reveal differences in soil moisture levels between. This study is expected to provide insights into the strategy of calibration and validation of SMAP in tropical environments. Keywords: soil moisture, SMAP, passive microwave, remote sensing, temporal resolution Share Link | Plain Format | Corresponding Author (Fong Chi Chan)

16	Topic A: General Remote Sensing	ABS-73
Benchmarking Phase Filtering Techniques for Coherence Enhancement and Persistent Scatterer Selection in PSInSAR Yan Akhbar Pamungkas, Shou Hao Chiang Center for Space and Remote Sensing Research, National Central University, Taiwan Department of Urban and Regional Planning, Universitas Brawijaya, Indonesia Abstract The accuracy of Persistent Scatterer Interferometric Synthetic Aperture Radar (PSInSAR) relies on effective coherence estimation to identify optimal Persistent Scatterer Point Candidates (PSCs). This study aims to evaluate and compare the performance of various phase filtering techniques, namely Boxcar, Goldstein, GAMMA, and Lee Filtering, with no filtering as the baseline, to compute the Spatial and Temporal Coherence for PSCs selection. A total of 173 Sentinel 1A Single Look Complex (SLC) images in ascending orbit with IW2 subswath and VV polarization images over Jakarta Capital City, Indonesia, were used. The processing workflow is composed of orbit correction, coregistration, wrapped interferogram generation, coherence estimation, and PSCs mask generation under different filtering scenarios. The results demonstrate significant variations in coherence performance among phase filtering methods, with spatial coherence consistently outperforming temporal coherence (spatial mean: 0.41 - 0.76 & temporal mean: 0.07 - 0.12). GAMMA filtering yields the highest spatial coherence (mean = 0.76), indicating its superior ability to suppress speckle while preserving coherent signals in urban environments. In contrast, the no-filter scenario produces the lowest coherence values, highlighting the essential of filtering in PSInSAR processing. Goldstein and Lee filters deliver reduced spatial coherence (means = 0.42 and 0.46), resulting in less detail and accuracy in persistent scatterer identification. Notably, Boxcar filtering achieves a spatial coherence (mean = 0.72) comparable to GAMMA, suggesting selective coherence preservation but with potential trade-offs in spatial detail. These findings underscore the critical influence of filter selection on the quality of coherence computation and have significant implications for enhancing PSInSAR accuracy in urban deformation studies. Future work will focus on comparing the results with unwrapped and cleaned phase as the input. Keywords: PSInSAR, Coherence Estimation, Phase Filtering, Persistent Scatterer, Sentinel-1 Share Link | Plain Format | Corresponding Author (Yan Akhbar Pamungkas)

17	Topic A: General Remote Sensing	ABS-335
High-Resolution PM2.5 Modeling and Respiratory Disease Correlation in Chiang Mai Using Himawari-9 and Ground Observations Katanchalee Taweepornwattanakul and Wataru Takeuchi Institute of Industrial Science, The University of Tokyo, Japan Abstract Haze pollution driven by fine particulate matter (PM2.5) is a recurring environmental and health crisis in Thailand, especially during the dry season. In 2023, Chiang Mai endured one of its most persistent haze events in recent years, with sustained PM2.5 concentrations affecting both residents and tourism. While national air quality trends over the past decade (2015-2024) indicate gradual improvement, annual averages remain above the World Health Organization (WHO) guideline, and seasonal peaks continue to impose acute health risks. This study develops a high-temporal-resolution PM2.5 estimation model by integrating Himawari-9 aerosol optical depth (AOD, Level 2, 10-minute resolution) with ground-based PM2.5 measurements from the Pollution Control Department. Meteorological parameters-air temperature, relative humidity, rainfall, wind speed, and surface pressure-are included as covariates to improve model accuracy. All datasets undergo quality control, spatial reprojection, and subsetting to the Chiang Mai domain before model calibration. Estimated PM2.5 values will be aggregated to monthly means and correlated with monthly respiratory disease statistics, including major conditions such as chronic obstructive pulmonary disease (COPD), asthma, and pneumonia. Using an event-focused ecological design with interrupted time-series analysis, haze months (February-April 2023) will be compared with non-haze months, adjusting for meteorology and seasonality. This preliminary study represents the first integration of high-frequency Himawari-9 AOD, ground PM2.5 data, and monthly respiratory health records in Northern Thailand, offering a scalable framework and initial evidence to support targeted interventions and future multi-year investigations. Keywords: Himawari-9, PM2.5, AOD, Ground observations, Respiratory health Share Link | Plain Format | Corresponding Author (Katanchalee Taweepornwattanakul)

