• Journal of Institute of Control, Robotics and Systems
• /
• v.13 no.8
• /
• pp.804-809
• /
• 2007

Initialization problem is defined for indoor mobile robot as a whole process from arrival to normal operation in a new environment. The unstructured environment make the process much more difficult compared to industrial robot in structured environments. We propose a simple and efficient initialization process based on artificial stars on ceiling. Important task points and paths connecting task points are defined based on the corresponding artificial stars. This approach can be used for all kinds of indoor mobile robots with landmarks used for indoor localization.

• The Journal of Korean Institute of Electromagnetic Engineering and Science
• /
• v.11 no.2
• /
• pp.177-185
• /
• 2000

The broadband EM absorbers composed of dielectrics and magnetic materials are designed. The performance of the partial initialization genetic algorithm(PIGA) is improved with three factors such as partial initialization ratios, initialization starting points and scale factors. At the frequency range over 3∼10 GHz, the optimized electromagnetic absorbers designed by using the improved PIGA. The design results obtained by enhanced PIGA's are superior to that of the using GA presented by E. Michielssen with regard to the total depth of composite materials, the depth of magnetics and maximum reflection coefficients.

• International Journal of Advanced Culture Technology
• /
• v.7 no.4
• /
• pp.283-288
• /
• 2019

Neural networks have been reborn as a Deep Learning thanks to big data, improved processor, and some modification of training methods. Neural networks used to initialize weights in a stupid way, and to choose wrong type activation functions of non-linearity. Weight initialization contributes as a significant factor on the final quality of a network as well as its convergence rate. This paper discusses different approaches to weight initialization. MNIST dataset is used for experiments for comparing their results to find out the best technique that can be employed to achieve higher accuracy in relatively lower duration.

• Journal of the Korean Society for Railway
• /
• v.7 no.1
• /
• pp.9-13
• /
• 2004

Railway passenger demand forecasts may be used directly, or as inputs to other optimization models use them to produce estimates of other activities. The optimization models require demand forecasts at the most detailed level. In this environment exponential smoothing forecasting methods such as Holt-Winters are appropriate because it is simple and inexpensive in terms of computation. There are several initialization methods for Holt-Winters Model. The purpose of this paper is to compare the initialization methods for Holt-Winters model.

• Journal of Institute of Control, Robotics and Systems
• /
• v.22 no.10
• /
• pp.833-838
• /
• 2016

This paper presents a performance evaluation of various feature-initialization algorithms for forward-viewing mono-camera based simultaneous localization and mapping (SLAM), specifically in indoor environments. For mono-camera based SLAM, the position of feature points cannot be known from a single view; therefore, it should be estimated from a feature initialization method using multiple viewpoint measurements. The accuracy of the feature initialization method directly affects the accuracy of the SLAM system. In this study, four different feature initialization algorithms are evaluated in simulations, including linear triangulation; depth parameterized, linear triangulation; weighted nearest point triangulation; and particle filter based depth estimation algorithms. In the simulation, the virtual feature positions are estimated when the virtual robot, containing a virtual forward-viewing mono-camera, moves forward. The results show that the linear triangulation method provides the best results in terms of feature-position estimation accuracy and computational speed.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• v.11 no.1
• /
• pp.123-145
• /
• 2017

Due to the increasing demand for spectrum resources, cognitive radio networks and dynamic spectrum access draw a lot of research into efficiently utilizing limited spectrum resources. To set up cluster-based CR ad-hoc common channels, conventional methods require a relatively long time to successfully exchange the initialization messages. In this paper, we propose a fast and reliable common channel initialization protocol for CR ad-hoc networks. In the proposed method, the cluster head sequentially broadcasts a system activation signal through its available channels with a predetermined correlation pattern. To detect the cluster head's broadcasting channels and to join the cluster, each member node implements fast Fourier transform (FFT) and computes autocorrelation of an FFT bin sequence for each available channel of the member node. This is compared to the predetermined reference pattern. The join request and channel decision procedures are also presented in this paper. In a simulation study, the performance of the proposed method is evaluated.

