• Journal of KIISE:Software and Applications
• /
• v.36 no.11
• /
• pp.901-911
• /
• 2009

In this paper, we propose a construction technique of two-level composite gradient map to automatically track a face with large movement in successive frames. Our method is composed of three main steps. First, the gradient maps with two-level resolution are generated for fast convergence of active contour. Second, to recognize the variations of face between successive frames and remove the neighbor background, weighted composite gradient map is generated by combining the composite gradient map and difference mask of previous and current frames. Third, to prevent active contour from converging local minima, the energy slope is generated by using closing operation. In addition, the fast closing operation is proposed to accelerate the processing time of closing operation. For performance evaluation, we compare our method with previous active contour model-based face tracking methods using a visual inspection, robustness test and processing time. Experimental results show that our method can effectively track the face with large movement and robustly converge to the optimal position even in frames with complicated background.

• The KIPS Transactions:PartB
• /
• v.17B no.4
• /
• pp.303-308
• /
• 2010

In this paper, we propose a medical image registration technique combining the gradient vector flow and modified conditional entropy. The registration is conducted by the use of a measure based on the entropy of conditional probabilities. To achieve the registration, we first define a modified conditional entropy (MCE) computed from the joint histograms for the area intensities of two given images. In order to combine the spatial information into a traditional registration measure, we use the gradient vector flow field. Then the MCE is computed from the gradient vector flow intensity (GVFI) combining the gradient information and their intensity values of original images. To evaluate the performance of the proposed registration method, we conduct experiments with our method as well as existing method based on the mutual information (MI) criteria. We evaluate the precision of MI- and MCE-based measurements by comparing the registration obtained from MR images and transformed CT images. The experimental results show that the proposed method is faster and more accurate than other optimization methods.

• The Journal of Korean Institute of Communications and Information Sciences
• /
• v.23 no.4
• /
• pp.787-795
• /
• 1998

In this paper, we have improved the speed of gradient operation, which needs to calculate Optical Flow for estimating a moving object, and proposed a method which estimate the contour of a moving object by the logical operationg of Optical Flow and edge in noisy images. The proposed method, which recognize to a moving ogject and traking a moving object, using logical operation of Optical Flow and edge in low-level has a advantage that is simpler than the known method for moving objects estimation. In addition, we have simulated several images using method I and method II on improved Gradient operation speed. When we have compared the average value of total operation time, method I is improved with 12% of operation speed compared with the known method, method II is improved with 38% operation speed.

• Journal of the Institute of Electronics Engineers of Korea CI
• /
• v.40 no.6
• /
• pp.95-105
• /
• 2003

A novel reference point detection method is proposed by exploiting tile gradient probabilistic model that captures the curvature information of fingerprint. The detection of reference point is accomplished through searching and locating the points of occurrence of the most evenly distributed gradient in a probabilistic sense. The uniformly distributed gradient texture represents either the core point itself or those of similar points that can be used to establish the rigid reference from which to map the features for recognition. Key benefits are reductions in preprocessing and consistency of locating the same points as the reference points even when processing arch type fingerprints. Moreover, the new feature extraction method is proposed by improving the existing feature extraction using filterbank method. Experimental results indicate the superiority of tile proposed scheme in terms of computational time in feature extraction and verification rate in various noisy environments. In particular, the proposed gradient probabilistic model achieved 49% improvement under ambient noise, 39.2% under brightness noise and 15.7% under a salt and pepper noise environment, respectively, in FAR for the arch type fingerprints. Moreover, a reduction of 0.07sec in reference point detection time of the GPM is shown possible compared to using the leading the poincare index method and a reduction of 0.06sec in code extraction time of the new filterbank mettled is shown possible compared to using the leading the existing filterbank method.

• Journal of the Korean Data and Information Science Society
• /
• v.28 no.6
• /
• pp.1245-1255
• /
• 2017

The online learning is a process of obtaining the solution for a given objective function where the data is accumulated in real time or in batch units. The stochastic gradient descent method is one of the most widely used for the online learning. This method is not only easy to implement, but also has good properties of the solution under the assumption that the generating model of data is homogeneous. However, the stochastic gradient method could severely mislead the online-learning when the homogeneity is actually violated. We assume that there are two heterogeneous generating models in the observation, and propose the a new stochastic gradient method that mitigate the problem of the heterogeneous models. We introduce a robust mini-batch optimization method using statistical tests and investigate the convergence radius of the solution in the proposed method. Moreover, the theoretical results are confirmed by the numerical simulations.

