3D Object Detection with Attention:Shell-Based Modeling

LIDAR point cloud-based 3D object detection aims to sense the surrounding environment by anchoring objects with the Bounding Box(BBox).However,under the three-dimensional space of autonomous driving scenes,the previous object detection methods,due to the pre-processing of the original LIDAR point cloud into voxels or pillars,lose the coordinate information of the original point cloud,slow detection speed,and gain inaccurate bounding box positioning.To address the issues above,this study proposes a new two-stage network structure to extract point cloud features directly by PointNet++,which effectively preserves the original point cloud coordinate information.To improve the detection accuracy,a shell-based modeling method is proposed.It roughly determines which spherical shell the coordinates belong to.Then,the results are refined to ground truth,thereby narrowing the localization range and improving the detection accuracy.To improve the recall of 3D object detection with bounding boxes,this paper designs a self-attention module for 3D object detection with a skip connection structure.Some of these features are highlighted by weighting them on the feature dimensions.After training,it makes the feature weights that are favorable for object detection get larger.Thus,the extracted features are more adapted to the object detection task.Extensive comparison experiments and ablation experiments conducted on the KITTI dataset verify the effectiveness of our proposed method in improving recall and precision.

Automatic Generation Method of Geological Cross-Sections in Dredging Engineering Based on 3D Solid NURBS Models

An automatic generation method of geological cross-sections in dredging engineering based on 3D geological solid models is presented.The 3D geological models are built applying the non-uniform rational B-splines(NURBS) technique,and a 2D profile can be calculated and generated automatically through Boolean operation to meet the demands of dredging projects.Moreover,an automatic marking method for geological attributes is put forward based on database technology,and the geological attributes include the profile name,scale,horizontal and vertical relative coordinates,geological lithology,and 2D standard lithology legend.At the same time,the automatic marking method can also provide an interactive mode for geological engineers to edit and modify the profile in the modeling system.Practical engineering applications show that the automatic generation method is a simple,flexible,fast and precise visual graphics rendering process that can create 2D standard profiles automatically and efficiently.This method also provides a convenient support tool for geological engineering digital analysis.

Framework system and research flow of uncertainty in 3D geological structure models

Uncertainty in 3D geological structure models has become a bottleneck that restricts the development and application of 3D geological modeling.In order to solve this problem during periods of accuracy assessment,error detection and dynamic correction in 3D geological structure models,we have reviewed the current situation and development trends in 3D geological modeling.The main context of uncertainty in 3D geological structure models is discussed.Major research issues and a general framework system of uncertainty in 3D geological structure models are proposed.We have described in detail the integration of development practices of 3D geological modeling systems,as well as the implementation process for uncertainty evaluation in 3D geological structure models.This study has laid the basis to build theoretical and methodological systems for accuracy assessment and error correction in 3D geological models and can assist in improving 3D modeling techniques under complex geological conditions.

A Study of Prospective Barriers,Benefits and Measures for Building Information Modeling(BIM)Adoption in Nepalese AEC Industry

The construction industry needs modern construction methodology and technology to improve sustainability and production performance.Building Information Modelling(BIM)technology supports improving the quality of products by reducing design and construction defects,risks to the health and safety of workers,and reduce overall project cost and delivery time.The BIM has capabilities,but it is still undiscovered and unable to exploit the full scale of its benefits in the Architectural Engineering and Construction(AEC)industry.There is a trend to adopt the BIM level 1,which is limited to 2D and only in a few cases 3D models uses in the design and construction of residential and commercial buildings,particularly in Nepal.Hence,this paper focuses on providing insight into the BIM benefits and identifies the potential barriers while adopting BIM Level 3 in Nepal.This was accomplished by developing a 4DBIM model of a multi-story residential building in Nepal and conducting the industry survey via focus group with the AEC professionals based on the developed 4DBIM model.A comprehensive literature review was conducted and presented the findings of the BIM benefits and barriers while adopting BIM.The study found that commercial and governmental projects can immediately be adopted BIM technology.It is concluded that the unavailability of skilled BIM users and the lack of proper policies for BIM adoption are key barriers in Nepal.Hence,the new policy is required to achieve and exploit the full scale of the BIM benefits and improve the project delivery in terms of quality,cost and time including the health and safety of workers and the sustainability of the AEC industry.