18	Topic A: General Remote Sensing	ABS-339
Random Forest-Based Flood Hazard Modeling: Analysis of the Impact of Population Growth and Land Cover Change in the Takkalasi Watershed, South Sulawesi Nur Dwiyanti Utari (a*), Roland Alexander Barkey (b), Andang Suryana Soma (b) a) Department Of Regional and Development Planning, Graduate Student, Hasanuddin University. *nurdwiyantiutari[at]gmail.com b) Faculty of Forestry, Hasanuddin University. Abstract Rapid population growth has triggered massive land cover changes, particularly through the conversion of natural areas such as forests and agricultural lands into settlements. These changes significantly reduce the soil^s ability to absorb rainwater, increase surface runoff, and exacerbate flood risks. This study develops a flood vulnerability assessment model using the Random Forest algorithm, considering influencing factors such as rainfall, topography, land cover, drainage capacity, geology, and river proximity. The training data was constructed from historical flood event datasets complemented by spatial predictor variables. The model achieved an accuracy of 85% based on cross-validation, with land cover and rainfall intensity being the most significant predictors. The objective of this study is to analyze the relationship between population growth, land cover changes, and the increased frequency and intensity of floods, using the Random Forest (RF) machine learning algorithm for flood hazard mapping in the Takkalasi Watershed, South Sulawesi, Indonesia. The results indicate that rainfall in upstream and downstream areas, flat topography in downstream regions, inadequate drainage capacity, and land cover changes are the dominant factors determining the extent and depth of flooding. The integrated approach developed in this study offers an efficient method for flood risk mapping by utilizing available spatial data, with the potential for application in other watershed areas. In practice, these findings can support the development of risk-based spatial planning and the formulation of effective disaster mitigation strategies, particularly in regions with limited hydrological data availability. Keywords: Flood modeling, land cover changes, Random Forest, Takkalasi Watershed Share Link | Plain Format | Corresponding Author (Nur Dwiyanti Utari)

19	Topic A: General Remote Sensing	ABS-85
Geometric Accuracy Improvement of Geostationary Environment Monitoring Spectrometer by Pixel Offset Adjustment Seunghyeok Choi (a), Taejung Kim (b*) a) Program in Smart City Engineering, Inha University 100 Inha-ro, Incheon 22212, Republic of Korea b) Dept. of Geoinformatic Engineering, Inha University 100 Inha-ro, Incheon 22212, Republic of Korea *tezid[at]inha.ac.kr Abstract The Geostationary Environment Monitoring Spectrometer (GEMS) is the world^s first geostationary hyperspectral environmental satellite launched in 2020. It carries out air quality monitoring over the East Asian region by collecting hyperspectral images with spatial resolution of around 7 kms. However, due to factors such as sensor misalignment, attitude instability, and thermal deformation, positional errors can occur in GEMS images. These errors degrade the reliability of products and cause difficulty in time-series analyses and fusion with other satellite data. In this study, we propose a method to evaluate and correct the geometric accuracy of GEMS images using reference images with spatial resolution of 1 km from the Advanced Meteorological Imager (AMI), a geostationary meteorological satellite. First, spatial collocation is performed between GEMS and AMI images to match their spatial resolution and acquisition time. Then, the collocated GEMS image is shifted within a \pm5 pixel range in vertical and horizontal directions to perform global matching. Cross-correlation is computed for each shift condition and the shift with the highest correlation is defined as the optimal offset. This offset is applied to the GEMS image. Finally, correlation is recalculated on the adjusted image to evaluate geometric accuracy before and after correction. The experiment was carried out using daily collected GEMS and AMI images from January 1, 2023 to May 31, 2025. Results showed that average pixel root mean squared error (RMSE) decreased from 1.36 pixels to 0.69 pixels. In most cases, geometric errors were reduced to within 1 pixel. These results confirmed that geometric accuracy of GEMS could be effectively improved by ptimal pixel shift offsets derived from the matching between GEMS and AMI images. Keywords: Satellite Image- Geometric Correction- GEMS- AMI- Image Matching Share Link | Plain Format | Corresponding Author (Seunghyeok Choi)

