3D OBJECT FEATURE EXTRACTION BASED ON SHAPE SIMILARITY

2003 ◽  
Vol 12 (01) ◽  
pp. 37-56 ◽  
Author(s):  
JEONG-JUN SONG ◽  
FOROUZAN GOLSHANI
Keyword(s):  
Feature Extraction ◽  
Hybrid Method ◽  
Visual Information ◽  
3D Model ◽  
Fundamental Problem ◽  
Spatial Arrangement ◽  
3D Models ◽  
Shape Similarity ◽  
Object Surface ◽  
3D Object

We introduce two complementary feature extraction methods for shape similarity based retrieval of 3D object models. The proposed methods lead us to achieve effectiveness and robustness in searching similar 3D models, and eventually support two essential query modes, namely, query by 3D model and query by 2D image. Our feature extraction scheme is inspired by the observation of human behavior in recognizing 3D objects. The process of extracting spatial arrangement from a 3D object can be considered as using human tactile sensation without visual information. On the other hand, the process of extracting 2D features from multiple views can be considered as examining an object by moving the viewing points (or camera positions). We propose a hybrid method of 3D model identification by object-centered feature extraction, which utilizes the Extended Gaussian Image (EGI) surface normal distribution and distance distributions between object surface points and origin. Another technique need in parallel is a hybrid method using view-centered features, which adopts simple geometric attributes such as circularity, rectangularity and eccentricity. To generate a signature for view-centered features, we have measured distances of a feature between different views and constructing histogram of the distance. We also address the fundamental problem of obtaining sample points on an object surface, which is important to extract reliable features from the object model.

Consistent Mesh Segmentation Based on Shape Diameter Function and EM

2014 ◽  
Vol 1049-1050 ◽  
pp. 1417-1420
Author(s):  
Hui Jia ◽  
Guo Hua Geng ◽  
Jian Gang Zhang
Keyword(s):  
Computer Graphics ◽  
3D Model ◽  
Shape Function ◽  
3D Models ◽  
Gaussian Mixture ◽  
Mesh Segmentation ◽  
Shape Similarity ◽  
Research Focus ◽  
Model Segmentation ◽  
New Research

3D model segmentation is a new research focus in the field of computer graphics. The segmentation algorithm of this paper is consistent segmentation which is about a group of 3D model with shape similarity. A volume-based shape-function called the shape diameter function (SDF) is used to on behalf of the characteristics of the model. Gaussian mixture model (GMM) is fitting k Gaussians to the SDF values, and EM algorithm is used to segment 3D models consistently. The experimental results show that this algorithm can effectively segment the 3D models consistently.

scholarly journals Voxel-Based 3D Object Reconstruction from Single 2D Image Using Variational Autoencoders

Mathematics ◽  
2021 ◽  
Vol 9 (18) ◽  
pp. 2288
Author(s):  
Rohan Tahir ◽  
Allah Bux Sargano ◽  
Zulfiqar Habib
Keyword(s):  
3D Reconstruction ◽  
3D Model ◽  
3D Models ◽  
Object Reconstruction ◽  
Data Set ◽  
3D Object ◽  
Practical Applications ◽  
3D Object Reconstruction ◽  
First Time ◽  
Memory Efficient

In recent years, learning-based approaches for 3D reconstruction have gained much popularity due to their encouraging results. However, unlike 2D images, 3D cannot be represented in its canonical form to make it computationally lean and memory-efficient. Moreover, the generation of a 3D model directly from a single 2D image is even more challenging due to the limited details available from the image for 3D reconstruction. Existing learning-based techniques still lack the desired resolution, efficiency, and smoothness of the 3D models required for many practical applications. In this paper, we propose voxel-based 3D object reconstruction (V3DOR) from a single 2D image for better accuracy, one using autoencoders (AE) and another using variational autoencoders (VAE). The encoder part of both models is used to learn suitable compressed latent representation from a single 2D image, and a decoder generates a corresponding 3D model. Our contribution is twofold. First, to the best of the authors’ knowledge, it is the first time that variational autoencoders (VAE) have been employed for the 3D reconstruction problem. Second, the proposed models extract a discriminative set of features and generate a smoother and high-resolution 3D model. To evaluate the efficacy of the proposed method, experiments have been conducted on a benchmark ShapeNet data set. The results confirm that the proposed method outperforms state-of-the-art methods.

