Geometric Attributes Research Articles

A lightweight grape detection model in natural environments based on an enhanced YOLOv8 framework.

Grapefruit and stem detection play a crucial role in automated grape harvesting. However, the dense arrangement of fruits in vineyards and the similarity in color between grape stems and branches pose challenges, often leading to missed or false detections in most existing models. Furthermore, these models' substantial parameters and computational demands result in slow detection speeds and difficulty deploying them on mobile devices. Therefore, we propose a lightweight TiGra-YOLOv8 model based on YOLOv8n. Initially, we integrated the Attentional Scale Fusion (ASF) module into the Neck, enhancing the network's ability to extract grape features in dense orchards. Subsequently, we employed Adaptive Training Sample Selection (ATSS) as the label-matching strategy to improve the quality of positive samples and address the challenge of detecting grape stems with similar colors. We then utilized the Weighted Interpolation of Sequential Evidence for Intersection over Union (Wise-IoU) loss function to overcome the limitations of CIoU, which does not consider the geometric attributes of targets, thereby enhancing detection efficiency. Finally, the model's size was reduced through channel pruning. The results indicate that the TiGra-YOLOv8 model's mAP(0.5) increased by 3.33% compared to YOLOv8n, with a 7.49% improvement in detection speed (FPS), a 52.19% reduction in parameter count, and a 51.72% decrease in computational demand, while also reducing the model size by 45.76%. The TiGra-YOLOv8 model not only improves the detection accuracy for dense and challenging targets but also reduces model parameters and speeds up detection, offering significant benefits for grape detection.

Adaptive Spline Surface Fitting With Arbitrary Topological Control Mesh.

Reconstructing a spline surface from a given arbitrary topological triangle mesh is a fundamental and challenging problem in computer-aided design and engineering. This article introduces a novel surface fitting method utilizing G-NURBS capable of handling control meshes with arbitrary topologies. This method employs adaptive control point adjustment, guided by the geometric attributes of the input model, ensuring precise representation of sharp features such as edges and corners. Two primary strategies are employed: A parameter correspondence approach designed for sharp features and a control mesh iterative refinement technique that incorporates geometrical feature information. The proposed method has been tested and evaluated on various CAD models to demonstrate its effectiveness. This method can achieve higher fitting accuracy while faithfully preserving the geometrical features with fewer control points.

Probing Disk Ice Content and Polycyclic Aromatic Hydrocarbon Emission through Multiband MagAO+Clio Images of HD 141569

We present resolved images of the inner disk component around HD 141569A using the Magellan adaptive optics system with the Clio2 1–5 μm camera, offering a glimpse of a complex system thought to be in a short evolutionary phase between protoplanetary and debris disk stages. We use a reference star along with the Karhunen–Loéve image projection (KLIP) algorithm for point-spread function subtraction to detect the disk inward to about 0.″24 (∼25 au assuming a distance of 111 pc) at high signal-to-noise ratios at L′ (3.8 μm), Ls (3.3 μm), and narrowband Ice (3.1 μm). We identify an arc or spiral arm structure at the southeast extremity, consistent with previous studies. We implement forward modeling with a simple disk model within the framework of a Markov Chain Monte Carlo sampler to better constrain the geometrical attributes and photometry using our KLIP-reduced disk images. We then leverage these modeling results to facilitate a comparison of the measured brightness in each passband to find a reduction in scattered light from the disk in the Ice filter, implying significant absorption due to water ice in the dust. Additionally, our best-fit disk models exhibit peak brightness in the southwestern, back-scattering region of the disk, which we suggest to be possible evidence of 3.3 μm polycyclic aromatic hydrocarbon emission. However, we point out the need for additional observations with bluer filters and more complex modeling to confirm these hypotheses.

Graph neural network-based identification of ditch matching patterns across multi-scale geospatial data

Ditches are vital to water system data. To ease the task of matching multi-scale ditch data and enhance accuracy (Acc), it is essential to discern ditch data matching patterns. Despite its importance, limited research has been conducted on ditch matching patterns, and the primary reasons for variations in multi-scale geographical entities’ matching patterns remain underexplored. Here, we introduce a supervised graph neural network (GNN) method, targeting the direct analysis of 1:10,000 ditch characteristics to identify the matching patterns between 1:10,000 and 1:25,000 datasets. The ditch network is depicted as a graph structure, with nodes representing ditch segments and edges indicating their connections. Subsequently, each ditch segment’s geometric, semantic, and topological attributes are computed as node attributes, and their matching patterns with the 1:25,000 dataset are labeled as node annotations. Ditch segment matching patterns are then categorized using supervised learning. Experiments using Overijssel, Netherlands’ ditch data reveal that this method achieves a 97.3% classification Acc, outperforming other GNN methods by 25.6–26.3% and traditional machine learning methods by 33%. These findings underscore the efficacy and superiority of the proposed supervised GNN approach in pinpointing ditch matching patterns.