3D打印模型在脊柱外科住院医师规范化培训中的创新应用

目的探讨3 D打印模型在脊柱外科住院医师规范化培训(简称住培)中的应用效果。方法选取2023年1月—2023年12月在广东省人民医院脊柱外科轮转的46名二年级外科学专业住院医师作为研究对象,通过抽签方式随机分为观察组和对照组,每组各23名。观察组采用3 D打印模型辅助教学,对照组则采用传统教学。经过1个月的教学培训,通过出科考试、综合考核、教学满意度及自我鉴定问卷调查等方式多维度评估教学效果。采用独立样本t检验进行数据分析。结果观察组的出科考试成绩、综合考核评分及教学满意度均显著优于对照组(均P<0.0001)。此外,观察组自我鉴定部分的学习兴趣、自学能力、团队合作能力及综合分析能力也均显著高于对照组(P<0.001)。结论3 D打印模型辅助教学模式能有效提升脊柱外科住培的教学质量,并显著提高住院医师的教学满意度、综合能力及自我评价,为脊柱外科住培教学提供了新的有效路径。

3DV quality model based depth maps for view synthesis in FTV system

Depth maps are used for synthesis virtual view in free-viewpoint television （FTV） systems. When depth maps are derived using existing depth estimation methods, the depth distortions will cause undesirable artifacts in the synthesized views. To solve this problem, a 3D video quality model base depth maps （D-3DV） for virtual view synthesis and depth map coding in the FTV applications is proposed. First, the relationships between distortions in coded depth map and rendered view are derived. Then, a precisely 3DV quality model based depth characteristics is develop for the synthesized virtual views. Finally, based on D-3DV model, a multilateral filtering is applied as a pre-processed filter to reduce rendering artifacts. The experimental results evaluated by objective and subjective methods indicate that the proposed D-3DV model can reduce bit-rate of depth coding and achieve better rendering quality.

Research and development of fringe projection-based methods in 3D shape reconstruction

This paper discusses current research and development of fringe projection-based techniques. A system based on Fourier transform profilometry (FTP) is proposed for three-dimensional (3D) shape recovery. The system improves the method of phase unwrapping to gain accurate 3D shapes of objects. The method uses a region-growing algorithm for the path prediction guided by the quality map to increase the recovering accuracy and provides a fast and simple tool for 3D shape recovery. The shape measurement and data recovery are integrated to offer a new method of 3D modelling. Examples are presented to verify the feasibility of the proposed method.

Contour Detection-Based Realistic Finite-Difference-Time-Domain Models for Microwave Breast Cancer Detection

In this paper, a collection of three-dimensional(3D)numerical breast models are developed based on clinical magnetic resonance images(MRIs). A hybrid contour detection method is used to create the contour, and the internal space is filled with different breast tissues, with each corresponding to a specified interval of MRI pixel intensity. The developed models anatomically describe the complex tissue structure and dielectric properties in breasts. Besides, they are compatible with finite-difference-time-domain(FDTD)grid cells. Convolutional perfect matched layer(CPML)is applied in conjunction with FDTD to simulate the open boundary outside the model. In the test phase, microwave breast cancer detection simulations are performed in four models with varying radiographic densities. Then, confocal algorithm is utilized to reconstruct the tumor images. Imaging results show that the tumor voxels can be recognized in every case, with 2 mm location error in two low density cases and 7 mm─8 mm location errors in two high density cases, demonstrating that the MRI-derived models can characterize the individual difference between patients' breasts.

Space-Temporal Variation in Underground Water Some Quality Parameters in Klodzko Water Intake Area Using Statistical and Geostatistical Methods （SW Part of Poland）

This paper presents selected results of research connected with the development of a （3D） geostatistical hydrogeochemical model of the Klodzko Drainage Basin, dedicated to the spatial and time variation in the selected quality parameters of underground water in the Klodzko water intake area （SW part of Poland） [1-6]. The research covers the period 1977-2012. Spatial analyses of the variation in different quality parameters, between others, Fe [gFe/m3], Mn [gMn/m3], ammonium ion [gNH4＋/m3] contents and oxidation capacity [gO2/m3], were carried out on the basis of the chemical determinations of the quality parameters of underground water samples taken from the wells in the water intake area [2-4]. Spatial and time variation in the parameters was analyzed on the basis of archival data （period 1977-1999） for 22 （pump and siphon） wells, later data obtained （November 2011） from tests of water taken from 14 existing wells and the latest data （January 2012） acquired from 3 new piezometers, which were made in other locations in the relevant area. Thematic databases, containing original data on coordinates X, Y （latitude, longitude） and Z （terrain elevation and time-years） and on regionalized variables, i.e. the underground water quality parameters in the Klodzko water intake area determined for different analytical configurations （22 wells, 14 wells, 14 wells ＋ 3 piezometers）, were created [2]. Both archival data （acquired in the years 1977-1999） and the latest data （collected in 2011-2012） were analyzed. These data were subjected to spatial analyses [2-6] using statistical and geostatistical methods [7-12]. The evaluation of basic statistics of the investigated quality parameters, including their histograms of distributions, scatter diagrams between these parameters and also correlation coefficients r, were presented in this article. The directional semivariogram function and the ordinary （block） kriging procedure were used to build the 3D geostatistical model. The geostatistical parameters of the theoretical models of directional semivariograms of the studied water quality parameters, calculated along the time interval and the well depth （taking into account the terrain elevation）, were used in the ordinary （block） kriging estimation. The obtained results of estimation, allowed to determine the levels of increased values Z＊ of studied underground water quality parameters [2, 4-6]. Generally, the behaviour of the underground water quality parameters has been found to vary in space and time. Thanks to the spatial analyses of the variation in the quality parameters in the Klodzko underground water intake area some regularities （trends） in the variation in water quality have been identified.