20	Topic A: General Remote Sensing	ABS-344
Characterizing Burned Savanna Signatures from Sentinel-1 SAR Backscatter and H-Alpha Dual-Pol Decomposition: Case Study of the 2023 Bromo Fire Athar Abdurrahman Bayanuddin, Retnadi Heru Jatmiko, Nur Muhammad Farda Remote Sensing Graduate Study Program, Faculty of Geography, Universitas Gadjah Mada, Indonesia. Directorate of Laboratory Management, Research Facilities, and Science and Technology Park, National Research and Innovation Agency (BRIN), Indonesia. Department of Geographic Information Science, Faculty of Geography, Universitas Gadjah Mada, Indonesia Abstract Forest and land fires in Indonesia have increasingly shifted toward Eastern Indonesia, significantly impacting extensive savanna ecosystems. Synthetic Aperture Radar (SAR) imagery offers promise as a complementary or alternative approach for burned area mapping, yet its application in Indonesian savannas remains limited. This study investigates changes in backscatter and polarimetric decomposition (H-Alpha) parameters over burned savanna areas, using pre- and post-fire Sentinel-1 C-band data from the 2023 Bromo fire. Backscatter (VV, VH) and H-Alpha decomposition values were extracted from samples representing burned low-vegetation and tree-dominated areas. Change values (delta (d)) were calculated and assessed for statistical significance using a two-tailed Wilcoxon signed-rank test. The influence of slope-relative orientation and pre-fire Enhanced Vegetation Index (EVI) on these changes was also examined. Results show statistically significant changes in both backscatter and H-Alpha parameters for burned areas. In low-vegetation areas, Entropy (H) (d median = &#8722-0.0049, p < 0.05), Alpha (d median = &#8722-0.3841 degrees, p < 0.05), and VV and VH backscatter (d median = &#8722-0.1708 dB and 0.3305 dB, respectively, p < 0.05) all changed significantly. In tree-dominated areas, Alpha (d median = &#8722-1.1986 degrees, p < 0.05) and VH backscatter (d median = &#8722-1.0518 dB, p < 0.05) showed strong declines, while VV backscatter changes were not significant. Both topography and pre-fire EVI influenced the magnitude of changes, with distinct patterns across vegetation types-particularly a decreasing trend in delta H and delta Alpha with increasing pre-fire EVI in tree-covered areas. This study confirms the capability of Sentinel-1 SAR imagery to detect and characterize burn scars in savanna landscapes, revealing distinct post-fire signatures for different vegetation types. These findings provide a foundation for enhancing fire-mapping algorithms tailored to tropical savannas, with particular relevance to Eastern Indonesia. Keywords: Sentinel-1 SAR, H-Alpha decomposition, Polarimetric SAR, Savanna fires, Burned area mapping Share Link | Plain Format | Corresponding Author (Athar Abdurrahman Bayanuddin)