Microstructure generation algorithm and micromechanics of curved fiber composites with random waviness

During multiscale modeling and analysis of composite materials, accurately capturing the distinctive microstructural features of the material is essential. This is achieved by generating a representative volume element (RVE). Fiber reinforcements in composites can exhibit various shapes, including straight, slightly wavy, or complex shapes, depending on the inherent properties or the manufacturing process. Fibers are characterized based on their tortuosity, spread, and orientation distribution. It is crucial to account for these geometrical attributes during the microstructure generation and subsequent micromechanics modelling as they significantly influence the effective properties of the composite. The present work describes a novel algorithm capable of modeling a broad range of 3D periodic microstructures with fibers having random waviness. The effects of various parameters in the algorithm are studied to understand their influence on fiber shapes. This is followed by finite element analysis of the generated RVEs using the embedded element technique to quantify the microstructure- mechanical properties relationships. Finally, guidelines on the usage of the proposed model to generate fibers with desired microstructure is elucidated. A comprehensive framework for modeling and analyzing microstructures with complex fiber geometry is outlined.

Prediction of Transonic Flow over Cascades via Graph Embedding Methods on Large-Scale Point Clouds

In this research, we introduce a deep-learning-based framework designed for the prediction of transonic flow through a linear cascade utilizing large-scale point-cloud data. In our experimental cases, the predictions demonstrate a nearly four-fold speed improvement compared to traditional CFD calculations while maintaining a commendable level of accuracy. Taking advantage of a multilayer graph structure, the framework can extract both global and local information from the cascade flow field simultaneously and present prediction over unstructured data. In line with the results obtained from the test datasets, we conducted an in-depth analysis of the geometric attributes of the cascades reconstructed using our framework, considering adjustments made to the geometric information of the point cloud. We fine-tuned the input using 1603 data points and quantified the contribution of each point. The outcomes reveal that variations in the suction side of the cascade have a significantly more substantial influence on the field results compared to the pressure side and explain the way graph neural networks work for cascade flow-field prediction, enhancing the comprehension of graph-based flow-field prediction among developers and proves the potential of graph neural networks in flow-field prediction on large-scale point clouds and design.

Autonomous Trajectory Planning for Spray Painting on Complex Surfaces Based on a Point Cloud Model.

Using teach pendants or offline programming methods can generate tool paths for robot manipulators to carry out production activities, such as spray painting on objects of different geometries. This task, in which the complexity of painting the surface is one of the main challenges, requires highly skilled operators. In addition, the time spent setting up a robot task can be justified for the mass production of the same workpiece. However, it is inconvenient for low-production and high-variation production lines. In order to overcome these challenges, this study presents an algorithm to autonomously generate robot trajectories for a spray-painting process applied to objects with complex surfaces based on input 3D point cloud data. A predefined spherical mesh wraps the object, organizing the geometrical attributes into a structured data set. Subsequently, the region of interest is extracted and isolated from the model, which serves as the basis for the automatic path-planning operation. A user-friendly graphical user interface (GUI) is developed to define input parameters, visualize the point cloud model and the generated trajectory, simulate paint quality using a color map, and ultimately generate the robot's code. A 3D sensor is used to localize the pose of the workpiece ahead of the robot and adjust the robot's trajectory. The efficacy of the proposed approach is validated first by using various workpieces within a simulated environment and second by employing a real robot to execute the motion task.

