• IEMEK Journal of Embedded Systems and Applications
• /
• v.16 no.5
• /
• pp.215-223
• /
• 2021

3D data can be categorized into two parts : Euclidean data and non-Euclidean data. In general, 3D data exists in the form of non-Euclidean data. Due to irregularities in non-Euclidean data such as mesh and point cloud, early 3D deep learning studies transformed these data into regular forms of Euclidean data to utilize them. This approach, however, cannot use memory efficiently and causes loses of essential information on objects. Thus, various approaches that can directly apply deep learning architecture to non-Euclidean 3D data have emerged. In this survey, we introduce various deep learning methods for mesh and point cloud data. After analyzing the operating principles of these methods designed for irregular data, we compare the performance of existing methods for shape classification and segmentation tasks.

• Smart Structures and Systems
• /
• v.30 no.5
• /
• pp.501-511
• /
• 2022

Three-dimensional (3D) models have become crucial for improving civil infrastructure analysis, and they can be used for various purposes such as damage detection, risk estimation, resolving potential safety issues, alarm detection, and structural health monitoring. 3D point cloud data is used not only to make visual models but also to analyze the states of structures and to monitor them using semantic data. This study proposes automating the generation of high-quality 3D point cloud data and removing noise using deep learning algorithms. In this study, large-format aerial images of civilian infrastructure, such as cut slopes and dams, which were captured by drones, were used to develop a workflow for automatically generating a 3D point cloud model. Through image cropping, downscaling/upscaling, semantic segmentation, generation of segmentation masks, and implementation of region extraction algorithms, the generation of the point cloud was automated. Compared with the method wherein the point cloud model is generated from raw images, our method could effectively improve the quality of the model, remove noise, and reduce the processing time. The results showed that the size of the 3D point cloud model created using the proposed method was significantly reduced; the number of points was reduced by 20-50%, and distant points were recognized as noise. This method can be applied to the automatic generation of high-quality 3D point cloud models of civil infrastructures using aerial imagery.

• Journal of the Korea Institute of Information and Communication Engineering
• /
• v.26 no.11
• /
• pp.1660-1665
• /
• 2022

Recently, many industries are using three dimensional scanning technology. As the performance of the 3D scanner gradually improves, a sampling step to reduce a point data or a remove step to remove a part determined to be noise are generally performed in post processing. However, total point data by long time scanning cannot be processed at once in spite of performing such those additional processes. In general, a method using a multi threaded environment is widely used, but as the scanning process work time increases, the processing performance gradually decreases due to various environmental conditions and accumulated operations. This paper proposes a method to initially remove point data judged to be unnecessary by calculating accumulated fast point feature histogram values from coming point data of the 3D scanner in real time. The entire 3D scanning process can be reduced using this approach.

• International Journal of Fuzzy Logic and Intelligent Systems
• /
• v.14 no.2
• /
• pp.84-90
• /
• 2014

This paper presents the results of three-dimensional face point cloud smoothing based on a modified anisotropic diffusion method. The focus of this research was to obtain a 3D face point cloud with a smooth texture and number of vertices equal to the number of vertices input during the smoothing process. Different from other methods, such as using a template D face model, modified anisotropic diffusion only uses basic concepts of convolution and filtering which do not require a complex process. In this research, we used 6D point cloud face data where the first 3D point cloud contained data pertaining to noisy x-, y-, and z-coordinate information, and the other 3D point cloud contained data regarding the red, green, and blue pixel layers as an input system. We used vertex selection to modify the original anisotropic diffusion. The results show that our method has improved performance relative to the original anisotropic diffusion method.

• Proceedings of the IEEK Conference
• /
• 2003.11a
• /
• pp.183-186
• /
• 2003

A vision sensor should be calibrated prior to infer a Euclidian shape reconstruction. A point to point calibration. also referred to as a hard calibration, estimates calibration parameters by means of a set of 3D to 2D point pairs. We proposed a new method for determining a set of 3D to 2D pairs for the structured light hard calibration. It is simply determined based on epipolar geometry between camera image plane and projector plane, and a projector calibrating grid pattern. The projector calibration is divided two stages; world 3D data acquisition Stage and corresponding 2D data acquisition stage. After 3D data points are derived using cross ratio, corresponding 2D point in the projector plane can be determined by the fundamental matrix and horizontal grid ID of a projector calibrating pattern. Euclidian reconstruction can be achieved by linear triangulation. and experimental results from simulation are presented.