21	Topic A: General Remote Sensing	ABS-90
Super-Resolution of Advanced Himawari Imager Data Using SRCNN Yohei Kato(a),Masayuki Matsuoka(a*) a) Department of Information Engineering, Mie University, 1577 Kurima-machiya,Tsu, 514-8507 *matsuoka[at]info.mie-u.ac.jp Abstract Recent advancements in deep learning have enabled notable improvements in satellite imagery resolution through super-resolution techniques. This study focuses on enhancing the spatial resolution of Himawari Advanced Imager (AHI) data using a convolutional neural network, specifically the Super-Resolution Convolutional Neural Network (SRCNN). The objective is to generate high-resolution images from their low-resolution ones while preserving key structural and radiometric information. The dataset consists solely of Himawari imagery, covering six spectral bands: B01, B02, B04, B05, B06, and B15. High-resolution (HR) and low-resolution (LR) image pairs were created and divided into smaller tiles based on original image sizes for model training and evaluation. The SRCNN model was trained to learn the mapping from LR to HR features. To assess super-resolution performance, several image quality metrics were employed. Global metrics such as Peak Signal-to-Noise Ratio (PSNR) and Structural Similarity Index Measure (SSIM) were used to evaluate fidelity and structural consistency. Brightness and contrast were also analyzed for radiometric evaluation. Furthermore, Local Contrast was introduced to assess the preservation of local details and textures. The results showed that the SRCNN-based method improved both visual and quantitative image quality across all bands. Comparisons for each band indicated that enhancement effectiveness varied depending on spectral characteristics. These findings suggest future improvements could be achieved through network optimization and band-aware training strategies. This research demonstrates the feasibility of applying deep learning-based super-resolution to Himawari satellite images. It holds potential for enhancing applications such as weather monitoring, disaster assessment, and environmental observation. In particular, generating high-resolution data from existing images can support faster, more accurate analysis in time-critical scenarios like natural disasters. Keywords: CNN-based super resolution, image quality evaluation, spectral band Share Link | Plain Format | Corresponding Author (Yohei Kato)

22	Topic A: General Remote Sensing	ABS-351
Using Remote Sensing for Mapping Land Use/Land Cover of The City of Baubau, Southeast of Sulawesi Rini Anggraini, S Baja, R Neswati 1 Post Graduate Student of Regional Planning and Development, Hasanuddin University, Makassar, Indonesia 2 Department of Soil Science, Hasanuddin University, Makassar, Indonesia Abstract Land use and land cover (LULC) mapping is a crucial instrument in spatial planning and environmental management, particularly in urban areas with high spatial dynamics. This study aims to update the LULC map of BauBau City (study area 28,619 ha) using Landsat 8 imagery (September 2024). The method employed is supervised classification using the Maximum Likelihood Classification (MLC) algorithm, supplemented by manual interpretation to enhance classification accuracy. The research process includes data preprocessing (geometric, radiometric, and atmospheric corrections), image classification, and accuracy testing using the stratified random sampling approach at 300 reference points. The initial classification results identified five land cover classes, which were then refined through manual interpretation to produce five main classes: forest, agriculture, built up land, open land, and water bodies. The evaluation yielded an overall accuracy of 91.7% and a Kappa coefficient of 0.843, indicating an extreme level of classification suitability for field conditions. This combined approach has proven effective in enhancing the spatial and thematic representation of LULC mapping and can support sustainable, data-driven urban development planning. Keywords: LULC, Landsat 8, MLC classification, manual interpretation, spatial accuracy Share Link | Plain Format | Corresponding Author (Rini Anggraini)