• Journal of Korea Multimedia Society
• /
• v.21 no.12
• /
• pp.1396-1406
• /
• 2018

Lane detection is a key function in developing autonomous vehicle technology. In this paper, we propose a lane marker detection algorithm robust to environmental variation targeting low cost embedded computing devices. The proposed algorithm consists of two phases: initialization phase which is slow but has relatively higher accuracy; and the tracking phase which is fast and has the reliable performance in a limited condition. The initialization phase detects lane markers using a set of filters utilizing the various features of lane markers. The tracking phase uses Kalman filter to accelerate the lane marker detection processing. In a tracking phase, we measure the reliability of the detection results and switch it to initialization phase if the confidence level becomes below a threshold. By combining the initialization and tracking phases we achieved high accuracy and acceptable computing speed even under a low cost computing resources in which we cannot use the computing intensive algorithm such as deep learning approach. Experimental results show that the detection accuracy is about 95% on average and the processing speed is about 20 frames per second with Raspberry Pi 3 which is low cost device.

• 제어로봇시스템학회:학술대회논문집
• /
• 2005.06a
• /
• pp.2042-2047
• /
• 2005

Many service robots which is interacting with human at home and in buildings have been developed. Few of them are shown in of the United States and of Japan. These robots are supposed to have a powerful indoor navigation performance in places where human beings live and work. The overall capability of service robots to move around in this environment is called environment correspondence, in which localization problem to find the accurate position and orientation is the most critical problem. While users set up a proper or a best environment for industrial robots, but for services robots at home and in buildings, it is very difficult to change the environment for robots. The expanded workspace due to mobility is difficult to be covered by means of those used for industrial robots because the cost increases and human beings do not want their environment to be changed for robots. This fact has made many researchers study efficient and effective environment correspondence problems. Among these problems, localization is the most difficult. Goal of localization study includes (1) Accurate detection of position and orientation (2) Minimum cost of the additional devices (3) Minimum change of human environment. In this study, as a solution of the above, we propose "Artificial Stars" which are attached on room ceiling as landmarks. In addition, we solve an adoption problem raised when a robot is delivered to a customer site and before it can perform its full navigation capability. We call this as "Initialization Problem" of service robots. We solve the initialization problem for both cases of environment with the map and without map. The proposed system is experimented and has shown how well it handles the initialization problem.

• Atmosphere
• /
• v.32 no.4
• /
• pp.395-409
• /
• 2022

A new soil moisture initialization scheme is applied to the Korea Meteorological Administration (KMA) Global Seasonal forecasting system version 6 (GloSea6). It is designed to ingest the microwave soil moisture retrievals from Soil Moisture Active Passive (SMAP) radiometer using the Local Ensemble Transform Kalman Filter (LETKF). In this technical note, we describe the procedure of the newly-adopted initialization scheme, the change of soil moisture states by assimilation, and the forecast skill differences for the surface temperature and precipitation by GloSea6 simulation from two preliminary experiments. Based on a 4-year analysis experiment, the soil moisture from the land-surface model of current operational GloSea6 is found to be drier generally comparing to SMAP observation. LETKF data assimilation shows a tendency toward being wet globally, especially in arid area such as deserts and Tibetan Plateau. Also, it increases soil moisture analysis increments in most soil levels of wetness in land than current operation. The other experiment of GloSea6 forecast with application of the new initialization system for the heat wave case in 2020 summer shows that the memory of soil moisture anomalies obtained by the new initialization system is persistent throughout the entire forecast period of three months. However, averaged forecast improvements are not substantial and mixed over Eurasia during the period of forecast: forecast skill for the precipitation improved slightly but for the surface air temperature rather degraded. Our preliminary results suggest that additional elaborate developments in the soil moisture initialization are still required to improve overall forecast skills.

• 제어로봇시스템학회:학술대회논문집
• /
• 2003.10a
• /
• pp.752-757
• /
• 2003

Recently, 3D structure recovery through self-calibration of camera has been actively researched. Traditional calibration algorithm requires known 3D coordinates of the control points while self-calibration only requires the corresponding points of images, thus it has more flexibility in real application. In general, self-calibration algorithm results in the nonlinear optimization problem using constraints from the intrinsic parameters of the camera. Thus, it requires initial value for the nonlinear minimization. Traditional approaches get the initial values assuming they have the same intrinsic parameters while they are dealing with the situation where the intrinsic parameters of the camera may change. In this paper, we propose new initialization method using the minimum 2 images. Proposed method is based on the assumption that the least violation of the camera’s intrinsic parameter gives more stable initial value. Synthetic and real experiment shows this result.