Keechul Jung

Frame segmentation used MLP-based X-Y recursive for mobile cartoon content

With rapid growth of the mobile industry, the limitation of small screen mobile is attracting a lot of researchers attention for transforming on/off-line contents into mobile contents. Frame segmentation for limited mobile browsers is the key point of off-line contents tranformation. The X-Y recursive cut algorithm has been widely used for frame segmentation in document analysis. However, this algorithm has drawbacks for cartoon images which have various image types and image with noises, especially the online cartoon contents obtain during scanning. In this paper, we propose a method to segment on/off-line cartoon contents into fitted frames for the mobile screen. This makes the x-y recursive cut algorithm difficult to find the exact cutting point. Therefore we use a method by combining two concepts: an X-Y recursive cut algorithm to extract candidate segmenting positions which shows a good performance on noises free contents, and Multi-Layer Perceptrons (MLP) concept use on candidate for verification. These methods can increase the accuracy of the frame segmentation and feasible to apply on various off-line cartoon images with frames.

Automatic mobile content conversion using semantic image analysis

An approach to knowledge-assisted semantic offline content re-authoring based on an automatic content conversion (ACC) ontology infrastructure is presented. Semantic concepts in the context are defined in ontology, text detection (e.g. connected component based), feature (e.g. texture homogeneity), feature parameter (e.g. texture model distribution), clustered feature (e.g. k-manes algorithm). We will show how the adaptation of the layout can facilitate browsing with mobile devices, especially small-screen mobile phones. In a second stage we address the topic of content personalization by providing a personalization scheme that is based on the ontology technology. Our experiment shows that the proposed ACC is more efficient than the existing methods in providing mobile comic contents.

Flying cake: Augmented game on mobile devices

In the current age of ubiquitous computing via high bandwidth networks, wearable and hand-held mobile devices with small cameras and wireless communication will become widespread in the near future. Thus, research on augmented games for mobile devices has recently attracted a lot of attention. Most existing augmented games use a traditional “backpack” system and “pattern marker”. However, ‘backpack’ systems are expensive, cumbersome, and inconvenient to use, while the use of a pattern marker means the game can only be played at a previously installed location. Hence, this article proposes an augmented game, called Flying Cake, where face regions are used to create virtual objects (characters) without a predefined pattern marker, plus the location of the virtual objects are measured relative to the real world on a small mobile PDA instead of using cumbersome hardware. Flying Cake is an augmented shooting game with two playing modes: (1) single player, where the player attacks a virtual character overlaid on images captured by a PDA camera in the physical world; and (2) two players, where each player attacks a virtual character in an image received via a wireless LAN from their opponent. The virtual character overlaps a face region obtained using a real-time face-detection technique. As a result, Flying Cake provides an exciting experience for players on the basis of a new game paradigm where the user interacts with both the physical world captured by a PDA camera and the virtual world.

Dimension reduction in 3d gesture recognition using meshless parameterization

3D gesture recognition offers more details data but leads to computational hurdles which do not support real-time gesture recognition application. In this paper, we introduce a method of dimension reduction for 3D gesture recognition. Our method uses meshless parameterization to perform dimension reduction in modeling process and extracts gesture data, in order to reduce the computation complexity. In addition, this method also maintains the 3D gesture information and result novel features vectors for 3D gesture recognition. The computational efficiency of dimension reduction and by using novel features vectors makes 3D gesture recognition more possible to achieve real-time performance.

3D posture representation using meshless parameterization with cylindrical virtual boundary

3D data is getting popular which offers more details and accurate information for posture recognition. However, it leads to computational hurdles and is not suitable for real time application. Therefore, we introduce a dimension reduction method using meshless parameterization with cylindrical virtual boundary for 3D posture representation. The meshless parameterization is based on convex combination approach which has good properties, such as fast computation and one-to-one mapping characteristic. This method depends on the number of boundary points. However, 3D posture reconstruction using silhouettes extraction from multiple cameras had resulted various number of boundary points. Therefore, a cylindrical virtual boundary points is introduced to overcome the inconsistency of 3D reconstruction boundary points. The proposed method generates five slices of 2D parametric appearance to represent a 3D posture for recognition purpose.