23	Topic A: General Remote Sensing	ABS-96
Refractive-Aware Gaussian Splatting for Geometrically Accurate and Photorealistic 3D Reconstruction of Bathymetry from Aerial Imagery Taiki Uno (a*), Sohei Kobayashi (b) a) Graduate School of Engineering, Kyoto University Nishikyo-ku, Kyoto 615-8540, Japan *uno.taiki.65r[at]st.kyoto-u.ac.jp b) Disaster Prevention Research Institute, Kyoto University Gokasho, Uji, Kyoto, 611-0011, Japan Abstract The 3D reconstruction of geometry in shallow water has a wide range of applications, such as monitoring bed morphology changes and conducting hazard simulations. In addition, the detailed texture of the reconstructed data is useful for bed surface classification and aquatic habitat quantification. In this context, bathymetric surveying using photogrammetry from aerial images captured by Unmanned Aerial Vehicles (UAV) is gaining attention as a particularly efficient method. However, it is fundamentally challenged by light refraction at the air-water interface, which invalidates the geometric principles of photogrammetry. Existing methods either rely on iterative post-processing or employ deep learning models that lack physical guarantees and explainability. We solve this challenge by introducing a refraction-aware 3D Gaussian Splatting framework that incorporates an optically accurate model of two-media refraction directly into the reconstruction pipeline. Our key innovation is a differentiable coordinate transformation that analytically models light refraction, mapping 3D Gaussians from their true underwater positions to their apparent space for each aerial view. This enables end-to-end optimization, simultaneously solving for dense scene geometry and detailed appearance while maintaining the efficiency of standard 3D Gaussian Splatting. We evaluated our method on a simulated UAV dataset of a riverbed, rendered with physically-based ray tracing to isolate refractive effects from other optical phenomena. Our approach achieved a geometric F1-score of 96% (with a 10 cm error threshold at a depth scale of 10 m). Furthermore, in novel view synthesis, we obtained photorealistic views with a Peak Signal-to-Noise Ratio (PSNR) of 25.9 dB and a Structural Similarity Index Measure (SSIM) of 0.93. By creating 3D models that are both photorealistic in appearance and dense and geometrically precise in structure, our method addresses a key challenge in aquatic remote sensing from aerial imagery. This work enables cost-effective, high-frequency monitoring of riverbeds, lakeshores, and seashores under calm surface conditions. Keywords: Bathymetry- Gaussian Splatting- 3D Reconstruction- Two-Media Photogrammetry- Refraction Correction Share Link | Plain Format | Corresponding Author (Taiki Uno)

24	Topic A: General Remote Sensing	ABS-98
Enhancing Visual SLAM Accuracy in Low-Light Environments via Image Enhancement Techniques Yusuke Ito (a) , Masayuki Matsuoka (a*) a) Department of Information Engineering, Mie University 1577 Kurima-machiya, Tsu, Mie 514-8507, Japan *matsuoka[at]info.mie-u.ac.jp Abstract The advancement of autonomous driving technologies has intensified the need for accurate self-localization and real-time environmental mapping. Simultaneous Localization and Mapping (SLAM) addresses this need by enabling robots or vehicles to build a map of an unknown environment while simultaneously determining their position within it. Among various types of SLAM, Visual SLAM (VSLAM) is notable for its low-cost implementation with monocular or stereo cameras. However, VSLAM systems are seriously limited in low-light environments, where the number of detectable visual features declines significantly, leading to reduced localization accuracy and potential tracking failure. This research aims to improve VSLAM performance under such challenging lighting conditions by integrating image enhancement modules into the VSLAM pipeline. In particular, we explore deep learning-based enhancement techniques, including those using Generative Adversarial Networks (GANs), to preprocess input images and improve feature visibility. The Oxford RobotCar Dataset is used, which includes sequences captured under varying illumination conditions. We compare conventional VSLAM with enhanced versions that incorporate di&#64256-erent image enhancement methods. Performance is evaluated based on two primary metrics: the number of extracted feature points and the computational time required for processing. Experimental results demonstrate that the enhanced VSLAM systems consistently outperform the baseline in low-light environments, showing increased feature point extraction without significant degradation in execution speed. These findings suggest that image enhancement can serve as a viable solution for improving VSLAM robustness in visually degraded settings. The results of this study have potential applications in autonomous vehicles operating at night, as well as in robotic vision systems deployed in low-visibility environments such as tunnels, disaster zones, or outer space. Future work will focus on optimizing these systems for real-time performance and further enhancing localization accuracy by incorporating deep learning techniques into feature extraction and matching processes. Keywords: Simultaneous Localization and Mapping, Generative Adversarial Networks, autonomous driving car, illumination condition Share Link | Plain Format | Corresponding Author (Yusuke Ito)