Experimental and numerical study on the Izbash equation coefficients in rough single fractures

The Izbash equation has been widely used in the subsurface applications. However, the Izbash equation is still empirical, and its coefficients (scaling factor λ and power exponent M) have not been systematically characterized and quantified. In this study, laboratory experiments and numerical simulations of fluid flow across a wide range of hydraulic gradients (J = 0–4) in horizontal rough fractures were conducted to comprehensively characterize and quantify the influence of fracture geometric attributes and fluid inertial effects on λ and M. The results showed that λ increased with fracture relative roughness (RSD). The fluid inertial effect (quantified by the non-Darcy effect factor E and Re) had a two-stage influence on λ. When the fluid flow was laminar, λ increased with E. However, when the fluid flow regime starts to transition from laminar flow to turbulent flow, λ decreased with increasing E. M is positively correlated with RSD and the fluid inertia effect E. We found that the transition of flow regime from laminar to turbulent flow depended on whether the recirculation zones are fully developed. The fully developed recirculation zones determine the distortions of the velocity field and flow field, which induced the turbulent flow. The quantitative models of λ and M were obtained based on numerical simulations, which quantified the coupling influence of the fracture geometric property and fluid inertial effect. The validity of quantitative models was verified by laboratory experiments. Our work provided a new understanding of the Izbash coefficients and laid a foundation for theoretical background exploration of the Izbash equation.

Curvature analysis and its correlation with faults and fractures in presalt carbonates, Santos Basin, Brazil

Presalt lacustrine carbonate reservoirs in the Santos Basin, Brazil, can have a high degree of fracturing. Properly characterizing such properties is important due to their influence on the porosity–permeability system. It may play a key role in the so-called excess permeability features and the flow behaviour, especially in the Aptian Alagoas Stage. Therefore, 3D characterization of the fracture system improves the reservoir model. For this purpose, correctly identifying the fault and fracture zones in seismic data is crucial. This task is generally performed manually and guided by amplitude interpretation or geometric seismic attributes. However, due to the seismic resolution of the pre-salt seismic datasets, many faults and fractured zones are not identified in the seismic amplitude domain or by standard geometric attributes such as coherence. Improving the resolution of seismic data and its geometric attributes can positively impact the delineation of fractured zones, structural lineaments, and their connectivity. Curvature (k) attributes have a high potential to reveal vertical to subvertical subseismic-scale structures, which is the main type of strain generated from the tectonic style in the Pre-salt rocks. Among the curvature attributes, the most-positive (k1) and the most-negative (k2) attributes are some of the most commonly used. The present study evaluates the use of seismic inverse Q-filter and three different dip estimation techniques based on Fast Fourier Transform (FFT), Principal Component Analysis (PCA), and Phase Gradient (PG). It also adds two postfiltering techniques (median and recursive Gaussian filters) to obtain a consistent most-positive curvature. We then analyse two analytical models that correlate the volumetric fracture intensity, measured on Borehole Image Logs (BHI), with the curvature. Then, we perform a comprehensive study to understand the role of seismic-scale strain (translated into curvature) and mechanical stratigraphy in the development of fractures in the presalt, where we address that in addition to seismic-scale deformation, the contents of silica, clay, and organic matter are also relevant variables. These results can help to guide future studies of fracture characterization to prospect naturally fractured reservoirs or to estimate fracture intensity for dual-porosity reservoir modelling on the Santos presalt carbonates.

Experimental and Statistical Study on the Formation Characteristics and Discrimination Criteria of River Blockages Caused by Landslides

The discrimination of river blockages is very important for the risk assessment of landslide disasters and secondary hazards. Experimental studies and statistical analyses were carried out to explore the formation process and discriminant criteria of river blockages caused by landslides. An adjustable slide chute was designed and built to conduct forty-five landslide experiments. According to the experimental results, river blockage was identified as having six types based on the differences between the water depth and the height of the landslide dam, and the degree of river blockage increased from 70% to 100% as the chute angle, particle size, and landslide volume increased. It is also found that landslide volume controls the landslide dam height and degree of blockage, and particle size and slide angle control both the landslide velocity as it enters the river and the cross-section shape of the landslide dam. To investigate more influence factors, a statistical investigation of 60 real landslide cases was carried out, and it revealed that some geometric attributes related to landslide volume have the highest correlation with river blockage, especially landslide thickness. Finally, an improved probability model was proposed to assess the possibility of complete blockage, and it has overall accuracies of 91.1% and 83.3% when applied to predict experimental landslide cases and real landslide cases, respectively.