3D Model Registration Based on Feature Extraction

2011 ◽  
Vol 299-300 ◽  
pp. 1091-1094 ◽  
Author(s):  
Jiang Zhu ◽  
Yuichi Takekuma ◽  
Tomohisa Tanaka ◽  
Yoshio Saito
Keyword(s):  
Feature Extraction ◽  
Point Cloud ◽  
3D Model ◽  
Point Cloud Data ◽  
Coordinate Systems ◽  
High Precision Measurement ◽  
Registration Method ◽  
Cloud Data ◽  
Cad System ◽  
Complicated Model

Currently, design and processing of complicated model are enabled by the progress of the CAD/CAM system. In shape measurement, high precision measurement is performed using CMM. In order to evaluate the machined part, the designed model made by CAD system the point cloud data provided by the measurement system are analyzed and compared. Usually, the designed CAD model and measured point cloud data are made in the different coordinate systems, it is necessary to register those models in the same coordinate system for evaluation. In this research, a 3D model registration method based on feature extraction and iterative closest point (ICP) algorithm is proposed. It could efficiently and accurately register two models in different coordinate systems, and effectively avoid the problem of localized solution.

DVFENet: Dual-branch Voxel Feature Extraction Network for 3D Object Detection

2021 ◽  
Author(s):  
Yunqian He ◽  
Guihua Xia ◽  
Yongkang Luo ◽  
Li Su ◽  
Zhi Zhang ◽  
...  
Keyword(s):  
Feature Extraction ◽  
Object Detection ◽  
3D Object ◽  
3D Object Detection

scholarly journals UNDERWATER PHOTOGRAMMETRY DIGITAL SURFACE MODEL (DSM) OF THE SUBMERGED SITE OF THE ANCIENT LIGHTHOUSE NEAR QAITBAY FORT IN ALEXANDRIA, EGYPT

2019 ◽  
Vol XLII-2/W10 ◽  
pp. 1-8 ◽  
Author(s):  
M. Abdelaziz ◽  
M. Elsayed
Keyword(s):  
3D Model ◽  
Low Cost ◽  
Data Gathering ◽  
Archaeological Site ◽  
3D Models ◽  
Surface Model ◽  
Digital Surface Model ◽  
Architectural Elements ◽  
3D Photogrammetry ◽  
First Time

<p><strong>Abstract.</strong> Underwater photogrammetry in archaeology in Egypt is a completely new experience applied for the first time on the submerged archaeological site of the lighthouse of Alexandria situated on the eastern extremity of the ancient island of Pharos at the foot of Qaitbay Fort at a depth of 2 to 9 metres. In 2009/2010, the CEAlex launched a 3D photogrammetry data-gathering programme for the virtual reassembly of broken artefacts. In 2013 and the beginning of 2014, with the support of the Honor Frost Foundation, methods were developed and refined to acquire manual photographic data of the entire underwater site of Qaitbay using a DSLR camera, simple and low cost materials to obtain a digital surface model (DSM) of the submerged site of the lighthouse, and also to create 3D models of the objects themselves, such as statues, bases of statues and architectural elements. In this paper we present the methodology used for underwater data acquisition, data processing and modelling in order to generate a DSM of the submerged site of Alexandria’s ancient lighthouse. Until 2016, only about 7200&amp;thinsp;m<sup>2</sup> of the submerged site, which exceeds more than 13000&amp;thinsp;m<sup>2</sup>, was covered. One of our main objectives in this project is to georeference the site since this would allow for a very precise 3D model and for correcting the orientation of the site as regards the real-world space.</p>

scholarly journals FROM ARCHIVE DOCUMENTATION TO ONLINE 3D MODEL VISUALIZATION OF NO LONGER EXISTING STRUCTURES: THE TURIN 1911 PROJECT

2020 ◽  
Vol XLIII-B2-2020 ◽  
pp. 837-844
Author(s):  
D. Einaudi ◽  
A. Spreafico ◽  
F. Chiabrando ◽  
C. Della Coletta
Keyword(s):  
Cultural Heritage ◽  
3D Model ◽  
3D Models ◽  
World's Fair ◽  
Existing Structures ◽  
Present Generation ◽  
Technical Issues ◽  
Digital Documentation ◽  
3D Documentation ◽  
Model Visualization

Abstract. Rebuilding the past of cultural heritage through digitization, archiving and visualization by means of digital technology is becoming an emerging issue to ensure the transmission of physical and digital documentation to future generations as evidence of culture, but also to enable present generation to enlarge, facilitate and cross relate data and information in new ways. In this global effort, the digital 3D documentation of no longer existing cultural heritage can be essential for the understanding of past events and nowadays, various digital techniques and tools are developing for multiple purposes.In the present research the entire workflow, starting from archive documentation collection and digitization to the 3D models metrically controlled creation and online sharing, is considered. The technical issues to obtain a detail 3D model are examined stressing limits and potentiality of 3D reconstruction of disappeared heritage and its visualization exploiting three complexes belonging to 1911 Turin World’s Fair.