25	Topic A: General Remote Sensing	ABS-361
Mapping the Dynamics of Vegetation Density Using the NDVI Index in Palopo City (2019-2023) Muh. Azriel Putra Rizal Department Of Geography, Universitas Negeri Makassar Abstract Rapid population growth and increasing land demand for housing and economic activities have intensified the reduction of green open spaces, particularly in medium-sized cities that are rarely studied compared to large metropolitan areas. This research addresses that gap by examining vegetation dynamics in Palopo City, South Sulawesi, from 2019 to 2023 using the Normalized Difference Vegetation Index (NDVI) combined with a descriptive-qualitative approach. The analysis reveals a significant decline in vegetation density, particularly in peri-urban areas where residential and commercial expansion has accelerated. Unlike most previous studies that focus on major urban centers in Java, this study provides new insights into vegetation changes in an underrepresented medium-sized city, emphasizing the role of urbanization and population mobility in driving land-use conversion. The novelty of this research lies in its multi-year temporal observation of NDVI in Palopo, which enriches the empirical understanding of vegetation dynamics in Eastern Indonesia while highlighting the urgent need to integrate environmental considerations into urban spatial planning. The findings are crucial for policymakers and urban planners to balance socio-economic growth with ecological sustainability, ensuring that urban development does not irreversibly compromise green infrastructure and ecosystem services. Keywords: NDVI- Sattelite- Temporal- Sentinel 2A- vegetation dynamics Share Link | Plain Format | Corresponding Author (Muh. Azriel Putra Rizal)

26	Topic A: General Remote Sensing	ABS-107
Passive remote sensing of marine liquid cloud geometric thickness using the O2-O2 band: first results from TROPOMI Wenwu Wang1,2, Chong Shi1*, Jian Xu3, Shuai Yin1, Huazhe Shang1, Yutong Wang1,2, Chenqian Tang1, Ruijie Yao1,2, Guangyu Shi4, Husi Letu1* 1Key Laboratory of Remote Sensing and Digital Earth, Aerospace Information Research Institute, Chinese Academy of Sciences. Beijing 100101, China. 2University of Chinese Academy of Sciences, Beijing 100049, China. 3National Space Science Center, Chinese Academy of Sciences, Beijing 100190, China. 4State Key Laboratory of Numerical Modeling for Atmospheric Sciences and Geophysical Fluid Dynamics, Institute of Atmospheric Physics, Chinese Academy of Sciences, Beijing 100029, China. Abstract Observations on cloud geometric thickness are crucial for understanding the radiative balance and aerosol indirect radiative effects, and currently, cloud geometric thickness retrieval studies for passive instruments remain constrained due to the lack of the understanding of the incident radiation penetrability. In this work, we firstly analyse the relationship between the cloud droplets distribution and the incident radiation penetrability based on physical model, and then fully utilize the advantages of hyperspectral O4 measurements to build a physically-based machine learning model to retrieve the cloud geometric thickness. The algorithm retrieves cloud geometric thickness from TROPOMI observations for the first time, and the retrievals are compared with the cloud geometric thickness from active observations. It is found that the mean absolute error of the retrievals using 2B-CLDPROF-LIDAR cloud-top height as input is 0.49 km, which shows the potential of O4 band to retrieve cloud geometric thickness. Keywords: hypespectral, cloud geometric thickness, O4 band Share Link | Plain Format | Corresponding Author (Wenwu Wang)