A novel infrastructure of digital storytelling theme museums based on RFID systems

This paper suggests a storytelling service infrastructure using RFID systems for theme museums with relevant multimedia contents providing an active experience space and drawing more interest. The proposed storytelling system uses RFID tags, a wireless LAN, and a mobile device for the theme museum, and relevant technical issues involving an overall architecture of RFID system will be introduced in this paper. Therefore, the proposed storytelling service infrastructure could be further applicable for games and animation to experience the people realistic scene through providing additional story and virtual information related to each object using RFID, a wireless LAN and a remote storytelling server.

Automatic cartoon image re-authoring using SOFM

According to the growth of the mobile industry, a lot of on/off-line contents are being converted into mobile contents. Although the cartoon contents especially are one of the most popular mobile contents, it is difficult to provide users with the existing on/off-line contents without any considerations due to the small size of the mobile screen. In existing methods to overcome the problem, the cartoon contents on mobile devices are manually produced by computer software such as Photoshop. In this paper, we automatically produce the cartoon contents fitting for the small screen, and introduce a clustering method useful for variety types of cartoon images as a prerequisite stage for preserving semantic meaning. Texture information which is useful for gray-scale image segmentation gives us a good clue for semantic analysis and self-organizing feature maps (SOFM) is used to cluster similar texture information. Besides we automatically segment the clustered SOFM outputs using agglomerative clustering. In our experimental results, combined approaches show good results of clustering in several cartoons.

Real-time image correction for interactive environment

In recent years, projectors are undergoing a transformation as they evolve from static output devices to portable, or communication systems. However, the projection images appear distorted unless the projector is precisely aligned to the projection screen. Generally many projection-based systems are corrected for oblique projection distortion using calibration methods (e.g., warping function). Computing a warping function uses fiducials or a special pattern projected to the screen. These methods are unable to automatically calibrate in real-time especially when a projector is moving. In this paper, we proposed an automatic calibration system in real-time. Our system makes up of a function to correct images that are only based on the degree angle of a mirror set up in front of the projector without attaching fiducials to the screen. Therefore the system can be used to display regardless of screen condition in parallelepiped room. This operation process is very simple and fast. Therefore our system provides a function of correcting movies during the movement of the projector (27.3fps). An advantage of the proposed method is the system can be easily applied to ubiquitous computing like interactive presentation or game.

Dynamic game level generation using on-line learning

In recent years, many researchers are attracted to computer games research. Capable gamers can easily get bored, while beginners tend to give up after trying several times because the game does not correspond to their level of interest. Therefore, this paper proposes that the users play pattern to be modeled on the basis of probability and level designer will dynamically generates the gaming level accordingly. We analyze users play pattern and design pattern based on GMM (probability model) and dynamically generate the level with online learning technique adapting the reinforcement technique. The play pattern is modeled using GMM and in order to create game level dynamically, the method of updating the weight of enemy creation using online script is proposed. Finally, we apply our proposed method to a 2D shooting game and introduce users play pattern leading to design pattern in the game.

Effective image retrieval for the m-learning system

In this paper, we propose augmented learning contents (ALC) with the blended learning on mobile devices. It augments on-line contents by indexing the corresponding off-line contents using traditional pattern recognition method, which results in a minimize of labors for conversion. Among the pattern recognition method marker-based is one of most general approach. However it must reconstruct the off-line contents with pattern markers. To solve both drawbacks that use of the pattern markers and difficulty of the color-based image retrieval by means of a low-resolution PDA camera, we used for a shape-based system. CBIR based on object shapes is used instead of pattern markers to link off-line contents with on-line, and shapes are represented by a differential chain code with estimated new starting points to obtain rotation-invariant representation, which is suited to low computational resources of mobile devices. Consequently, the ALC can provide learner with a fast and accurate multimedia contents (video, audio, text) on static off-line contents using mobile devices without space limitation.