As a convenience to ASPRS members, in-press peer reviewed articles approved for publication in forthcoming issues of PE&RS are listed below.
Manuscript number: 23-00034R2
IMU and Bluetooth Data Fusion to Achieve Submeter Position Accuracy in Indoor Positioning
Ugur Acar
Indoor navigation applications have become widespread in recent years with the ability of mobile phones which determine the position. Due to the inefficiency of global positioning system (GPS) indoors, other positioning methods have been developed based on local networks using technologies such as Bluetooth, wireless networks, ultra-wideband signals, ultrasonic signals, and radio frequency identification modules. Various technologies yield high or medium accuracy. Combining data from multiple sources via fusion enhances location precision. In this study, indoor positions were estimated using trilateration with Bluetooth devices, and the accuracy was improved by applying filters to the data from inertial measurement unit (IMU) sensors on the phone. As a result of combining Bluetooth and IMU data with data fusion, submeter accuracy was achieved. The results obtained were tested at Yildiz Technical University-Istanbul Turkey. It was determined that 92% of the data was obtained with submeter accuracy.
Manuscript number: 23-00036R2
Rice Identification Under Complex Surface Conditions with CNN and Integrated Remote Sensing Spectral-Temporal-Spatial Features
Tianjiao Liu, Jiankui Chen, Li Zhang, and Dong Li
Accurate and effective rice identification has great significance for the sustainable de-velopment of agricultural management and food security. This paper proposes an accu-rate rice identification method that can solve the confused problem between fragmented rice fields and the surroundings in complex surface areas. The spectral, temporal, and spatial features extracted from the created Sentinel-2 time series were integrated and collaboratively displayed in the form of visual images, and a convolutional neural net-work model embedded with integrated information was established to further mine the key information that distinguishes rice from other types. The results showed that the overall accuracy, precision, recall, and F1-score of the proposed method for rice identi-fication reached 99.4%, 99.5%, 99.5%, and 99.5%, respectively, achieving a better per-formance than the support vector machine classifier. Therefore, the proposed method can effectively reduce the confusion between rice and other types and accurately ex-tract rice distribution information under complex surface conditions.
Manuscript number: 23-00031R2
Combination of Terrestrial Laser Scanning and Unmanned Aerial Vehicle Photogrammetry for Heritage Building Information Modeling: A Case Study of Tarsus St. Paul Church
Şafak Fidan, Ali Ulvi, Abdurahman Yasin Yiğit, Seda Nur Gamze Hamal, and Murat Yakar
Cultural heritage building information modeling (HBIM) is an emerging process allowing us to reconstruct built heritage virtually. The data of a digitally documented cultural heritage building offers significant advantages as it is accessible and modifiable by all professionals involved in the same or different projects. The most important factor affecting the accuracy and precision of the HBIM model is the ability to collect complete and accurate information about the physical structure. Combining terrestrial laser scanning (TLS) and unmanned aerial vehicle (UAV) photogrammetry point clouds is one of the most efficient ways to capture accurate digital data on the building. This study provides the foundation for creating an HBIM model for cultural heritage the coupling of spatial data with TLS and UAV. This paper aims to generate synergy between TLS and UAV point cloud data and ensure that the spatial database contains sufficient data to model historical objects with HBIM tendencies.
Manuscript number: 23-00017R2
Quantifying Impacts of Regional Multiple Factors on Spatiotemporal the Mechanisms for Spatio-temporal changes of Net Primary Vegetation Productivity and Net Ecosystem Productivity: An Example in the Jianghuai River Basin, China
Huimin Chen, Benlin Wang, Liangfeng Zheng, and AmirReza Shahtahmassebi
Despite much valuable research on the mechanisms for spatio-temporal changesof net primary vegetation productivity (NPP) and net ecosystem productivity (NEP), there is a paucity of information on assessing impacts of regional multiple factors on spatiotemporal researchs of NPP and NEP in the complex environment. This study attempts to bridge this information gap using the Jianghuai Basin in China as a case study. Using a field campaign, remotely sensed imagery, socioeconomic data, and meteorological parameters, we developed a framework based on the Carnegie–Ames–Stanford Approach (CASA) model, correlation technique, trend analysis, and landscape metrics to measure spatiotemporal changes in NPP and NEP from 2001 to 2018. The derived changes were then linked to regional multiple factors including climate, landscape factors, human activity, and land use change. The results of the research can provide a scientific basis for vegetation evaluation, ecosystem assessment, and other aspects of the region.
Manuscript number: 23-00023R2
Terrain Complexity and Maximal Poisson-Disk Sampling-Based Digital Elevation Model Simplification
Jingxian Dong, Fan Ming, Twaha Kabika, Jiayao Jiang, Siyuan Zhang, Aliaksandr Chervan, Zhukouskaya Natallia, and Wenguang Hou
With the rapid development of lidar, the accuracy and density of the Digital Elevation Model (DEM) point clouds have been continuously improved. However, in some applications, dense point cloud has no practical meaning. How to effectively sample from the dense points and maximize the preservation of terrain features is extremely important. This paper will propose a DEM sampling algorithm that utilizes terrain complexity and maximal Poisson-disk sampling to extract key feature points for adaptive DEM sampling. The algorithm estimates terrain complexity based on local terrain variation and prioritizes points with high complexity for sampling. The sampling radius is inversely proportional to terrain complexity, while ensuring that points within the radius of accepted samples are not considered new samples. This way makes more points of concern in the rugged regions. The results show that the proposed algorithm has higher global accuracy than the classic six sampling methods.