27	Topic A: General Remote Sensing	ABS-109
Accuracy Assessment of Multi-Temporal Intertidal DEM in Taiwan Using ICESat-2 Ying-Cih Shih(a), Kuo-Hsin Tseng(a,b) a) Department of Civil Engineering, National Central University, Taiwan b) Center for Space and Remote Sensing Research, National Central University, Taiwan Abstract The steep eastern terrain and gentle western slopes of Taiwan have shaped extensive sandy coastlines and numerous wetlands along the western shore. Situated in a subtropical region with variable climatic conditions, the coastlines are frequently affected by typhoons, tides, and coastal currents, resulting in continual shoreline changes. These factors highlight the importance of timely observation and analysis of intertidal zones. To capture such dynamic changes, this study utilizes Sentinel-2 multispectral imagery from 2018 to 2024 to reconstruct intertidal topography and applies ICESat-2 laser altimetry data to validate elevation accuracy. The objective is to identify the optimal time span for reconstruction and determine the most suitable temporal scale for intertidal monitoring in Taiwan. The study area is located at the Dadu River Estuary Wetland. ICESat-2 ATL03 point cloud data from June 2020 to January 2024 were collected, with the lowest-tide observation used as the starting point for reconstructing intertidal topography across various time scales. A Modified Normalized Difference Water Index (MNDWI) was first applied to delineate the land-water boundary. Then, inundation probabilities were calculated and combined with tidal models to estimate elevation values. The reconstructed intertidal topography was validated using ICESat-2 elevation profiles. Results show that RMSE values across different observation periods range from 0.3 to 0.7 meters, with optimal accuracy achieved at the nine-month period. These findings demonstrate that the proposed method effectively evaluates reconstruction performance across varying time spans and helps identify the optimal temporal scale for intertidal monitoring in Taiwan, offering valuable guidance for future wetland and coastal management. Keywords: ICESat-2, Sentinel-2, Dadu Estuary, intertidal topography, time-series analysis Share Link | Plain Format | Corresponding Author (YING CIH SHIH)

28	Topic A: General Remote Sensing	ABS-111
A Comparison of SuperPoint + SuperGlue and SIFT-Based Pipelines for Sparse-View Structure-from-Motion Shoya Morizaki(a), Masayuki Matsuoka (a*) a) Department of Information Engineering, Mie University 1577 Kurima-machiya, Tsu, 514-8507 Japan * matsuoka[at]info.mie-u.ac.jp Abstract Structure-from-Motion (SfM) is a widely used technique for reconstructing three-dimensional structures from two-dimensional images, with applications in cultural heritage preservation, urban planning, and environmental monitoring. However, its performance often deteriorates under sparse-view conditions where image coverage is limited by restricted viewpoints or flight paths, such as in UAV-based surveys or narrow indoor environments. To address this challenge, we explored the integration of machine learning-based keypoint detection and matching methods-SuperPoint and SuperGlue-into the SfM workflow, and compared their performance with that of conventional SIFT-based pipelines implemented in OpenMVG. In this approach, SuperPoint was used to extract robust keypoints, while SuperGlue performs context-aware feature matching. The resulting matching points were then passed to OpenMVG for geometric verification, incremental reconstruction, and camera pose estimation. Our comparative study evaluated three configurations: (1) a fully traditional SIFT-based pipeline, (2) a hybrid approach combining SuperPoint and OpenMVG, and (3) a deep-learning-based pipeline combining SuperPoint, SuperGlue, and OpenMVG. The evaluation focused on reconstruction success rate, point cloud density, and camera pose accuracy under varying conditions of image overlap and viewpoint sparsity. Initial trials suggested that the deep learning-based approach may offer improved performance in challenging conditions involving occlusion and limited image data. This study aims to contribute to a deeper understanding of how recent advances in feature extraction and image matching algorithms can be integrated into classical SfM frameworks to enhance robustness and accuracy, especially in scenarios constrained by sparse visual input. Keywords: Structure-from-Motion, SuperPoint, SuperGlue, Learned Features, Sparse-view Reconstruction Share Link | Plain Format | Corresponding Author (Shoya Morizaki)

