• KIPS Transactions on Software and Data Engineering
• /
• v.11 no.12
• /
• pp.509-516
• /
• 2022

Accurate pose prediction of objects in 3D space is an important visual recognition technique widely used in many applications such as scene understanding in both indoor and outdoor environments, robotic object manipulation, autonomous driving, and augmented reality. Most previous works for object pose estimation have the limitation that they require an exact 3D CAD model for each object. Unlike such previous works, this paper proposes a novel neural network model that can predict the poses of unknown objects based on only their RGB color images without the corresponding 3D CAD models. The proposed model can obtain depth maps required for unknown object pose prediction by using an adaptive depth estimator, AdaBins,. In this paper, we evaluate the usefulness and the performance of the proposed model through experiments using benchmark datasets.

• The Journal of Korea Robotics Society
• /
• v.14 no.2
• /
• pp.139-149
• /
• 2019

In this paper, we present auto-annotation tool and synthetic dataset using 3D CAD model for deep learning based object detection. To be used as training data for deep learning methods, class, segmentation, bounding-box, contour, and pose annotations of the object are needed. We propose an automated annotation tool and synthetic image generation. Our resulting synthetic dataset reflects occlusion between objects and applicable for both underwater and in-air environments. To verify our synthetic dataset, we use MASK R-CNN as a state-of-the-art method among object detection model using deep learning. For experiment, we make the experimental environment reflecting the actual underwater environment. We show that object detection model trained via our dataset show significantly accurate results and robustness for the underwater environment. Lastly, we verify that our synthetic dataset is suitable for deep learning model for the underwater environments.

• Proceedings of the Korean Society of Precision Engineering Conference
• /
• 2003.06a
• /
• pp.1055-1058
• /
• 2003

The robustness and reliability of vision algorithms is the key issue in robotic research and industrial applications. In this paper robust real time visual tracking in complex scene is considered. A common approach to increase robustness of a tracking system is the use of different model (CAD model etc.) known a priori. Also fusion or multiple features facilitates robust detection and tracking of objects in scenes of realistic complexity. Voting-based fusion of cues is adapted. In voting. a very simple or no model is used for fusion. The approach for this algorithm is tested in a 3D Cartesian robot which tracks a toy vehicle moving along 3D rail, and the Kalman filter is used to estimate the motion parameters. namely the system state vector of moving object with unknown dynamics. Experimental results show that fusion of cues and motion estimation in a tracking system has a robust performance.

• Journal of KIISE:Databases
• /
• v.29 no.2
• /
• pp.110-118
• /
• 2002

This paper presents a 3-D graphic database system based on XML that supports content-based retrievals of 3-D images, Most of graphics application systems are currently centered around the processing of 2-D images and research works on 3-D graphics are mainly concerned about the visualization aspects of 3-D image. They do not support the semantic modeling of 3-D objects and their spatial relations. In our data model, 3-D images are represented as compositions of 3-D graphic objects with associated spatial relations. Complex 3-D objects are mode]ed using a set of primitive 3-D objects rather than the lines and polygons that are found in traditional graphic systems. This model supports content-based retrievals of scenes containing a particular object or those satisfying certain spatial relations among the objects contained in them. 3-D images are stored in the database as XML documents using 3DGML DTD that are developed for modeling 3-D graphic data. Finally, this paper describes some examples of query executed in our Web-based prototype database system.

• 제어로봇시스템학회:학술대회논문집
• /
• 2002.10a
• /
• pp.99.5-99
• /
• 2002

$\textbullet$ 3-D object tracking framework $\textbullet$ Using fast stereo vision system for range image $\textbullet$ Using CONDENSATION algorithm to tracking object $\textbullet$ For recognizing object, superquardrics model is used $\textbullet$ Our target object is like coils in steel works

• The Journal of Korea Robotics Society
• /
• v.18 no.3
• /
• pp.346-351
• /
• 2023

With the release of numerous open driving datasets, the demand for domain adaptation in perception tasks has increased, particularly when transferring knowledge from rich datasets to novel domains. However, it is difficult to solve the change 1) in the sensor domain caused by heterogeneous LiDAR sensors and 2) in the environmental domain caused by different environmental factors. We overcome domain differences in the semi-supervised setting with 3-stage model parameter training. First, we pre-train the model with the source dataset with object scaling based on statistics of the object size. Then we fine-tine the partially frozen model weights with copy-and-paste augmentation. The 3D points in the box labels are copied from one scene and pasted to the other scenes. Finally, we use the knowledge distillation method to update the student network with a moving average from the teacher network along with a self-training method with pseudo labels. Test-Time Augmentation with varying z values is employed to predict the final results. Our method achieved 3rd place in ECCV 2022 workshop on the 3D Perception for Autonomous Driving challenge.