Mechanical behavior investigation for quenching and partitioning steel dissimilar resistance spot welds

Resistance spot welding is still the most common joining process for autobody structure assembly. It has many advantages and is simple to use, but the main challenge is controlling the process parameters that affect mechanical properties and weld quality. Advanced high-strength steels were developed to meet the demands of automakers such as lighter and stronger autobody. Quenching and partitioning (Q&P 980) steel is a third-generation advanced high strength steel with a combination of high strength and high ductility. In this study, effect of resistance spot welding parameters on the mechanical performance of dissimilar combination of Q&P980 steel and SPFC780Y high strength steel was investigated. Welding current, welding time, electrode pressure and holding time were chosen as the most important process parameters influencing the weld quality. A comprehensive investigation was conducted on geometrical attributes such as nugget size, fusion zone width, fusion zone area, electrode indentation and heat affected zone area and their correlations with peak load (Pmax), failure energy, fracture energy, and elongation at peak load (Lmax). In contrast the failure modes and fracture behavior were studied using scanning electron microscopy. It was observed that using high levels of welding current and welding time while maintaining low levels of electrode pressure and holding time result in a maximum peak load, energy absorption and maximum (Lmax) with pullout failure mode (PF) and this was demonstrated for large nugget size to a critical limit. The methods and findings of this research can be used as a basis for further research and development in the field of dissimilar resistance spot welding.

Enhancing Energy Efficiency in Jet Ejectors: A Computational Fluid Dynamics Investigation

Abstract: This study investigates the analysis of the operating modes of two distinct jet ejector types, namely, Constant Rate of Momentum Change (CRMC), Constant Pressure Mixing (CPM). Computational Fluid Dynamics (CFD) was employed to thoroughly examine and evaluate their respective geometrical attributes. Through our comprehensive scrutiny, we ascertained the notably superior performance of CRMC ejectors. This superiority is primarily linked to the continuous presence of compression shock waves within the flow processes of CRMC, which consequently lead to comparable or even diminished critical condenser pressures. The core objective of this research is to advance the application of jet-ejector refrigeration technology, especially for the efficient harnessing of low-grade waste heat, thereby contributing to enhanced energy efficiency and sustainable cooling solutions. We concentrated on two operational modes, CRMC and CPM, and employed CFD to analyze the geometry of these ejectors. We found a particular configuration called RSO-2 that demonstrated exceptional performance after carrying out extensive experiments encompassing 53 different parameter combinations using the Taguchi Design of Experiments. With regard to the throat inlet, RSO-2 demonstrated an exceptional mass flow rate of 5.28E+05 J kg-1 and a Mach number of 1.5418 at the throat exit. Furthermore, a strong correlation was found between the diameter of the throat and the radius of the mixing chamber, as revealed by our thorough sensitivity analysis. This discovery offers important new information for the development of jet ejector designs.

CurvAGN: Curvature-based Adaptive Graph Neural Networks for Predicting Protein-Ligand Binding Affinity

Accurately predicting the binding affinity between proteins and ligands is crucial for drug discovery. Recent advances in graph neural networks (GNNs) have made significant progress in learning representations of protein-ligand complexes to estimate binding affinities. To improve the performance of GNNs, there frequently needs to look into protein-ligand complexes from geometric perspectives. While the “off-the-shelf” GNNs could incorporate some basic geometric structures of molecules, such as distances and angles, through modeling the complexes as homophilic graphs, these solutions seldom take into account the higher-level geometric attributes like curvatures and homology, and also heterophilic interactions.To address these limitations, we introduce the Curvature-based Adaptive Graph Neural Network (CurvAGN). This GNN comprises two components: a curvature block and an adaptive attention guided neural block (AGN). The curvature block encodes multiscale curvature informaton, then the AGN, based on an adaptive graph attention mechanism, incorporates geometry structure including angle, distance, and multiscale curvature, long-range molecular interactions, and heterophily of the graph into the protein-ligand complex representation. We demonstrate the superiority of our proposed model through experiments conducted on the PDBbind-V2016 core dataset.

Analysis of Pile Group under Vertical Loading in Clay and Sand using FEM

Abstract: Pile foundations are the preferred choice for supporting heavy structures when hard bedrock is situated at considerable depths beneath weak soil conditions. The ultimate load-carrying capacity of these piles hinges on several factors, including material composition, soil properties, geometric attributes, and the mechanism of load transfer (end bearing or friction). This research focuses on investigating the load-bearing capacity and settlement characteristics of end-bearing piles in both soft clay and medium sand. The study explores these parameters by varying pile diameter, length, and spacing within a pile group. Utilizing ANSYS for analysis, the study draws comparisons between clay and sand. The study's findings reveal that clay exhibits a higher load-carrying capacity compared to sand. Additionally, it highlights that sand experiences significantly greater settlement compared to clay when an optimal pile diameter of 0.4 meters and a length of 4 meters are employed.