29	Topic A: General Remote Sensing	ABS-132
Precise Building Boundary Extraction Task Using Deep Learning method Lin Y. H.1, Tsai F.2* 1Student, Department of Civil Engineering, National Central University, Taiwan 2*Professor, Center for Space and Remote Sensing Research, National Central University, Taiwan 0123kan[at]gmail.com, *ftsai[at]csrsr.ncu.edu.tw Abstract Buildings stand out as the most important elements in urban areas, and their morphology and spatial characteristics are distinctly captured in high-resolution satellite imagery. Although LiDAR point clouds enable more precise 3D reconstructions of urban structures, their high cost and lack of real-time availability impose significant constraints. In contrast, satellite images offer shorter revisit intervals and more cost-effective coverage. Deep learning methods introduce an innovative, automated pipeline for interpreting remote sensing data. Among these approaches, Mask R-CNN has proven effective at extracting building boundaries from high-resolution satellite imagery. Recent studies have emphasized producing smoother, more consistent boundary geometries and strengthening the robustness of network across diverse urban scenes. In this study, we used Mask R-CNN as the base method for feature extraction. We experimented with different datasets offering varied spatial characteristics, aiming to fine-tune the model for high-resolution satellite imagery in Taiwan and integrate a boundary-regularization module to produce cleaner, sharper building boundaries. Future work will focus on developing a fully automated framework tailored for Taiwan, aiming to achieve higher accuracy in building footprint extraction and Level-of-Detail (LoD)-2 building modeling. Keywords: Building boundary extraction, Deep learning, Mask R-CNN Share Link | Plain Format | Corresponding Author (Yun-Hao Lin)

30	Topic A: General Remote Sensing	ABS-138
Estimation of Terrestrial Albedo from Multiple Reflectance Observations in a Single Day for SGLI Albedo Products Susaki, J.(a*), Ishii, Y.(b), Kuriki, A.(c), Kimura, Y.(d) and Ono, A.(e) (a) Professor, Graduate School of Engineering, Kyoto University, Japan (b) Assistant Professor, Graduate School of Engineering, Kyoto University, Japan (c) Technical Staff, Graduate School of Engineering, Kyoto University, Japan (d) Student, Graduate School of Engineering, Kyoto University, Japan (e) Associate Professor, Faculty of Humanities, Tenri University, Japan Abstract This paper examines algorithms for estimating terrestrial albedo for the products of the Global Change Observation Mission - Climate (GCOM-C) / Second-generation Global Imager (SGLI), which was launched in December 2017 by the Japan Aerospace Exploration Agency (JAXA). We focused on an approach using parameters derived from a bidirectional reflectance distribution function (BRDF) model. In operational products of the Moderate Resolution Imaging Spectroradiometer (MODIS), kernel-driven BRDF model parameters are estimated from multiple sets of reflectance and they are applied to estimate the land surface albedo. However, one of the challenges in this algorithm is to require multiple reflectance over several or dozens of days in order to achieve robust parapmeter estimation. Instead, another approach has been proposed that estimates the land surface albedo from a single set of reflectance with multi-regression models. The multi-regression models are derived for an arbitrary geometry from datasets of simulated albedo and multi-angular reflectance. Our preliminary experiments using in situ multi-temporal data demonstrated that this algorithm requires information about the land cover of the pixel of interest, and the variance of its estimated albedo is sensitive to the observation geometry. Therefore, in this research, we propose an approach that estimates the land surface albedo from multiple reflectance observed within several hours in a single day. We conducted field measurement in barren land, Tottori Sand Dunes, Japan. We calculated the land surface albedos from those data sets, and compared the albedos with the actual albedos and those obtained using the traditional approach. We found that the error of the albedos estimated by the proposed approach is acceptable. In near future, we^ll examine the performance of the proposed method by using actual satellite images. Keywords: Terrestrial albedo, BRDF, In-situ measurement, narrowband-to-broadband (NTB) conversion Share Link | Plain Format | Corresponding Author (Junichi Susaki)

Page 1 (data 1 to 30 of 339) | Displayed ini 30 data/page
1 2 3 4 5 6 7 8 9 10 11 12 NEXT >>

Powered By Konfrenzi Ultimate 1.832M-Build8 © 2007-2025 All Rights Reserved