Kyungboo Jung

Virtual Tactical Map with Tangible Augmented Reality Interface

We propose a new augmented reality (AR)-based approach for training using virtual sand table representations of military battlefields. A virtual tactical map (VTM) can archive simple actions such as moving a marker by hand, to provide more organic realizations in virtual military training. The new tangible AR interface provides a content-authoring tool that is natural, intuitive, and user-friendly. AR-interfaced VTMs exhibit multiple possibilities for military learning and training applications.

Interactive auto-stereoscopic display with efficient and flexible interleaving

We propose a system that displays a three-dimensional (3D) scene on a 3D display device, consisting of a flat-panel display and a slanted parallax barrier, which can directly manipulate virtual 3D objects in real time. The proposed system consists of a 3D scene display part and a user interaction part. First, we propose a multi-view image interleaving method that considers the distance between a user and a display device for a slanted parallax barrier-based 3D display. Second, we define hand motion parameters for convenient hand motion analysis, and we propose a hand gesture recognition method that uses dynamic time warping (DTW) with specified hand motion parameters for user interaction. Thus, the proposed system makes it possible to see in any position, and viewers can perform event actions without regard to the shape or timing of hand motions.

SERP: SURF enhancer for repeated pattern

This paper proposes an object-matching method for repetitive patterns. Mismatching problems occur when descriptor-based features like SURF or SIFT are applied to repeated image patterns due to the use of the usual distance-ratio test. To overcome this, we first classify SURF descriptors in the image using mean-shift clustering. The repetitive features are grouped into a single cluster, and each non-repetitive feature has its own cluster. We then evaluate the similarity between the converged modes (descriptors) resulting from mean-shift clustering. We thus generate a new descriptor space that has a distinct and reliable descriptor for each cluster, and we use these to find correlations between images. We also calculate the homography between two images using the descriptors to guarantee correctness of the match. Experiments with repeated patterns show that this method improves recognition rates. This paper shows the results of applying this method to building recognition; the technique can be extended to matching various repeated patterns in textiles and geometric patterns.

Marker-based tangible interfaces for 3D reconstruction

In this paper, we proposed a novel marker-based tangible interface for various users to manipulate the object with intuitive and simple approaches during an authoring application of augmented reality. The proposed method makes use of marker as intuitive interface to obtain 3D geometric information of 3D reconstruction. 3D geometric information of an object surface is acquired by touching the object directly with the proposed tangible interfaces. The tangible interfaces not only support 3D reconstruction for graphical modeling but also offer features information which is used for augmented reality. Finally, we verify efficiency of the proposed method with demonstration of an augmented reality application using the proposed method.

A Multi-View Images Interleaving for Slanted Parallax Barrier based Display Device

Flat panel-based parallax barrier or lenticular based 3D display devices that have been developed recently are designed to feel depth. In order to see a three-dimensional(3D) image by the display device, a multi-view image displayed on the flat panel must be regenerated from images of multi-views using a subsampling method. Previous subsampling methods are focused on reducing crosstalk. In this paper, we focus on a misalignment that is occurred on manufacture process of slanted parallax barrier based autostereoscopic display device. Therefore, we propose a interleaving method that considers tilt of slanted parallax barrier, aperture size, and distance between an autostereoscopic display device and a viewer to see a 3D image regardless of a viewer position.

Adaptive Hand Color Extraction Using Wave Motion Detection

Correct hand segmentation in image sequences is the most important aspect of hand-gesture interfaces. Existing hand-segmentation methods may be classified into two general approaches. One uses predefined skin color and the other uses a learning process. The skin color-learning method gives good results but its excessive processing time precludes its use in real time. In the predefined color approach, automatically modifying the predefined color when the environment changes would increase convenience and efficiency. Identifying a change in the environment usually requires user intervention. We propose a method by which the user can easily modify the predefined skin color any time in any environment simply by waving his/her hand in front of the camera. The hand interface system detects the waving hand region and efficiently extracts the hand color for use in hand-gesture interfaces.

Virtual Duplication of Rigid Objects from Image Sequences

Augmented reality is a type of reproduction that adds virtual objects to real scenes. Recently, remarkable growth in the use of augmented reality has been reported and a number of applications are being developed. Although numerous applications have been proposed, many challenges remain. In particular, lack of harmony between the virtual object and real scene is crucial to the performance of augmented reality. Therefore, research into ways to improve this drawback is required. In this paper, we propose a novel paradigm called ldquovirtual duplicationrdquo that enables the movement of objects freely between virtual and real worlds. The proposed method copies objects from a real scene directly. It is possible to generate harmonious objects because the method uses the texture of a real scene. Furthermore, three-dimensional (3D) information is easier to obtain using the proposed environment for acquiring 3D models. Unlike previous methods, our proposed method does not depend on the users ability.