scholarly journals VERIFICATION OF POTENCY OF AERIAL DIGITAL OBLIQUE CAMERAS FOR AERIAL PHOTOGRAMMETRY IN JAPAN

2016 ◽  
Vol XLI-B1 ◽  
pp. 63-68 ◽  
Author(s):  
Ryuji Nakada ◽  
Masanori Takigawa ◽  
Tomowo Ohga ◽  
Noritsuna Fujii
Keyword(s):  
3D Model ◽  
3D Models ◽  
Aerial Photographs ◽  
Digital Cameras ◽  
Digital Mapping ◽  
Aerial Photogrammetry ◽  
Public Survey ◽  
Survey Accuracy ◽  
Left And Right ◽  
Simultaneous Adjustment

Digital oblique aerial camera (hereinafter called “oblique cameras”) is an assembly of medium format digital cameras capable of shooting digital aerial photographs in five directions i.e. nadir view and oblique views (forward and backward, left and right views) simultaneously and it is used for shooting digital aerial photographs efficiently for generating 3D models in a wide area. &lt;br&gt;&lt;br&gt; For aerial photogrammetry of public survey in Japan, it is required to use large format cameras, like DMC and UltraCam series, to ensure aerial photogrammetric accuracy. &lt;br&gt;&lt;br&gt; Although oblique cameras are intended to generate 3D models, digital aerial photographs in 5 directions taken with them should not be limited to 3D model production but they may also be allowed for digital mapping and photomaps of required public survey accuracy in Japan. &lt;br&gt;&lt;br&gt; In order to verify the potency of using oblique cameras for aerial photogrammetry (simultaneous adjustment, digital mapping and photomaps), (1) a viewer was developed to interpret digital aerial photographs taken with oblique cameras, (2) digital aerial photographs were shot with an oblique camera owned by us, a Penta DigiCAM of IGI mbH, and (3) accuracy of 3D measurements was verified.

scholarly journals ENHANCED RGB-D MAPPING METHOD FOR DETAILED 3D MODELING OF LARGE INDOOR ENVIRONMENTS

2016 ◽  
Vol III-1 ◽  
pp. 151-158
Author(s):  
Shengjun Tang ◽  
Qing Zhu ◽  
Wu Chen ◽  
Walid Darwish ◽  
Bo Wu ◽  
...  
Keyword(s):  
3D Modeling ◽  
Visual Information ◽  
3D Models ◽  
Measurement Range ◽  
Mapping Method ◽  
Depth Information ◽  
Indoor Environments ◽  
Recovery Method ◽  
Depth Measurement ◽  
Rgb Image

RGB-D sensors are novel sensing systems that capture RGB images along with pixel-wise depth information. Although they are widely used in various applications, RGB-D sensors have significant drawbacks with respect to 3D dense mapping of indoor environments. First, they only allow a measurement range with a limited distance (e.g., within 3&thinsp;m) and a limited field of view. Second, the error of the depth measurement increases with increasing distance to the sensor. In this paper, we propose an enhanced RGB-D mapping method for detailed 3D modeling of large indoor environments by combining RGB image-based modeling and depth-based modeling. The scale ambiguity problem during the pose estimation with RGB image sequences can be resolved by integrating the information from the depth and visual information provided by the proposed system. A robust rigid-transformation recovery method is developed to register the RGB image-based and depth-based 3D models together. The proposed method is examined with two datasets collected in indoor environments for which the experimental results demonstrate the feasibility and robustness of the proposed method

scholarly journals MODEL BASED VIEWPOINT PLANNING FOR TERRESTRIAL LASER SCANNING FROM AN ECONOMIC PERSPECTIVE

2016 ◽  
Vol XLI-B5 ◽  
pp. 607-614
Author(s):  
D. Wujanz ◽  
F. Neitzel
Keyword(s):  
Laser Scanning ◽  
3D Model ◽  
Acquisition Time ◽  
Small Subset ◽  
Economic Perspective ◽  
Object Surface ◽  
Basic Algorithm ◽  
Laser Scanners ◽  
Planning Algorithm ◽  
Viewpoint Planning

Despite the enormous popularity of terrestrial laser scanners in the field of Geodesy, economic aspects in the context of data acquisition are mostly considered intuitively. In contrast to established acquisition techniques, such as tacheometry and photogrammetry, optimisation of the acquisition configuration cannot be conducted based on assumed object coordinates, as these would change in dependence to the chosen viewpoint. Instead, a combinatorial viewpoint planning algorithm is proposed that uses a given 3D-model as an input and simulates laser scans based on predefined viewpoints. The method determines a suitably small subset of viewpoints from which the sampled object surface is preferably large. An extension of the basic algorithm is proposed that only considers subsets of viewpoints that can be registered to a common dataset. After exemplification of the method, the expected acquisition time in the field is estimated based on computed viewpoint plans.

Sign in / Sign up

Export Citation Format

Share Document