The Influence and Implications of Climate Change on Water Quality in a Large Water Reservoir in the Southwest, USA

Maintaining water quality in large reservoirs is crucial to ensure continued delivery of high-quality water to consumers for municipal and agricultural needs. Lake Mead, a large reservoir in the desert southwest, USA, is projected to be affected by both loss of volume and rising air temperatures through the end of the 21st century. In this study, reductions in lake volume, coupled with downscaled climate projections for rising air temperatures through the end of the 21st century, are incorporated into the 3D hydrodynamic and water quality model for Lake Mead. If current management practices continue in the future, simulations indicate water temperatures will increase in all scenarios and could increase by as much 2℃under the most pessimistic scenarios, but nutrient loads would not increase to concerning levels. Releases from the dam to downstream users are projected to be much warmer, and warmer water temperatures and significant dissolved oxygen in the water column are expected to cause challenges for ecosystem and recreation in the future. Surprisingly, during the Winter and Autumn, retention of heat in Lake Mead is more pronounced at higher surface elevations than the lower elevations as expected. The effects of these projections on the lake water quality and consequently, lake management decisions, are discussed.

3-D Geological Modeling-Concept,Methods and Key Techniques

3-D geological modeling plays an increasingly important role in Petroleum Geology, Mining Geology and Engineering Geology. The complexity of geological conditions requires different modeling methods in different situations. This paper summarizes the general concept of geological modeling; compares the characteristics of borehole-based modeling, cross-section based modeling and multi- source interactive modeling; analyses key techniques in 3-D geological modeling; and highlights the main difficulties and directions of future studies.

On reservoir boundary detection with marine CSEM

After decades of research and development, Marine Controlled Source ElectroMagnetics （MCSEM） has come into the application phase for oil and gas exploration. However, presently 3D modeling of MCSEM is far from practical with simple models and much computing time. Based on a forward modeling study of 3D frequency-domain MCSEM over a complicated target body and its sensitivity analysis, we derive a method which can delineate the edges of the resistive reservoir. We use the second derivative of the magnitude versus offset （MVO） curve to define the resistive reservoir edges. For the air-wave-dominated far field zone, we suggest using the less affected apparent resistivities in order to improve the adaptability of the method.

利用DEM在AutoCAD Civil 3D下自动提取断面数据

Civil 3D能够根据所建立的三维模型自动生成各种施工图、所需文档及设计信息。为此,在AutoCAD Civil 3D下利用DEM,实现了纵横断面数据的自动提取。实践证明,该方法适用于通行困难且线路较长的工程测量项目中,而且能大大减少外业工作的时间。

3D Bounding Box Proposal for on-Street Parking Space Status Sensing in Real World Conditions

Vision-based technologies have been extensively applied for on-street parking space sensing,aiming at providing timely and accurate information for drivers and improving daily travel convenience.However,it faces great challenges as a partial visualization regularly occurs owing to occlusion from static or dynamic objects or a limited perspective of camera.This paper presents an imagery-based framework to infer parking space status by generating 3D bounding box of the vehicle.A specially designed convolutional neural network based on ResNet and feature pyramid network is proposed to overcome challenges from partial visualization and occlusion.It predicts 3D box candidates on multi-scale feature maps with five different 3D anchors,which generated by clustering diverse scales of ground truth box according to different vehicle templates in the source data set.Subsequently,vehicle distribution map is constructed jointly from the coordinates of vehicle box and artificially segmented parking spaces,where the normative degree of parked vehicle is calculated by computing the intersection over union between vehicle’s box and parking space edge.In space status inference,to further eliminate mutual vehicle interference,three adjacent spaces are combined into one unit and then a multinomial logistic regression model is trained to refine the status of the unit.Experiments on KITTI benchmark and Shanghai road show that the proposed method outperforms most monocular approaches in 3D box regression and achieves satisfactory accuracy in space status inference.