• Journal of the Korea Institute of Information and Communication Engineering
• /
• v.23 no.11
• /
• pp.1482-1484
• /
• 2019

We present a method for transmitting 3D object information in real time in a telepresence system. Three-dimensional object information consists of a large amount of point cloud data, which requires high performance computing power and ultra-wideband network transmission environment to process and transmit such a large amount of data in real time. In this paper, multiple users can transmit object motion and facial expression information in real time even in small network bands by using GANs (Generative Adversarial Networks), a non-supervised learning machine learning algorithm, for real-time transmission of 3D point cloud data. In particular, we propose the creation of an object similar to the original using only the feature information of 3D objects using conditional GANs.

• Journal of the Korean Association of Geographic Information Studies
• /
• v.21 no.4
• /
• pp.132-144
• /
• 2018

3D Geo-spatial information models have been widely used in the field of Civil Engineering, Medical, Computer Graphics, Urban Management and many other. Especially, in surveying and geo-spatial field, the demand for high quality 3D geospatial information and indoor spatial information is so highly increasing. However, it is so difficult to provide a low-cost and high efficiency service to the field which demand the highest quality of 3D model, because pre-constructed spatial data are composed of different formats and storage structures according to the application purpose of each institutes. In fact, the techniques to construct a high applicable 3D geo-spatial model is very expensive to collect and analyze geo-spatial data, but most demanders of 3D geo-spatial model never want to pay the high-cost to that. This study, therefore, suggest the effective way to construct 3D geo-spatial model with low-cost of construction. In general, the effective way to reduce the cost of constructing 3D geo-spatial model as presented in previous studies is to combine the raw data obtained from point cloud observatory and UAV imagery, however this method has some limitation of usage from difficulties to approve the use of raw data because of those have been managed separately by various institutes. To solve this problem, we developed the linking & management system for unifying a high-Resolution raw geo-spatial data based on the point cloud DB and apply this system to extract the basic database from 3D geo-spatial mode for the road database registration. As a result of this study, it can be provided six contents of main entries for road registration by applying the developed system based on the point cloud DB.

• Cartoon and Animation Studies
• /
• s.48
• /
• pp.199-215
• /
• 2017

Recently, a drone is used for the general purpose application although the drone was builtfor the military purpose. A drone is actively used for the creation of contents, and an image acquisition. In this paper, we develop a 3D library module platform using 3D mesh model data, which is generated by a drone image and its point cloud. First, a lot of 2D image data are taken by a drone, and a point cloud data is generated from 2D drone images. A 3D mesh data is acquired from point cloud data. Then, we develop a service library platform using a transformed 3D data for multi-purpose uses. Our platform with 3D data can minimize the cost and time of contents creation for special effects during the production of a movie, drama, or documentary. Our platform can contribute the creation of experts for the digital contents production in the field of a realistic media, a special image, and exhibitions.

• Journal of Korea Multimedia Society
• /
• v.23 no.2
• /
• pp.166-173
• /
• 2020

Multi-view images, which are widely used for providing free-viewpoint services, can enhance the quality of synthetic views when the number of views increases. However, there needs an efficient representation method because of the tremendous amount of data. In this paper, we propose a method for generating point cloud data for the efficient representation of multi-view color and depth images. The proposed method conducts sequential reconstruction of point clouds at each viewpoint as a method of deleting duplicate data. A 3D point of a point cloud is projected to a frame to be reconstructed, and the color and depth of the 3D point is compared with the pixel where it is projected. When the 3D point and the pixel are similar enough, then the pixel is not used for generating a 3D point. In this way, we can reduce the number of reconstructed 3D points. Experimental results show that the propose method generates a point cloud which can generate multi-view images while minimizing the number of 3D points.

• Journal of Information Processing Systems
• /
• v.12 no.3
• /
• pp.338-357
• /
• 2016

The recent advent of increasingly affordable and powerful 3D scanning devices capable of capturing high resolution range data about real-world objects and environments has fueled research into effective 3D surface reconstruction techniques for rendering the raw point cloud data produced by many of these devices into a form that would make it usable in a variety of application domains. This paper, therefore, provides an overview of the existing literature on surface reconstruction from 3D point clouds. It explains some of the basic surface reconstruction concepts, describes the various factors used to evaluate surface reconstruction methods, highlights some commonly encountered issues in dealing with the raw 3D point cloud data and delineates the tradeoffs between data resolution/accuracy and processing speed. It also categorizes the various techniques for this task and briefly analyzes their empirical evaluation results demarcating their advantages and disadvantages. The paper concludes with a cross-comparison of methods which have been evaluated on the same benchmark data sets along with a discussion of the overall trends reported in the literature. The objective is to provide an overview of the state of the art on surface reconstruction from point cloud data in order to facilitate and inspire further research in this area.