Vol.8 No.2 June 15, 2012

Recovering Drawing Order of Single-Stroke Handwritten Images Using Probabilistic Tabu Search (073-087)
        Takayuki Nagoya and Hiroyuki Fujioka
This paper considers the problem for recovering a drawing order of static handwritten images with single stroke. Such a stroke may include the so-called double-traced lines (D-lines). The problem is analyzed and solved by employing the graph theoretic approach. Then the central issue is to obtain the smoothest path of stroke from a graph model of input handwritten images. First, the graph model is constructed from an input images by image processing techniques. Then, we locally analyze the structure of graph. In particular, the method to identify D-lines is developed by introducing the idea of `D-line index'. The method enables us to transform any graphs with D-lines to semi-Eulerian graphs. Then, the restoration problem reduces to the problem of globally computing the maximum weight collection of perfect matchings. For solving such a problem, we propose a method using a probabilistic tabu search algorithm. The effectiveness and usefulness of the proposed method are examined by some experimental studies.

Analysis of Filed Intensity Distribution in Inhomogeneous Propagation Environment Based on Two-Ray Model (088-104)
        Junichi Honda, Kazunori Uchida, and Masafumi Takemats
This paper is concerned with an analysis of field intensity distribution caused by sensor nodes located on inhomogeneous terrestrial surfaces. First, we introduce 1-ray and 2-ray models with two modification factors to estimate the field intensity distribution. One of the two factors is an amplitude modification ($\alpha$) and the other is a distance order of propagation ($\beta$). By using the two factors, we can calculate field intensity distributions in complicated natural environment such as random rough surface. Then, we propose an estimation formula for analyzing electric fields in inhomogeneous propagation environments based on the conventional two models. Next we introduce an algorithm for radio communication distance based on the 1-ray and 2-ray models. In the numerical examples, we show the field intensity distribution caused by sensor nodes located randomly on inhomogeneous terrestrial surfaces, using 2-ray models. Finally, we discuss how many sensor nodes are needed to cover the field area in order to construct networks.

Study on X-Shaped Photonic Crystal Waveguide (105-113)
        Hiroshi Maeda, Yongmei Zhang, and Hiroyuki Terashima
For wavelength division multiplexing system, X-shaped waveguide situated in two dimensional pillar type photonic crystal structure was studied in this paper. Propagation and filtering characteristics of some types of the waveguide were experimentally investigated. Measurement was done by using a model in microwave frequency around 4 GHz with lattice period P=26.5 mm, diameter of pillar $\phi$=7.5 mm and its dielectric constant $\varepsilon_r$=36.0. First, symmetric X waveguide was measured to show its basic transmission characteristics. After the measurement, three cavities with a pair of dielectric rods were situated in each output waveguide to perform a filtering function. Next, asymmetric X waveguide with cavities was examined to improve filtering characteristics in each output port. For the asymmetric structure, higher extinction ratio and higher Q-factor were obtained, compared to symmetric structure. Finally, filtering characteristics of the asymmetric X-shaped waveguide was compared with the result of Y-shaped cascaded waveguide with 4 outputs. The asymmetric X-shaped waveguide showed better results with higher resonant peak amplitude than that of Y-shaped waveguide. Applying good resolution characteristics of 40MHz with respect to input frequency, proposed asymmetric X-branch circuit is available for signal add/drop device for wavelength division multiplexing (WDM) system.

A Tangible 3D Desktop Environment with Force Feedback (114-131)
        Yusuke Ouchi, Hiroaki Nishino, Tsuneo Kagawa, and Kouichi Utsumiya
Various 3D desktop environments have been proposed and implemented in recent years. They don’t, however, become a popular interaction method in HCI (human computer interface) research field. Restrictions appeared in mouse-based 3D operations such as difficulties in appropriately designating object’s positions and directions in a 3D task space are major obstacles. In this paper, we propose to replace the mouse by a haptic device, an interaction device enabling easy object manipulations in the 3D task space. The haptic device also allows users to touch the objects with depth perceptions and feel some tactile sensations. It presents some special effects like shocks, vibrations, and weights when they touch the objects. We designed and developed a 3D operational environment allowing the users to treat the desktop interface as if it is a real world property coexisting with the haptic device. We describe the implementation method and operational interface of the proposed new 3D desktop environment empowered by the haptic device and two experiments conducted for verifying the effectiveness of the proposed system.

Implicit Context Awareness by Face Recognition (132-148)
        Seiji Takeda, Tsutomu Terada, and Masahiko Tsukamoto
In recent years, technical improvements to sensors have attracted a great deal of attention, in particular due to the sensors' capability recognizing user contexts. In this paper, we propose an implicit context awareness system that identifies user context by sensing the context of surrounding environments. We implemented a prototype that recognizes user contexts by sensing surrounding people by two cameras We actually used the prototype in a variety of situations. Evaluation results showed that the system was effective and improved context recognition. Our method can be used to identify rich contexts that cannot be recognized by conventional methods.

Back to JMM Online Front Page