Deep Learning Logging Sedimentary Microfacies via Improved U-Net

Well logging data contain abundant information on stratigraphic sedimentology. Artificial identification is usually strongly subjective and time-consuming. Pattern recognition algorithms like SVM may not adequately capture the depth-related variations in logging curve shape. This paper defines logging sedimentary microfacies as unidirectional 2D image segmentation and builds an improved U-net model to meet the requirements of logging sedimentary microfacies acquaintance. The proposed model contains three characteristics: (1) It removes pooling layers to avoid the loss of spatial features; (2) it utilizes multi-scale convolution blocks for mining multi-scale spatial features in logging data; (3) a one dimensional convolution layer is added to achieve deep single-direction segmentation. In this model, a 2D image composed of several standardized logging curves is used as the network’s input. In addition, we propose an effective data enhancement method and calculate the geometric feature attributes of well logging curves to reduce the complexity of the data characteristics. We tested the model on manually annotated validation datasets. Our method automatically measures fine sedimentary microfacies characteristics, improving the accuracy of sedimentary microfacies identification and achieving the desired result. Additionally, the model was tested on unlabeled actual logging data, which shows the generalizability of this deep learning method on different datasets.

Unsupervised machine learning and multi-seismic attributes for fault and fracture network interpretation in the Kerry Field, Taranaki Basin, New Zealand

Unsupervised machine learning using an unsupervised vector quantization neural network (UVQ-NN) integrated with meta-geometrical attributes as a novel computation process as opposed to traditional methodologies is currently used effectively in the 3D seismic structural interpretation for high-resolution detection of fault patterns, fracture network zones, and small-scale faults (SSFs). This technology has a crucial role in locating prospective well sites and building a 3D structural model while saving time and cost. The innovation of the current workflow involves combining geostatistical and structural filtering, optimal geometrical seismic attributes, UVQ-NN for automatic major faults, fracture network zones, and SSFs volumes extraction due to the unavailability of well logs and cores. To sharpen the fault edges and discontinuities, a steered volume was first extracted. Structural filters were then applied to the 3D volume, first with a dip-steered median filter (DSMF), followed by a dip-steered diffusion filter (DSDF), and finally, both DSMF and DSDF were combined to generate the fault enhancement filter (FEF). After that, optimal geometrical attributes were computed and extracted, such as similarity, FEF on similarity, maximum curvature, polar dip, fracture density, and thinned fault likelihood (TFL) attributes. Finally, selected attributes were inserted as the input layer to the UVQ-NN to generate segmentation and matching volumes. On the other hand, the TFL was used with the voxel connectivity filter (VCF) for 3D automatic fault patches extraction. The results from the UVQ-NN and VCF identified the locations, orientations, and extensions of the main faults, SSFs, and fracture networks. The implemented approach is innovative and can be employed in the future for the identification, extraction, and classification of geological faults and fracture networks in any region of the world.

An intrinsic anisotropic feature of DTI images derived by geometric properties on the Riemannian manifold

Diffusion tensor imaging (DTI) is a promising technique for characterizing organic microstructural information by mapping a spatial point as a 3 × 3 symmetric matrix (called diffusion tensor). Analyzing DTI images directly is challenging since each element of the DTI image is a 3-order matrix rather than a scalar. It is meaningful to extract the scalar-valued feature for DTI image processing and analysis. In this paper, we propose an intrinsic anisotropic feature of DTI images using global geometric attributes of diffusion tensors on the Riemannian manifold. To measure the intrinsic similarity between diffusion tensors as symmetric matrices, we convert them into symmetric positive definite (SPD) matrices and use the normalized geodesic distance in the tensor manifold to define the visible intrinsic feature. Moreover, we extend the geodesic distance to make sure our feature can measure the intrinsic properties of diffusion tensors, and the neighborhood attributes and even the global information of a DTI image are utilized to define the proposed feature. To measure the anisotropy of DTI images, we define the average diffusion sphere at each point to measure the isotropic attributes and use the difference between the diffusion tensor and its average diffusion tensor to describe the anisotropic properties. It is experimentally demonstrated that the proposed feature performs adequately in characterizing various attributes of DTI images, including but not limited anisotropy, intrinsic geometric property, and average diffusion degree, etc. Compared with the most popular features, our anisotropy feature can represent DTI image properties more robustly and effectively since the contrast of various details in the feature is improved. Moreover, the results of fiber tracking defined by the anisotropy feature indicate that our feature performs better on representing substantial anisotropic attributes of DTI images and is potentially used in DTI image processing and analysis.