• Journal of Biomedical Engineering Research
• /
• v.17 no.1
• /
• pp.19-24
• /
• 1996

The purpose of this paper is to design high-order (or radial) surface gradient coil (SGC), which can provide multi-dimensional spatial selection. Although the spatial Selection with High-Order gradienT (SHOT) can provide a 2-D selection with only one selective RF pulse, the high-order gradient pro- duced by conventional cylindrical-shape coils has not been clinically useful due to the large selection size caused by the limited radial gradient intensity. However, by using the proposed high-order SGCs located near the imaging region, the size of volume selection can be reduced to a clinically useflll size of 1-2 cm in diameter by applying stronger radial gradient field with much less gradient driving power. So far radial SGCs have been designed by using the field component method and may cause distortion in the selection shapes. In this paper, by using the target field approach for the coil design, selected volumes became almost circular. A 40 cm-by-40 cm $z^2$_surface gradient coil has been designed and implemented by using the target field approach. Phantom and volunteer studies have been performed Experimental results using spatially localized MRI show good agreement to the theoretically predicted behavior.

• The Transactions of The Korean Institute of Electrical Engineers
• /
• v.56 no.1
• /
• pp.197-205
• /
• 2007

We have proposed a novel line detection method based on the estimated probability density function of gradient directions of edges. By estimating peaks of the density function, we determine groups of edges that have the same gradient direction. For edges in the same groups, we detect lines that correspond to peaks of the connectivity weighted distribution of the distances from the origin. In the experiments using the Data Matrix barcode images and LCD images, the proposed method showed better performance than conventional Methods in terms of the processing speed and accuracy.

• Proceedings of the KIEE Conference
• /
• 1987.07b
• /
• pp.1595-1597
• /
• 1987

Two gradient methods, steepest descent method and conjugate gradient descent method, are compar ed through application to vector linear predictors. It is found that the convergence rate of the conju-gate gradient descent method is much faster than that of the steepest descent method.

• Investigative Magnetic Resonance Imaging
• /
• v.13 no.1
• /
• pp.15-21
• /
• 2009

Purpose : A method to estimate a real k-space trajectory based on a circuit model of the gradient system is proposed for spiral imaging. The estimated k-space trajectory instead of the ideal trajectory is used in the reconstruction to improve the image quality in the spiral imaging. Materials and Methods : Since the gradient system has self resistance, capacitance, and inductance, as well as the mutual inductance between the magnet and the gradient coils, the generated gradient fields have delays and transient responses compared to the input waveform to the gradient system. The real gradient fields and their trajectory in k-space play an important role in the reconstruction. In this paper, the gradient system is modeled with R-L-C circuits, and real gradient fields are estimated from the input to the model. An experimental method to determine the model parameters (R, L, C values) is also suggested from the quality of the reconstructed image. Results : The gradient fields are estimated from the circuit model of the gradient system at 1.5 Tesla MRI system. The spiral trajectory obtained by the integration of the estimated gradient fields is used for the reconstruction. From experiments, the reconstructed images using the estimated trajectory show improved uniformity, reduced overshoots near the edges, and enhanced resolutions compared to those using the ideal trajectory without model. Conclusion : The gradient system was successfully modeled by the R-L-C circuits. Much improved reconstruction was achieved in the spiral imaging using the trajectory estimated by the proposed model.

• The Journal of Korean Institute of Communications and Information Sciences
• /
• v.31 no.9C
• /
• pp.882-888
• /
• 2006

Multiuser detection can significantly increase system capacity and improve service quality compared with the existing matched filter. In this paper, we introduce an method which efficiently calculates the maximum likelihood (ML) metric based on the gradient search (GS). The ML detection needs user powers as well as their spreading codes. A method is also proposed that allows us to detect data bits with the estimation of user powers when they are unknown. Computer simulation shows that the proposed method can nearly achieve the same performance as the GS with perfectly hewn user powers.