• Journal of the Korean Institute of Telematics and Electronics B
• /
• v.29B no.1
• /
• pp.94-101
• /
• 1992

The multilayer perceptron, trained by the error back-propagation learning rule, has been known as a mapping network which can represent arbitrary functions. However depending on the complexity of a function and the initial weights of the multilayer perceptron, the error back-propagation learning may fall into a local minimum or a flat area which may require a long learning time or lead to unsuccessful learning. To solve such difficulties in training the multilayer perceptron by standard error back-propagation learning rule, the paper proposes a learning method which progressively enlarges the learning domain from a small area to the entire region. The proposed method is devised from the investigation on the roles of hidden nodes and connection weights in the multilayer perceptron which approximates a function of one variable. The validity of the proposed method was illustrated through simulations for a function of one variable and a function of two variable with many extremal points.

• Communications for Statistical Applications and Methods
• /
• v.17 no.6
• /
• pp.845-852
• /
• 2010

The outcomes of counts commonly occur in the area of disease mapping for mortality rates or disease rates. A Poisson distribution is usually assumed as a model of disease rates in conjunction with a gamma prior. The small area typically refers to a small geographical area or demographic group for which very little information is available from the sample surveys. Under this situation the model-based estimation is very popular, in which the auxiliary variables from various administrative sources are used. The empirical Bayes estimator under Poissongamma model has been considered with its accuracy measures. An accuracy measure using a bootstrap samples adjust the underestimation incurred by the posterior variance as an estimator of true mean squared error. We explain the suggested method through a practical dataset of hitters in baseball games. We also perform a Monte Carlo study to compare the accuracy measures of mean squared error.

• Journal of the Korean Society for Precision Engineering
• /
• v.13 no.10
• /
• pp.78-87
• /
• 1996

In this paper, a remeshing criterion has been suggested in order to order to automate the remeshing decision during finite element analysis of metal forming. In order to use for the remeshing decision, two of remeshing criteria have been investigated. One is the use of error estimates based on errors in stresses and strain rate of the finite element solution and the other is the use of geometric characterisreics of distorted elements. As a result, the strain rate error estimate in power norm based on the former is found to give more valuable information about remeshing decision than the ones based on the latter. Examples are given to demon- strate the usefulness of the suggested eroor estimate as a remeshing criterion.

• Korean Journal of Computational Design and Engineering
• /
• v.9 no.4
• /
• pp.397-405
• /
• 2004

Texture mapping is the process of covering 3D models with texture images in order to increase the visual realism of the models. For proper mapping the coordinates of texture images need to coincide with those of the 3D models. When projective images from the camera are used as texture images, the texture image coordinates are defined by a camera calibration method. The texture image coordinates are determined by the relation between the coordinate systems of the camera image and the 3D object. With the projective camera images, the distortion effect caused by the camera lenses should be compensated in order to get accurate texture coordinates. The distortion effect problem has been dealt with iterative methods, where the camera calibration coefficients are computed first without considering the distortion effect and then modified properly. The methods not only cause to change the position of the camera perspective line in the image plane, but also require more control points. In this paper, a new iterative method is suggested for reducing the error by fixing the principal points in the image plane. The method considers the image distortion effect independently and fixes the values of correction coefficients, with which the distortion coefficients can be computed with fewer control points. It is shown that the camera distortion effects are compensated with fewer numbers of control points than the previous methods and the projective texture mapping results in more realistic image.

• The Journal of Korea Institute of Information, Electronics, and Communication Technology
• /
• v.3 no.1
• /
• pp.38-46
• /
• 2010

This paper describes a computer vision system based on active IR illumination for real-time gaze discrimination system. Unlike most of the existing gaze discrimination techniques, which often require assuming a static head to work well and require a cumbersome calibration process for each person, our gaze discrimination system can perform robust and accurate gaze estimation without calibration and under rather significant head movement. This is made possible by a new gaze calibration procedure that identifies the mapping from pupil parameters to screen coordinates using generalized regression neural networks (GRNNs). With GRNNs, the mapping does not have to be an analytical function and head movement is explicitly accounted for by the gaze mapping function. Furthermore, the mapping function can generalize to other individuals not used in the training. To further improve the gaze estimation accuracy, we employ a reclassification scheme that deals with the classes that tend to be misclassified. This leads to a 10% improvement in classification error. The angular gaze accuracy is about $5^{\circ}$horizontally and $8^{\circ}$vertically. The effectiveness of our gaze tracker is demonstrated by experiments that involve gaze-contingent interactive graphic display.