Automatic Generation of 3D Building Models with Multiple Roofs

Based on building footprints (building polygons) on digital maps, we are proposing the GIS and CG integrated system that automatically generates 3D building models with multiple roofs. Most building polygons’ edges meet at right angles (orthogonal polygon). The integrated system partitions orthogonal building polygons into a set of rectangles and places rectangular roofs and box-shaped building bodies on these rectangles. In order to partition an orthogonal polygon, we proposed a useful polygon expression in deciding from which vertex a dividing line is drawn. In this paper, we propose a new scheme for partitioning building polygons and show the process of creating 3D roof models.

Exploring the efficiency of image metric for assessing the visual quality of 3D mesh model

Recent developments in 3D graphics technology have led to extensive processes on 3D meshes(e.g.,compression,simplification,transmission and watermarking),these processes unavoidably cause the visual perceptual degradation of the 3D objects.The existing mesh visual quality evaluation metrics either require topology constrain or fail to reflect the perceived visual quality.Meanwhile,for the 3D objects that are observed on 2D screens by the users,it is reasonable to apply image metric to assess the distortion caused by mesh simplification.We attempt to explore the efficiency of image metric for assessing the visual fidelity of the simplified 3D model in this paper.For this purpose,several latest and most effective image metrics,2D snapshots,number and pooling algorithms are involved in our study,and finally tested on the IEETA simplification database.The statistical data allow the researcher to select the optimal parameter for this image-based mesh visual quality assessment and provide a new perspective for the design and performance assessment of mesh simplification algorithms.

Three-Dimensional Model Retrieval Using Dynamic Multi-Descriptor Fusion

In this paper, we propose a dynamic multi-descriptor fusion （DMDF） approach to improving the retrieval accuracy of 3-dimensional （3D） model retrieval systems. First, an independent retrieval list is generated by using each individual descriptor. Second, we propose an automatic relevant/irrelevant models selection （ARMS） approach to selecting the relevant and irrelevant 3D models automatically without any user interaction. A weighted distance, in which the weight associated with each individual descriptor is learnt by using the selected relevant and irrelevant models, is used to measure the similarity between two 3D models. Furthermore, a descriptor-dependent adaptive query point movement （AQPM） approach is employed to update every feature vector. This set of new feature vectors is used to index 3D models in the next search process. Four 3D model databases are used to compare the retrieval accuracy of our proposed DMDF approach with several descriptors as well as some well-known information fusion methods. Experimental results have shown that our proposed DMDF approach provides a promising retrieval result and always yields the best retrieval accuracy.

SIRIUS3D叠前深度偏移软件在BZ25—1构造上的应用

介绍了SIRIUS3D叠前深度偏移软件的技术特点及工作流程,分析了该软件的优点和不足.将该软件首次应用于BZ25-1构造,通过精细的速度反演,建立了较为准确的三维速度模型,最终叠前深度偏移结果较为理想,尤其是中浅层复杂断层的成像质量得到明显提高,为BZ25-1构造评价提供了重要依据.

2D/3D Isometric Transformation Using Spring-Mass System

This paper presents a general method for 2D/3D transformation, which can be efficiently used in three dimensional computer aided garment design. The method utilizes a uniform triangular spring_mass based deformable model. 2D to 3D transformation and 3D to 2D transformation both can be implemented on the same model. A general and efficient collision detection method is also briefly discussed in this paper.

Updating digital elevation models via change detection and fusion of human and remote sensor data in urban environments

OpenStreetMap(OSM)currently represents the most popular project of Volunteered Geographic Information(VGI):geodata are collected by common people and made available for public use.Airborne Laser Scanning(ALS)enables the acquisition of high-resolution digital elevation models that are used for many applications.This study combines the advantages of both ALS and OSM,offering a promising new approach that enhances data quality and allows change detection:the mainly up-to-date 2D data of OSM can be combined with the high-resolution–but rarely updated–elevation information provided by ALS.This case study investigates building objects of OSM and ALS data of the city of Bregenz,Austria.Data quality of OSM is discerned by the comparison of building footprints using different true positive definitions(e.g.overlapping area).High quality of OSM data is revealed,yet also limitations of each method with respect to heterogeneous regions and building outlines are identified.For the first time,an up-to-date Digital Surface Model(DSM)combining 2D OSM and ALS data is achieved.A multitude of applications such as flood simulations and solar potential assessments can directly benefit from this data combination,since their value and reliability strongly depend on an up-to-date DSM.