• International conference on construction engineering and project management
• /
• 2009.05a
• /
• pp.1468-1473
• /
• 2009

Efficient communication of construction information has been critical for successful project performance. Building Information Modeling (BIM) has appeared as a tool for efficient communication. Through 3D CAD objects, it is possible to check interception and collisions of each object in advance. In addition, 4D simulation based on 3D objects integrated with time information makes it realize to go over scheduling and to perceive potential errors in scheduling. However, current scheduling simulation is still at a stage of animation due to manual integration of 3D objects and scheduling data. Accordingly, this study aims to develop an integrated model of 3D CAD objects that automatically creates scheduling information.

• Journal of the Korean Society for Precision Engineering
• /
• v.23 no.5 s.182
• /
• pp.143-148
• /
• 2006

It is still very difficult to generate a geometry model and finite element model, which has complex and many free surface, even though 3D CAD solutions are applied. Furthermore, in the medical field, which is a big growth area of recent years, there is no drawing. For these reasons, making a geometry model, which is used in finite element analysis, is very difficult. To resolve these problems and satisfy the requests of the need to create a 3D digital file for an object where none had existed before, new technologies are appeared recently. Among the recent technologies, there is a growing interest in the availability of fast, affordable optical range laser scanning. The development of 3D laser scan technology to obtain 3D point cloud data, made it possible to generate 3D model of complex object. To generate CAD and finite element model using point cloud data from 3D scanning, surface reconstruction applications have widely used. In the early stage, these applications have many difficulties, such as data handling, model creation time and so on. Recently developed point-based surface generation applications partly resolve these difficulties. However there are still many problems. In case of large and complex object scanning, generation of CAD and finite element model has a significant amount of working time and effort. Hence, we concerned developing a good direct finite element model generation method using point cloud's location coordinate value to save working time and obtain accurate finite element model.

• The Journal of Korea Robotics Society
• /
• v.4 no.3
• /
• pp.185-191
• /
• 2009

This paper presents a new shape-based algorithm using affine category shape model for object category recognition and model learning. Affine category shape model is a graph of interconnected nodes whose geometric interactions are modeled using pairwise potentials. In its learning phase, it can efficiently handle large pose variations of objects in training images by estimating 2-D homography transformation between the model and the training images. Since the pairwise potentials are defined on only relative geometric relationship betweenfeatures, the proposed matching algorithm is translation and in-plane rotation invariant and robust to affine transformation. We apply spectral matching algorithm to find feature correspondences, which are then used as initial correspondences for RANSAC algorithm. The 2-D homography transformation and the inlier correspondences which are consistent with this estimate can be efficiently estimated through RANSAC, and new correspondences also can be detected by using the estimated 2-D homography transformation. Experimental results on object category database show that the proposed algorithm is robust to pose variation of objects and provides good recognition performance.

• Journal of the Korean Institute of Intelligent Systems
• /
• v.9 no.4
• /
• pp.436-443
• /
• 1999

Most of the 3D object reconstruction techniques divide the object into multiplane and approximate the surfaces of the object. The Marching Cubes Algorithm which initializes the mesh structure using a given isovalue. and Delaunay Tetrahedrisation are widely used. Deformable models are well-suited for general object reconstruction because they make little assumptions about the shape to recover and they can reconstruct objects *om various types of datasets. Now, many researchers are studying the reconstruction systems based on a deformable model. In this paper, we propose a novel method for reconstruction of 3D objects. This method, for a 3D object composed of curved planes, compresses the 3D object based on the adaptive simplexmesh technique. It changes the pre-defined mesh structure, so that it may approach to the original object. Also, we redefine the geometric characteristics such as curvatures. As results of simulations, we show reconstruction of the original object with high compression and concentration of vertices towards parts of high curvature in order to optimize the shape description.