• Journal of the Korean Society of Surveying, Geodesy, Photogrammetry and Cartography
• /
• v.36 no.2
• /
• pp.75-84
• /
• 2018

The purpose of this study is that the low-cost mobile mapping system using INS (Inertial Navigation System) based on MEMS (Micro Electro Mechanical System) could decipher the interpretation of road facility with the accuracy of x, y 0.546m plane error. Even though the MMS (Mobile Mapping System) technology as a new measurement technology has been used vividly to set up geographic information by some world leading surveying equipment manufacturers, the domestic technology is still in its beginning stage. Several domestic institutes and companies tried to catch up the leading technology but they just produced prototypes which needs more stabilization. Through this thesis, we developed low-cost mobile mapping system installed with INS based on MEMS after time synchronizing sensors for MMS such as LiDAR (Light Detection And Ranging), CCD (Charge Coupled Device), GPS/INS (Global Positioning System / Inertial Navigation System) and DMI (Distance Measurement Instrument).

• Journal of The Korean Association For Science Education
• /
• v.29 no.8
• /
• pp.861-873
• /
• 2009

This study investigated the types of analogies generated by science-gifted students, mapping errors on the chromatography, and the perceptions on generating analogy. The subjects were science-gifted eighth-graders (N=68) enrolled at four gifted-education centers in Seoul. Analyses of the results revealed that most science-gifted students represented the analogies using verbal and pictorial forms based on concrete and everyday experiences, and they elaborately explained them including the functional attributes and the causal relationships of the target concept. Science-gifted students selected the analogies having rich similarities or similar structure and principle between the target concept and the analogy as the best of the self-generated analogies. And they used mainly their 'school life' as analogy materials. The results of the mapping test showed that many science-gifted students had mapping errors such as 'failure to map' and 'overmapping'. They were found to have the positive perceptions on generating analogy. The present study will provide the basic information to develop an instructional model in generating analogy in an education program for science-gifted students.

• The Journal of the Institute of Internet, Broadcasting and Communication
• /
• v.9 no.5
• /
• pp.123-127
• /
• 2009

In this paper, we suggested new additional data transmission technology for T-DMB system using MOK-SS(Mary Orthogonal Keying- Spread Spectrum)mapping method. We applied revised watermarking method of TxID(Transmitter Identification) technology mainly had been used for conventional ATSC-DTV(Advanced Television Systems Committee - Digital TV). And proposed schemes enhanced additional data rate using MOK-SS mapping method. We certified availability of our proposed technology by using various simulation and error rate performance analysis.

• Journal of the Korean Institute of Telematics and Electronics B
• /
• v.31B no.2
• /
• pp.45-55
• /
• 1994

In this paper, a block based image approximation technique using the Self Affine System(SAS) from the fractal theory is suggested. Each block of an image is divided into 4 tiles and 4 affine mapping coefficients are found for each tile. To find the affine mapping cefficients that minimize the error between the affine transformed image block and the reconstructed image block, the matrix euation is solved by setting each partial differential coefficients to aero. And to ensure the convergence of coding block. 4 uniformly partitioned affine transformation is applied. Variable block size technique is employed in order to applynatural image reconstruction property of fractal image coding. Large blocks are used for encoding smooth backgrounds to yield high compression efficiency and texture and edge blocks are divided into smaller blocks to preserve the block detail. Affine mapping coefficinets are found for each block having 16$\times$16, 8$\times$8 or 4$\times$4 size. Each block is classified as shade, texture or edge. Average gray level is transmitted for shade bolcks, and coefficients are found for texture and edge blocks. Coefficients are quantized and only 16 bytes per block are transmitted. Using the proposed algorithm, the computational load increases linearly in proportion to image size. PSNR of 31.58dB is obtained as the result using 512$\times$512, 8 bits per pixel Lena image.

• Transactions of the Korean Society of Mechanical Engineers A
• /
• v.28 no.3
• /
• pp.230-237
• /
• 2004

Lane Sensing techniques based on vision sensors are regarded promising because they require little infrastructure on the highway except clear lane markers. However, they require more intelligent processing algorithms in vehicles to generate the previewed roadway from the vision images. In this paper, a lane sensing algorithm using vision sensors is developed to improve the sensing robustness. The parallel stereo-camera is utilized to regenerate the 3-dimensional road geometry. The lane geometry models are derived such that their parameters represent the road curvature, lateral offset and heading angle, respectively. The parameters of the lane geometry models are estimated by the Kalman filter and utilized to reconstruct the lane geometry in the global coordinate. The inverse perspective mapping from the image plane to the global coordinate considers roll and pitch motions of a vehicle so that the mapping error is minimized during acceleration, braking or steering. The proposed sensing system has been built and implemented on a 1/10-scale model car.