Manuscript number: 23-00033R2
I²-FaçadeNet: An Illumination-invariant Façade Recognition Network Leveraging Sparsely Gated Mixture of Multi-color Space Experts for Aerial Oblique Imagery
Shengzhi Huang, Han Hu, and Qing Zhu
Façade image recognition under complex illumination conditions is crucial for various applications, including urban three-dimensional modeling and building identification. Existing methods relying solely on Red-Green-Blue (RGB)RGB images are prone to texture ambiguity in complex illumination environments. Furthermore, façades display varying orientations and camera viewing angles, resulting in performance issues within the RGB color space. In this study, we introduce an illumination-invariant façade recognition network (I²-FaçadeNet) that leverages sparsely gated multi-color space experts for enhanced façade image recognition in challenging illumination environments. First, RGB façade images are converted into multi-color spaces to eliminate the ambiguous texture in complex illumination. Second, we train expert networks using separate channels of multi-color spaces. Finally, a sparsely gated mechanism is introduced to manage the expert networks, enabling dynamic activation of expert networks and the merging of results. Experimental evaluations leveraging both the International Society for Photogrammetry and Remote Sensing benchmark data sets and the Shenzhen data sets reveal that our proposed I²-FaçadeNet surpasses various depths of ResNet in façade recognition under complex illumination conditions. Specifically, the classification accuracy for poorly illuminated façades in Zurich improves by nearly 8%, while the accuracy for over-illuminated areas in Shenzhen increases by approximately 3%. Moreover, ablation studies conducted on façade images with complex illumination indicate that compared to traditional RGB-based ResNet, the proposed network achieves an accuracy improvement of 3% to 4% up to 100% for overexposed images and an accuracy improvement of 3% to 10% for underexposed images.
Manuscript number: 23-00043R2
Development of Soil-Suppressed Impervious Surface Area Index for Automatic Urban Mapping
Akib Javed, Zhenfeng Shao, Iffat Ara, Muhammad Nasar Ahmad, Enamul Huq, Nayyer Saleem, and Fazlul Karim
Expanding urban impervious surface area (ISA) mapping is crucial to sustainable development, urban planning, and environmental studies. Multispectral ISA mapping is challenging because of the mixed-pixel problems with bare soil. This study presents a novel approach using spectral and temporal information to develop a Soil-Suppressed Impervious Surface Area Index (SISAI) using the Landsat Operational Land Imager (OLI) data set, which reduces the soil but enhances the ISA signature. This study mapped the top 12 populated megacities using SISAI and achieved an overall accuracy of 0.87 with an F1-score of 0.85. It also achieved a higher Spatial Dissimilarity Index between the ISA and bare soil. However, it is limited by bare gray soil and shadows of clouds and hills. SISAI encourages urban dynamics and inter-urban comparison studies owing to its automatic and unsupervised methodology.
Manuscript number: 23-00055R2
Dual-Branch Networks Based on Contrastive Learning for Long-Tailed Remote Sensing
Lei Zhang, Lijia Peng, Pengfei Xia, Chuyuan Wei, Chengwei Yang, and Yanyan Zhang
Deep learning has been widely used in remote sensing image classification and achieves many excellent results. These methods are all based on relatively balanced data sets. However, in real-world scenarios, many data sets belong to the long-tailed distribution, resulting in poor performance. In view of the good performance of contrastive learning in long-tailed image classification, a new dual-branch fusion learning classification model is proposed to fuse the discriminative features of remote sensing images with spatial data, making full use of valuable image representation information in imbalance data. This paper also presents a hybrid loss, which solves the problem of poor discrimination of extracted features caused by large intra-class variation and inter-class ambiguity. Extended experiments on three long-tailed remote sensing image classification data sets demonstrate the advantages of the proposed dual-branch model based on contrastive learning in long-tailed image classification.
Manuscript number: 23-00056R2
Comparison of 3D Point Cloud Completion Networks for High Altitude Lidar Scans of Buildings
Marek Kulawiak
High altitude lidar scans allow for rapid acquisition of big spatial data representing entire city blocks. Unfortunately, the raw point clouds acquired by this method are largely incomplete due to object occlusions and restrictions in scanning angles and sensor resolution, which can negatively affect the obtained results. In recent years, many new solutions for 3D point cloud completion have been created and tested on various objects; however, the application of these methods to high-altitude lidar point clouds of buildings has not been properly investigated yet. In the above context, this paper presents the results of applying several state-of-the-art point cloud completion networks to various building exteriors acquired by simulated airborne laser scanning. Moreover, the output point clouds generated from partial data are compared with complete ground-truth point clouds. The performed tests show that the SeedFormer network trained on the ShapeNet-55 data set provides promising shape completion results.
Manuscript number: 23-00052R2
Crop Monitoring System Using MODIS Time-Series Data for Within-Season Prediction of Yield and Production of US Corn and Soybeans
Toshihiro Sakamoto
In terms of contribution to global food security, this study aimed to build a crop monitoring system for within-season yield prediction of US corn and soybeans by using the Moderate Resolution Imaging Spectroradiometer (time-series data, which consists of three essential core algorithms (crop phenology detection, early crop classification, and crop yield prediction methods)). Within-season predictions for 2018–2022 were then made to evaluate the performance of the proposed system by comparing it with the United States Department of Agriculture’s (USDA’s) monthly forecasts and the fixed statistical data. The absolute percentage errors of the proposed system for predicting national-level yield and production were less than 5% for all simulation years as of day of year (DOY) 279. The prediction accuracy as of DOY 247 and DOY 279 were comparable to the USDA’s forecasts. The proposed system would enable us to make a comprehensive understanding about overview of US corn and soybean crop condition by visualizing detail spatial pattern of good- or poor harvest regions on a within-season basis.