Programmable Droplet Bouncing on Bionic Functional Surfaces for Continuous Electricity Generation

AbstractThe natural phenomenon of droplets bouncing on various surfaces holds remarkable potential for applications like water transportation, self‐cleaning, antifreezing, etc. However, achieving precisely controlled patterned droplet bouncing on functional surfaces with accurately controlled factors like bouncing velocity and trajectory in three dimensions remains a formidable challenge. In this context, a concept of bionic hydrophobic functional surfaces composed of mushroom‐like microstructures is introduced. These microstructures are crafted using the projection microstereolithography (PµSL) based 3D printing technique, subsequently coated with a hydrophobic spray. By finely adjusting the geometric attributes and inclination angles of these micromushrooms, the ability is gained to meticulously manipulate the bouncing velocity and trajectory of water droplets. The most optimal performance is demonstrated by a droplet exhibiting a maximal jumping distance and height respectively of 2.5 and 7.1 mm with 50° inclined micromushrooms. Notably, these specially designed micromushrooms orchestrate diverse behaviors in droplet bouncing, encompassing patterned bouncing, antigravity jumps, and directional water transportation. Additionally, the functional surface's adaptable self‐cleaning capability facilitates the harnessing of energy from rainfall on large surfaces, offering potential applications in realms, such as self‐cleaning mechanisms, droplet capture, water conveyance, and clean energy generation.

An observational study on pedestrian and bicyclist violations at railroad grade crossings: Exploring the impact of geometrical and operational attributes

Introduction: In order to enhance the existing body of research regarding the prevalence of violations committed by pedestrians and bicyclists, this observational study utilized personal-specific, train-specific, operational, and behavioral data extracted from recorded videos at seven railroad grade crossings (RRGCs). Method: A total of 7,332 pedestrians and 1,684 bicyclist crossing events were observed during 279 gate operation phases at seven study locations. First, two separate violation models (binary logit) for pedestrians and bicyclists were developed to determine the influence of personal-specific characteristics and other behavioral traits on probability of pedestrians’ and bicyclists’ violation. Later, the impact of various geometrical (railway track, distance between gates, distance between gate and nearest track, and road width) and operational (train vehicle unit) parameters on the factors that caused non-motorist violations was examined by developing distinctbinary logit models. Results: The probability of pedestrians’and bicyclists’ violationwas found to be significantly influenced by train characteristics (train speed and train occupancy time), gate operational attributes (warning time and gate blockage duration), and personal-specificcharacteristics (gender, age, crossing speed, and looking both side). Furthermore, geometrical attributes along with train vehicle unit were observed to have impact over various factors that significantly influenced the behavior of pedestrians and bicyclists to violate. Practical Applications: The findings of this research provide essential information that could be applied in the initiatives of user awareness and enforcement program, emphasis on situational awareness and design of operational and geometricalelementsto significantlylimitthe effects of risky walking andbicycling behavior at RRGCs.

Learning geometric consistency and discrepancy for category-level 6D object pose estimation from point clouds

Category-level 6D object pose estimation aims to predict the position and orientation of unseen object instances, which is a fundamental problem in robotic applications. Previous works mainly focused on exploiting visual cues from RGB images, while depth images received less attention. However, depth images contain rich geometric attributes about the object’s shape, which are crucial for inferring the object’s pose. This work achieves category-level 6D object pose estimation by performing sufficient geometric learning from depth images represented by point clouds. Specifically, we present a novel geometric consistency and geometric discrepancy learning framework called CD-Pose to resolve the intra-category variation, inter-category similarity, and objects with complex structures. Our network consists of a Pose-Consistent Module and a Pose-Discrepant Module. First, a simple MLP-based Pose-Consistent Module is utilized to extract geometrically consistent pose features of objects from the pre-computed object shape priors for each category. Then, the Pose-Discrepant Module, designed as a multi-scale region-guided transformer network, is dedicated to exploring each instance’s geometrically discrepant features. Next, the NOCS model of the object is reconstructed according to the integration of consistent and discrepant geometric representations. Finally, 6D object poses are obtained by solving the similarity transformation between the reconstruction and the observed point cloud. Experiments on the benchmark datasets show that our CD-Pose produces superior results to state-of-the-art competitors.